|
{ |
|
"results": { |
|
"arc_challenge": { |
|
"acc,none": 0.2883959044368601, |
|
"acc_stderr,none": 0.01323839442242817, |
|
"acc_norm,none": 0.318259385665529, |
|
"acc_norm_stderr,none": 0.013611993916971451 |
|
}, |
|
"arc_easy": { |
|
"acc,none": 0.6123737373737373, |
|
"acc_stderr,none": 0.009997307914447608, |
|
"acc_norm,none": 0.5328282828282829, |
|
"acc_norm_stderr,none": 0.010237645778853858 |
|
}, |
|
"lambada_openai": { |
|
"perplexity,none": 8.77834507332179, |
|
"perplexity_stderr,none": 0.23405401593899378, |
|
"acc,none": 0.5783039006404036, |
|
"acc_stderr,none": 0.006880023561938465 |
|
}, |
|
"logiqa": { |
|
"acc,none": 0.21505376344086022, |
|
"acc_stderr,none": 0.01611524086412917, |
|
"acc_norm,none": 0.282642089093702, |
|
"acc_norm_stderr,none": 0.017661585370360625 |
|
}, |
|
"piqa": { |
|
"acc,none": 0.7176278563656148, |
|
"acc_stderr,none": 0.010502821668555377, |
|
"acc_norm,none": 0.7176278563656148, |
|
"acc_norm_stderr,none": 0.010502821668555377 |
|
}, |
|
"sciq": { |
|
"acc,none": 0.859, |
|
"acc_stderr,none": 0.011010914595992443, |
|
"acc_norm,none": 0.779, |
|
"acc_norm_stderr,none": 0.013127502859696244 |
|
}, |
|
"winogrande": { |
|
"acc,none": 0.595895816890292, |
|
"acc_stderr,none": 0.013791610664670847 |
|
}, |
|
"wsc": { |
|
"acc,none": 0.36538461538461536, |
|
"acc_stderr,none": 0.0474473339327792 |
|
} |
|
}, |
|
"configs": { |
|
"arc_challenge": { |
|
"task": "arc_challenge", |
|
"group": [ |
|
"ai2_arc" |
|
], |
|
"dataset_path": "ai2_arc", |
|
"dataset_name": "ARC-Challenge", |
|
"training_split": "train", |
|
"validation_split": "validation", |
|
"test_split": "test", |
|
"doc_to_text": "Question: {{question}}\nAnswer:", |
|
"doc_to_target": "{{choices.label.index(answerKey)}}", |
|
"doc_to_choice": "{{choices.text}}", |
|
"description": "", |
|
"target_delimiter": " ", |
|
"fewshot_delimiter": "\n\n", |
|
"num_fewshot": 0, |
|
"metric_list": [ |
|
{ |
|
"metric": "acc", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
}, |
|
{ |
|
"metric": "acc_norm", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
} |
|
], |
|
"output_type": "multiple_choice", |
|
"repeats": 1, |
|
"should_decontaminate": true, |
|
"doc_to_decontamination_query": "Question: {{question}}\nAnswer:" |
|
}, |
|
"arc_easy": { |
|
"task": "arc_easy", |
|
"group": [ |
|
"ai2_arc" |
|
], |
|
"dataset_path": "ai2_arc", |
|
"dataset_name": "ARC-Easy", |
|
"training_split": "train", |
|
"validation_split": "validation", |
|
"test_split": "test", |
|
"doc_to_text": "Question: {{question}}\nAnswer:", |
|
"doc_to_target": "{{choices.label.index(answerKey)}}", |
|
"doc_to_choice": "{{choices.text}}", |
|
"description": "", |
|
"target_delimiter": " ", |
|
"fewshot_delimiter": "\n\n", |
|
"num_fewshot": 0, |
|
"metric_list": [ |
|
{ |
|
"metric": "acc", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
}, |
|
{ |
|
"metric": "acc_norm", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
} |
|
], |
|
"output_type": "multiple_choice", |
|
"repeats": 1, |
|
"should_decontaminate": true, |
|
"doc_to_decontamination_query": "Question: {{question}}\nAnswer:" |
|
}, |
|
"lambada_openai": { |
|
"task": "lambada_openai", |
|
"group": [ |
|
"lambada" |
|
], |
|
"dataset_path": "EleutherAI/lambada_openai", |
|
"dataset_name": "default", |
|
"test_split": "test", |
|
"doc_to_text": "{{text.split(' ')[:-1]|join(' ')}}", |
|
"doc_to_target": "{{' '+text.split(' ')[-1]}}", |
|
"description": "", |
|
"target_delimiter": " ", |
|
"fewshot_delimiter": "\n\n", |
|
"num_fewshot": 0, |
|
"metric_list": [ |
|
{ |
|
"metric": "perplexity", |
|
"aggregation": "perplexity", |
|
"higher_is_better": false |
|
}, |
|
{ |
|
"metric": "acc", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
} |
|
], |
|
"output_type": "loglikelihood", |
|
"repeats": 1, |
|
"should_decontaminate": true, |
|
"doc_to_decontamination_query": "{{text}}" |
|
}, |
|
"logiqa": { |
|
"task": "logiqa", |
