| { | |
| "help": false, | |
| "algo": "APPO", | |
| "env": "GDY-MettaGrid", | |
| "experiment": "p2.dr6", | |
| "train_dir": "/workspace/metta/train_dir", | |
| "restart_behavior": "resume", | |
| "device": "gpu", | |
| "seed": 0, | |
| "num_policies": 1, | |
| "async_rl": true, | |
| "serial_mode": false, | |
| "batched_sampling": false, | |
| "num_batches_to_accumulate": 2, | |
| "worker_num_splits": 2, | |
| "policy_workers_per_policy": 1, | |
| "max_policy_lag": 50, | |
| "num_workers": 32, | |
| "num_envs_per_worker": 2, | |
| "batch_size": 16384, | |
| "num_batches_per_epoch": 1, | |
| "num_epochs": 1, | |
| "rollout": 256, | |
| "recurrence": 256, | |
| "shuffle_minibatches": false, | |
| "gamma": 0.99, | |
| "reward_scale": 1.0, | |
| "reward_clip": 1000.0, | |
| "value_bootstrap": false, | |
| "normalize_returns": true, | |
| "exploration_loss_coeff": -0.002, | |
| "value_loss_coeff": 0.976, | |
| "kl_loss_coeff": 0.0, | |
| "aux_loss_coeff": 0.0, | |
| "exploration_loss": "symmetric_kl", | |
| "gae_lambda": 0.95, | |
| "ppo_clip_ratio": 0.1, | |
| "ppo_clip_value": 1.0, | |
| "with_vtrace": false, | |
| "vtrace_rho": 1.0, | |
| "vtrace_c": 1.0, | |
| "optimizer": "adam", | |
| "adam_eps": 1e-06, | |
| "adam_beta1": 0.9, | |
| "adam_beta2": 0.999, | |
| "max_grad_norm": 4.0, | |
| "learning_rate": 0.0001, | |
| "lr_schedule": "constant", | |
| "lr_schedule_kl_threshold": 0.008, | |
| "lr_adaptive_min": 1e-06, | |
| "lr_adaptive_max": 0.01, | |
| "obs_subtract_mean": 0.0, | |
| "obs_scale": 1.0, | |
| "normalize_input": false, | |
| "normalize_input_keys": null, | |
| "decorrelate_experience_max_seconds": 150, | |
| "decorrelate_envs_on_one_worker": true, | |
| "actor_worker_gpus": [], | |
| "set_workers_cpu_affinity": true, | |
| "force_envs_single_thread": false, | |
| "default_niceness": 0, | |
| "log_to_file": true, | |
| "experiment_summaries_interval": 10, | |
| "flush_summaries_interval": 30, | |
| "stats_avg": 100, | |
| "summaries_use_frameskip": true, | |
| "heartbeat_interval": 20, | |
| "heartbeat_reporting_interval": 180, | |
| "train_for_env_steps": 9999999999999, | |
| "train_for_seconds": 10000000000, | |
| "save_every_sec": 120, | |
| "keep_checkpoints": 2, | |
| "load_checkpoint_path": null, | |
| "init_checkpoint_path": "train_dir/baseline.v0.5.4/checkpoint_p0/checkpoint_000349863_5732155392.pth", | |
| "load_checkpoint_kind": "latest", | |
| "save_milestones_sec": -1, | |
| "save_best_every_sec": 5, | |
| "save_best_metric": "reward", | |
| "save_best_after": 100000, | |
| "benchmark": false, | |
| "encoder_mlp_layers": [ | |
| 512, | |
| 512 | |
| ], | |
| "encoder_conv_architecture": "convnet_simple", | |
| "encoder_conv_mlp_layers": [ | |
| 512 | |
| ], | |
| "use_rnn": true, | |
| "rnn_size": 512, | |
| "rnn_type": "gru", | |
| "rnn_num_layers": 1, | |
| "decoder_mlp_layers": [], | |
| "nonlinearity": "elu", | |
| "policy_initialization": "orthogonal", | |
| "policy_init_gain": 1.0, | |
| "actor_critic_share_weights": true, | |
| "adaptive_stddev": true, | |
| "continuous_tanh_scale": 0.0, | |
| "initial_stddev": 1.0, | |
| "use_env_info_cache": false, | |
| "env_gpu_actions": false, | |
| "env_gpu_observations": true, | |
| "env_frameskip": 1, | |
