cot-eval-results / data /allenai /tulu-2-dpo-7b /orig /results_24-03-26-10:11:02.json
ggbetz's picture
Upload results for model allenai/tulu-2-dpo-7b (#163)
7ae125d verified
raw
history blame
8.77 kB
{
"results": {
"logiqa2_base": {
"acc,none": 0.3276081424936387,
"acc_stderr,none": 0.011841329714669953,
"alias": "logiqa2_base"
},
"logiqa_base": {
"acc,none": 0.3083067092651757,
"acc_stderr,none": 0.018471759300608265,
"alias": "logiqa_base"
},
"lsat-ar_base": {
"acc,none": 0.23478260869565218,
"acc_stderr,none": 0.028009647070930125,
"alias": "lsat-ar_base"
},
"lsat-lr_base": {
"acc,none": 0.26862745098039215,
"acc_stderr,none": 0.01964651988859971,
"alias": "lsat-lr_base"
},
"lsat-rc_base": {
"acc,none": 0.34944237918215615,
"acc_stderr,none": 0.029124821619700387,
"alias": "lsat-rc_base"
}
},
"configs": {
"logiqa2_base": {
"task": "logiqa2_base",
"group": "logikon-bench",
"dataset_path": "logikon/logikon-bench",
"dataset_name": "logiqa2",
"test_split": "test",
"doc_to_text": "def doc_to_text(doc) -> str:\n \"\"\"\n Answer the following question about the given passage.\n \n Passage: <passage>\n \n Question: <question>\n A. <choice1>\n B. <choice2>\n C. <choice3>\n D. <choice4>\n [E. <choice5>]\n \n Answer:\n \"\"\"\n k = len(doc[\"options\"])\n choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n prompt = \"Answer the following question about the given passage.\\n\\n\"\n prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n for choice, option in zip(choices, doc[\"options\"]):\n prompt += f\"{choice.upper()}. {option}\\n\"\n prompt += \"\\n\"\n prompt += \"Answer:\"\n return prompt\n",
"doc_to_target": "{{answer}}",
"doc_to_choice": "{{options}}",
"description": "",
"target_delimiter": " ",
"fewshot_delimiter": "\n\n",
"num_fewshot": 0,
"metric_list": [
{
"metric": "acc",
"aggregation": "mean",
"higher_is_better": true
}
],
"output_type": "multiple_choice",
"repeats": 1,
"should_decontaminate": false,
"metadata": {
"version": 0.0
}
},
"logiqa_base": {
"task": "logiqa_base",
"group": "logikon-bench",
"dataset_path": "logikon/logikon-bench",
"dataset_name": "logiqa",
"test_split": "test",
"doc_to_text": "def doc_to_text(doc) -> str:\n \"\"\"\n Answer the following question about the given passage.\n \n Passage: <passage>\n \n Question: <question>\n A. <choice1>\n B. <choice2>\n C. <choice3>\n D. <choice4>\n [E. <choice5>]\n \n Answer:\n \"\"\"\n k = len(doc[\"options\"])\n choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n prompt = \"Answer the following question about the given passage.\\n\\n\"\n prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n for choice, option in zip(choices, doc[\"options\"]):\n prompt += f\"{choice.upper()}. {option}\\n\"\n prompt += \"\\n\"\n prompt += \"Answer:\"\n return prompt\n",
"doc_to_target": "{{answer}}",
"doc_to_choice": "{{options}}",
"description": "",
"target_delimiter": " ",
"fewshot_delimiter": "\n\n",
"num_fewshot": 0,
"metric_list": [
{
"metric": "acc",
"aggregation": "mean",
"higher_is_better": true
}
],
"output_type": "multiple_choice",
"repeats": 1,
"should_decontaminate": false,
"metadata": {
"version": 0.0
}
},
"lsat-ar_base": {
"task": "lsat-ar_base",
"group": "logikon-bench",
"dataset_path": "logikon/logikon-bench",
"dataset_name": "lsat-ar",
"test_split": "test",
"doc_to_text": "def doc_to_text(doc) -> str:\n \"\"\"\n Answer the following question about the given passage.\n \n Passage: <passage>\n \n Question: <question>\n A. <choice1>\n B. <choice2>\n C. <choice3>\n D. <choice4>\n [E. <choice5>]\n \n Answer:\n \"\"\"\n k = len(doc[\"options\"])\n choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n prompt = \"Answer the following question about the given passage.\\n\\n\"\n prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n for choice, option in zip(choices, doc[\"options\"]):\n prompt += f\"{choice.upper()}. {option}\\n\"\n prompt += \"\\n\"\n prompt += \"Answer:\"\n return prompt\n",