|
"dataset_path": "EleutherAI/logiqa", |
|
"dataset_name": "logiqa", |
|
"training_split": "train", |
|
"validation_split": "validation", |
|
"test_split": "test", |
|
"doc_to_text": "<function doc_to_text at 0x7f6c84a4cdc0>", |
|
"doc_to_target": "<function doc_to_target at 0x7f6c84a4d120>", |
|
"doc_to_choice": "{{options}}", |
|
"description": "", |
|
"target_delimiter": " ", |
|
"fewshot_delimiter": "\n\n", |
|
"num_fewshot": 0, |
|
"metric_list": [ |
|
{ |
|
"metric": "acc", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
}, |
|
{ |
|
"metric": "acc_norm", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
} |
|
], |
|
"output_type": "multiple_choice", |
|
"repeats": 1, |
|
"should_decontaminate": true, |
|
"doc_to_decontamination_query": "{{context}}" |
|
}, |
|
"piqa": { |
|
"task": "piqa", |
|
"dataset_path": "piqa", |
|
"training_split": "train", |
|
"validation_split": "validation", |
|
"doc_to_text": "Question: {{goal}}\nAnswer:", |
|
"doc_to_target": "label", |
|
"doc_to_choice": "{{[sol1, sol2]}}", |
|
"description": "", |
|
"target_delimiter": " ", |
|
"fewshot_delimiter": "\n\n", |
|
"num_fewshot": 0, |
|
"metric_list": [ |
|
{ |
|
"metric": "acc", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
}, |
|
{ |
|
"metric": "acc_norm", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
} |
|
], |
|
"output_type": "multiple_choice", |
|
"repeats": 1, |
|
"should_decontaminate": true, |
|
"doc_to_decontamination_query": "goal" |
|
}, |
|
"sciq": { |
|
"task": "sciq", |
|
"dataset_path": "sciq", |
|
"training_split": "train", |
|
"validation_split": "validation", |
|
"test_split": "test", |
|
"doc_to_text": "{{support.lstrip()}}\nQuestion: {{question}}\nAnswer:", |
|
"doc_to_target": 3, |
|
"doc_to_choice": "{{[distractor1, distractor2, distractor3, correct_answer]}}", |
|
"description": "", |
|
"target_delimiter": " ", |
|
"fewshot_delimiter": "\n\n", |
|
"num_fewshot": 0, |
|
"metric_list": [ |
|
{ |
|
"metric": "acc", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
}, |
|
{ |
|
"metric": "acc_norm", |
|
"aggregation": "mean", |
|
"higher_is_better": true |
|
} |
|
], |
|
"output_type": "multiple_choice", |
|
"repeats": 1, |
|
"should_decontaminate": true, |
|
"doc_to_decontamination_query": "{{support}} {{question}}" |
|
}, |
|
"winogrande": { |
|
"task": "winogrande", |
|
"dataset_path": "winogrande", |
|
"dataset_name": "winogrande_xl", |
|
"training_split": "train", |
|
"validation_split": "validation", |
|
"doc_to_text": "<function doc_to_text at 0x7f6c84ac1000>", |
|
"doc_to_target": "<function doc_to_target at 0x7f6c84ac1360>", |
|
"doc_to_choice": "<function doc_to_choice at 0x7f6c84ac16c0>", |
|
"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": true, |
|
"doc_to_decontamination_query": "sentence" |
|
}, |
|
"wsc": { |
|
"task": "wsc", |
|
"group": [ |
|
"super-glue-lm-eval-v1" |
|
], |
|
"dataset_path": "super_glue", |
|
"dataset_name": "wsc.fixed", |
|
"training_split": "train", |
|
"validation_split": "validation", |
|
"doc_to_text": "<function default_doc_to_text at 0x7f6c84b48a60>", |
|
"doc_to_target": "label", |
|
"doc_to_choice": [ |
|
"no", |
|
"yes" |
|
], |
|
"description": "", |
|
"target_delimiter": " ", |
|
"fewshot_delimiter": "\n\n", |
|
"num_fewshot": 0, |
|
"metric_list": [ |
|
{ |
|
"metric": "acc" |
|
} |
|
], |
|
"output_type": "multiple_choice", |
|
"repeats": 1, |
|
"should_decontaminate": false |
|
} |
|
}, |
|
"versions": { |
|
"arc_challenge": "Yaml", |
|
"arc_easy": "Yaml", |
|
"lambada_openai": "Yaml", |
|
"logiqa": "Yaml", |
|
"piqa": "Yaml", |
|
"sciq": "Yaml", |
|
"winogrande": "Yaml", |
|
"wsc": "Yaml" |
|
}, |
|
"config": { |
|
"model": "hf", |
|
"model_args": "pretrained=usvsnsp/pythia-2.8b-ppo,dtype=float16", |
|
"batch_size": 1, |
|
"batch_sizes": [], |
|
"device": null, |
|
"use_cache": null, |
|
"limit": null, |
|
"bootstrap_iters": 100000 |
|
}, |
|
"git_hash": "0aa37743" |
|
} |