| "env_framestack": 1, | |
| "pixel_format": "CHW", | |
| "use_record_episode_statistics": false, | |
| "episode_counter": false, | |
| "with_wandb": true, | |
| "wandb_user": "platypus", | |
| "wandb_project": "metta", | |
| "wandb_group": null, | |
| "wandb_job_type": "SF", | |
| "wandb_tags": [], | |
| "with_pbt": false, | |
| "pbt_mix_policies_in_one_env": true, | |
| "pbt_period_env_steps": 5000000, | |
| "pbt_start_mutation": 20000000, | |
| "pbt_replace_fraction": 0.3, | |
| "pbt_mutation_rate": 0.15, | |
| "pbt_replace_reward_gap": 0.1, | |
| "pbt_replace_reward_gap_absolute": 1e-06, | |
| "pbt_optimize_gamma": false, | |
| "pbt_target_objective": "true_objective", | |
| "pbt_perturb_min": 1.1, | |
| "pbt_perturb_max": 1.5, | |
| "env_cfg": "{\"name\": \"GDY-MettaGrid\", \"_target_\": \"env.griddly.mettagrid.gym_env.MettaGridGymEnv\", \"max_action_value\": 10, \"hidden_features\": {\"grid_obs\": [\"agent:energy\", \"agent:hp\"]}, \"game\": {\"obs_width\": 11, \"obs_height\": 11, \"max_steps\": 5000, \"tile_size\": 16, \"num_agents\": 20, \"no_energy_steps\": 500, \"objects\": {\"agent\": {\"initial_energy\": [50, 250], \"max_energy\": [200, 250], \"max_inventory\": [3, 10], \"freeze_duration\": [5, 20], \"energy_reward\": false, \"hp\": 1, \"mortal\": false, \"upkeep\": {\"time\": 0, \"shield\": [0, 3]}, \"use_cost\": 0}, \"altar\": {\"hp\": [10, 50], \"cooldown\": [2, 5], \"use_cost\": [50, 250]}, \"converter\": {\"hp\": [10, 50], \"cooldown\": [2, 5], \"energy_output\": [50, 200], \"use_cost\": 0}, \"generator\": {\"hp\": [10, 50], \"cooldown\": [3, 10], \"initial_resources\": [5, 50], \"use_cost\": 0}, \"wall\": {\"density\": 0.01, \"hp\": [3, 20]}}, \"actions\": {\"move\": {\"cost\": [0, 2]}, \"rotate\": {\"cost\": [0, 1]}, \"jump\": {\"cost\": [1, 5]}, \"shield\": {\"cost\": [0, 2]}, \"drop\": {\"cost\": [0, 2]}, \"use\": {\"cost\": [0, 2]}, \"attack\": {\"cost\": [3, 10], \"damage\": [3, 10]}}, \"map\": {\"layout\": {\"rooms_x\": 2, \"rooms_y\": 2}, \"room\": {\"width\": 20, \"height\": 20, \"num_agents\": 5, \"objects\": {\"agent\": 5, \"altar\": 1, \"converter\": 3, \"generator\": 10, \"wall\": 40}, \"border\": 1}}}, \"kinship\": {\"team_size\": [1, 5], \"team_reward\": [0.0, 1.0]}}", | |
| "agent_cfg": "{\"_target_\": \"agent.metta_agent.MettaAgent\", \"observation_encoders\": {\"grid_obs\": {\"feature_names\": [], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 512, \"layers\": 4}, \"global_vars\": {\"feature_names\": [], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}, \"last_action\": {\"feature_names\": [\"last_action_id\", \"last_action_val\"], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}, \"last_reward\": {\"feature_names\": [\"last_reward\"], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}, \"kinship\": {\"feature_names\": [\"kinship\"], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}}, \"fc\": {\"layers\": 3, \"output_dim\": 512}, \"decoder\": {\"_target_\": \"agent.decoder.Decoder\"}, \"core\": {\"_target_\": \"sample_factory.model.core.ModelCoreRNN\", \"rnn_type\": \"gru\", \"rnn_num_layers\": 1, \"rnn_size\": 512}}", | |