"doc_to_target": "{{answer}}",
"doc_to_choice": "{{options}}",
"description": "",
"target_delimiter": " ",
"fewshot_delimiter": "\n\n",
"num_fewshot": 0,
"metric_list": [
{
"metric": "acc",
"aggregation": "mean",
"higher_is_better": true
}
],
"output_type": "multiple_choice",
"repeats": 1,
"should_decontaminate": false,
"metadata": {
"version": 0.0
}
},
"lsat-lr_base": {
"task": "lsat-lr_base",
"group": "logikon-bench",
"dataset_path": "logikon/logikon-bench",
"dataset_name": "lsat-lr",
"test_split": "test",
"doc_to_text": "def doc_to_text(doc) -> str:\n \"\"\"\n Answer the following question about the given passage.\n \n Passage: <passage>\n \n Question: <question>\n A. <choice1>\n B. <choice2>\n C. <choice3>\n D. <choice4>\n [E. <choice5>]\n \n Answer:\n \"\"\"\n k = len(doc[\"options\"])\n choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n prompt = \"Answer the following question about the given passage.\\n\\n\"\n prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n for choice, option in zip(choices, doc[\"options\"]):\n prompt += f\"{choice.upper()}. {option}\\n\"\n prompt += \"\\n\"\n prompt += \"Answer:\"\n return prompt\n",
"doc_to_target": "{{answer}}",
"doc_to_choice": "{{options}}",
"description": "",
"target_delimiter": " ",
"fewshot_delimiter": "\n\n",
"num_fewshot": 0,
"metric_list": [
{
"metric": "acc",
"aggregation": "mean",
"higher_is_better": true
}
],
"output_type": "multiple_choice",
"repeats": 1,
"should_decontaminate": false,
"metadata": {
"version": 0.0
}
},
"lsat-rc_base": {
"task": "lsat-rc_base",
"group": "logikon-bench",
"dataset_path": "logikon/logikon-bench",
"dataset_name": "lsat-rc",
"test_split": "test",
"doc_to_text": "def doc_to_text(doc) -> str:\n \"\"\"\n Answer the following question about the given passage.\n \n Passage: <passage>\n \n Question: <question>\n A. <choice1>\n B. <choice2>\n C. <choice3>\n D. <choice4>\n [E. <choice5>]\n \n Answer:\n \"\"\"\n k = len(doc[\"options\"])\n choices = [\"a\", \"b\", \"c\", \"d\", \"e\"][:k]\n prompt = \"Answer the following question about the given passage.\\n\\n\"\n prompt = \"Passage: \" + doc[\"passage\"] + \"\\n\\n\"\n prompt += \"Question: \" + doc[\"question\"] + \"\\n\"\n for choice, option in zip(choices, doc[\"options\"]):\n prompt += f\"{choice.upper()}. {option}\\n\"\n prompt += \"\\n\"\n prompt += \"Answer:\"\n return prompt\n",
"doc_to_target": "{{answer}}",
"doc_to_choice": "{{options}}",
"description": "",
"target_delimiter": " ",
"fewshot_delimiter": "\n\n",
"num_fewshot": 0,
"metric_list": [
{
"metric": "acc",
"aggregation": "mean",
"higher_is_better": true
}
],
"output_type": "multiple_choice",
"repeats": 1,
"should_decontaminate": false,
"metadata": {
"version": 0.0
}
}
},
"versions": {
"logiqa2_base": 0.0,
"logiqa_base": 0.0,
"lsat-ar_base": 0.0,
"lsat-lr_base": 0.0,
"lsat-rc_base": 0.0
},
"n-shot": {
"logiqa2_base": 0,
"logiqa_base": 0,
"lsat-ar_base": 0,
"lsat-lr_base": 0,
"lsat-rc_base": 0
},
"config": {
"model": "vllm",
"model_args": "pretrained=allenai/tulu-2-dpo-7b,revision=main,dtype=bfloat16,tensor_parallel_size=1,gpu_memory_utilization=0.8,trust_remote_code=true,max_length=2048",
"batch_size": "auto",
"batch_sizes": [],
"device": null,
"use_cache": null,
"limit": null,
"bootstrap_iters": 100000,
"gen_kwargs": null
},
"git_hash": "741db1c"
}