| "command_line": "--aux_loss_coeff=0 --recurrence=256 --rollout=256 --value_loss_coeff=0.976 --exploration_loss=symmetric_kl --exploration_loss_coeff=0.002 --policy_initialization=orthogonal --learning_rate=0.0001 --max_policy_lag=50 --nonlinearity=elu --load_checkpoint_kind=latest --normalize_input=False --seed=0 --batch_size=16384 --decorrelate_experience_max_seconds=150 --train_for_env_steps=9999999999999 --with_wandb=True --wandb_user=platypus --wandb_project=metta --experiment=p2.dr6 --init_checkpoint_path=train_dir/baseline.v0.5.4/checkpoint_p0/checkpoint_000349863_5732155392.pth --rnn_type=gru --rnn_num_layers=1 --rnn_size=512 --env=GDY-MettaGrid --env_cfg={\"name\": \"GDY-MettaGrid\", \"_target_\": \"env.griddly.mettagrid.gym_env.MettaGridGymEnv\", \"max_action_value\": 10, \"hidden_features\": {\"grid_obs\": [\"agent:energy\", \"agent:hp\"]}, \"game\": {\"obs_width\": 11, \"obs_height\": 11, \"max_steps\": 5000, \"tile_size\": 16, \"num_agents\": 20, \"no_energy_steps\": 500, \"objects\": {\"agent\": {\"initial_energy\": [50, 250], \"max_energy\": [200, 250], \"max_inventory\": [3, 10], \"freeze_duration\": [5, 20], \"energy_reward\": false, \"hp\": 1, \"mortal\": false, \"upkeep\": {\"time\": 0, \"shield\": [0, 3]}, \"use_cost\": 0}, \"altar\": {\"hp\": [10, 50], \"cooldown\": [2, 5], \"use_cost\": [50, 250]}, \"converter\": {\"hp\": [10, 50], \"cooldown\": [2, 5], \"energy_output\": [50, 200], \"use_cost\": 0}, \"generator\": {\"hp\": [10, 50], \"cooldown\": [3, 10], \"initial_resources\": [5, 50], \"use_cost\": 0}, \"wall\": {\"density\": 0.01, \"hp\": [3, 20]}}, \"actions\": {\"move\": {\"cost\": [0, 2]}, \"rotate\": {\"cost\": [0, 1]}, \"jump\": {\"cost\": [1, 5]}, \"shield\": {\"cost\": [0, 2]}, \"drop\": {\"cost\": [0, 2]}, \"use\": {\"cost\": [0, 2]}, \"attack\": {\"cost\": [3, 10], \"damage\": [3, 10]}}, \"map\": {\"layout\": {\"rooms_x\": 2, \"rooms_y\": 2}, \"room\": {\"width\": 20, \"height\": 20, \"num_agents\": 5, \"objects\": {\"agent\": 5, \"altar\": 1, \"converter\": 3, \"generator\": 10, \"wall\": 40}, \"border\": 1}}}, \"kinship\": {\"team_size\": [1, 5], \"team_reward\": [0.0, 1.0]}} --agent_cfg={\"_target_\": \"agent.metta_agent.MettaAgent\", \"observation_encoders\": {\"grid_obs\": {\"feature_names\": [], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 512, \"layers\": 4}, \"global_vars\": {\"feature_names\": [], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}, \"last_action\": {\"feature_names\": [\"last_action_id\", \"last_action_val\"], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}, \"last_reward\": {\"feature_names\": [\"last_reward\"], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}, \"kinship\": {\"feature_names\": [\"kinship\"], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}}, \"fc\": {\"layers\": 3, \"output_dim\": 512}, \"decoder\": {\"_target_\": \"agent.decoder.Decoder\"}, \"core\": {\"_target_\": \"sample_factory.model.core.ModelCoreRNN\", \"rnn_type\": \"gru\", \"rnn_num_layers\": 1, \"rnn_size\": 512}}", | |
| "cli_args": { | |
| "env": "GDY-MettaGrid", | |
| "experiment": "p2.dr6", | |
| "seed": 0, | |
| "max_policy_lag": 50, | |
| "batch_size": 16384, | |
| "rollout": 256, | |
| "recurrence": 256, | |
| "exploration_loss_coeff": 0.002, | |
| "value_loss_coeff": 0.976, | |
| "aux_loss_coeff": 0.0, | |
| "exploration_loss": "symmetric_kl", | |
| "learning_rate": 0.0001, | |
| "normalize_input": false, | |
| "decorrelate_experience_max_seconds": 150, | |
| "train_for_env_steps": 9999999999999, | |
| "init_checkpoint_path": "train_dir/baseline.v0.5.4/checkpoint_p0/checkpoint_000349863_5732155392.pth", | |
| "load_checkpoint_kind": "latest", | |
| "rnn_size": 512, | |
| "rnn_type": "gru", | |
| "rnn_num_layers": 1, | |
| "nonlinearity": "elu", | |
| "policy_initialization": "orthogonal", | |
| "with_wandb": true, | |
| "wandb_user": "platypus", | |
| "wandb_project": "metta", | |
| "env_cfg": "{\"name\": \"GDY-MettaGrid\", \"_target_\": \"env.griddly.mettagrid.gym_env.MettaGridGymEnv\", \"max_action_value\": 10, \"hidden_features\": {\"grid_obs\": [\"agent:energy\", \"agent:hp\"]}, \"game\": {\"obs_width\": 11, \"obs_height\": 11, \"max_steps\": 5000, \"tile_size\": 16, \"num_agents\": 20, \"no_energy_steps\": 500, \"objects\": {\"agent\": {\"initial_energy\": [50, 250], \"max_energy\": [200, 250], \"max_inventory\": [3, 10], \"freeze_duration\": [5, 20], \"energy_reward\": false, \"hp\": 1, \"mortal\": false, \"upkeep\": {\"time\": 0, \"shield\": [0, 3]}, \"use_cost\": 0}, \"altar\": {\"hp\": [10, 50], \"cooldown\": [2, 5], \"use_cost\": [50, 250]}, \"converter\": {\"hp\": [10, 50], \"cooldown\": [2, 5], \"energy_output\": [50, 200], \"use_cost\": 0}, \"generator\": {\"hp\": [10, 50], \"cooldown\": [3, 10], \"initial_resources\": [5, 50], \"use_cost\": 0}, \"wall\": {\"density\": 0.01, \"hp\": [3, 20]}}, \"actions\": {\"move\": {\"cost\": [0, 2]}, \"rotate\": {\"cost\": [0, 1]}, \"jump\": {\"cost\": [1, 5]}, \"shield\": {\"cost\": [0, 2]}, \"drop\": {\"cost\": [0, 2]}, \"use\": {\"cost\": [0, 2]}, \"attack\": {\"cost\": [3, 10], \"damage\": [3, 10]}}, \"map\": {\"layout\": {\"rooms_x\": 2, \"rooms_y\": 2}, \"room\": {\"width\": 20, \"height\": 20, \"num_agents\": 5, \"objects\": {\"agent\": 5, \"altar\": 1, \"converter\": 3, \"generator\": 10, \"wall\": 40}, \"border\": 1}}}, \"kinship\": {\"team_size\": [1, 5], \"team_reward\": [0.0, 1.0]}}", | |
| "agent_cfg": "{\"_target_\": \"agent.metta_agent.MettaAgent\", \"observation_encoders\": {\"grid_obs\": {\"feature_names\": [], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 512, \"layers\": 4}, \"global_vars\": {\"feature_names\": [], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}, \"last_action\": {\"feature_names\": [\"last_action_id\", \"last_action_val\"], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}, \"last_reward\": {\"feature_names\": [\"last_reward\"], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}, \"kinship\": {\"feature_names\": [\"kinship\"], \"normalize_features\": true, \"label_dim\": 4, \"output_dim\": 8, \"layers\": 2}}, \"fc\": {\"layers\": 3, \"output_dim\": 512}, \"decoder\": {\"_target_\": \"agent.decoder.Decoder\"}, \"core\": {\"_target_\": \"sample_factory.model.core.ModelCoreRNN\", \"rnn_type\": \"gru\", \"rnn_num_layers\": 1, \"rnn_size\": 512}}" | |
| }, | |
| "git_hash": "bd6dd5e8f31642adff63d12f41f4ddd3333616af", | |
| "git_repo_name": "https://github.com/daveey/metta.git", | |
| "wandb_unique_id": "p2.dr6_20240621_135940_570912" | |
| } |