404 lines
11 KiB
JSON
404 lines
11 KiB
JSON
|
{
|
||
|
|
"results": {
|
||
|
|
"arc_challenge": {
|
||
|
|
"acc,none": 0.31399317406143346,
|
||
|
|
"acc_stderr,none": 0.013562691224726305,
|
||
|
|
"acc_norm,none": 0.3515358361774744,
|
||
|
|
"acc_norm_stderr,none": 0.013952413699600938
|
||
|
|
},
|
||
|
|
"arc_easy": {
|
||
|
|
"acc,none": 0.6704545454545454,
|
||
|
|
"acc_stderr,none": 0.00964518419095386,
|
||
|
|
"acc_norm,none": 0.6127946127946128,
|
||
|
|
"acc_norm_stderr,none": 0.009995312065890353
|
||
|
|
},
|
||
|
|
"boolq": {
|
||
|
|
"acc,none": 0.6351681957186545,
|
||
|
|
"acc_stderr,none": 0.00841944098496366
|
||
|
|
},
|
||
|
|
"hellaswag": {
|
||
|
|
"acc,none": 0.4797849034056961,
|
||
|
|
"acc_stderr,none": 0.004985701593897999,
|
||
|
|
"acc_norm,none": 0.6389165504879506,
|
||
|
|
"acc_norm_stderr,none": 0.0047933305256562106
|
||
|
|
},
|
||
|
|
"lambada_openai": {
|
||
|
|
"perplexity,none": 4.456621453732154,
|
||
|
|
"perplexity_stderr,none": 0.10000881772967252,
|
||
|
|
"acc,none": 0.6712594605084417,
|
||
|
|
"acc_stderr,none": 0.006544612151352774
|
||
|
|
},
|
||
|
|
"openbookqa": {
|
||
|
|
"acc,none": 0.256,
|
||
|
|
"acc_stderr,none": 0.019536923574747605,
|
||
|
|
"acc_norm,none": 0.372,
|
||
|
|
"acc_norm_stderr,none": 0.0216371979857224
|
||
|
|
},
|
||
|
|
"piqa": {
|
||
|
|
"acc,none": 0.7524483133841132,
|
||
|
|
"acc_stderr,none": 0.010069703966857088,
|
||
|
|
"acc_norm,none": 0.763873775843308,
|
||
|
|
"acc_norm_stderr,none": 0.009908965890558214
|
||
|
|
},
|
||
|
|
"sciq": {
|
||
|
|
"acc,none": 0.893,
|
||
|
|
"acc_stderr,none": 0.009779910359847167,
|
||
|
|
"acc_norm,none": 0.838,
|
||
|
|
"acc_norm_stderr,none": 0.011657267771304427
|
||
|
|
},
|
||
|
|
"wikitext": {
|
||
|
|
"word_perplexity,none": 17.682958079421635,
|
||
|
|
"byte_perplexity,none": 1.6057045697141277,
|
||
|
|
"bits_per_byte,none": 0.6832064787735104
|
||
|
|
},
|
||
|
|
"winogrande": {
|
||
|
|
"acc,none": 0.6077348066298343,
|
||
|
|
"acc_stderr,none": 0.013722400462000888
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"configs": {
|
||
|
|
"arc_challenge": {
|
||
|
|
"task": "arc_challenge",
|
||
|
|
"group": [
|
||
|
|
"ai2_arc",
|
||
|
|
"multiple_choice"
|
||
|
|
],
|
||
|
|
"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",
|
||
|
|
"multiple_choice"
|
||
|
|
],
|
||
|
|
"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:"
|
||
|
|
},
|
||
|
|
"boolq": {
|
||
|
|
"task": "boolq",
|
||
|
|
"group": [
|
||
|
|
"super-glue-lm-eval-v1"
|
||
|
|
],
|
||
|
|
"dataset_path": "super_glue",
|
||
|
|
"dataset_name": "boolq",
|
||
|
|
"training_split": "train",
|
||
|
|
"validation_split": "validation",
|
||
|
|
"doc_to_text": "{{passage}}\nQuestion: {{question}}?\nAnswer:",
|
||
|
|
"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": true,
|
||
|
|
"doc_to_decontamination_query": "passage"
|
||
|
|
},
|
||
|
|
"hellaswag": {
|
||
|
|
"task": "hellaswag",
|
||
|
|
"group": [
|
||
|
|
"multiple_choice"
|
||
|
|
],
|
||
|
|
"dataset_path": "hellaswag",
|
||
|
|
"training_split": "train",
|
||
|
|
"validation_split": "validation",
|
||
|
|
"doc_to_text": "{% set text = activity_label ~ ': ' ~ ctx_a ~ ' ' ~ ctx_b.capitalize() %}{{text|trim|replace(' [title]', '. ')|regex_replace('\\[.*?\\]', '')|replace(' ', ' ')}}",
|
||
|
|
"doc_to_target": "{{label}}",
|
||
|
|
"doc_to_choice": "{{endings|map('trim')|map('replace', ' [title]', '. ')|map('regex_replace', '\\[.*?\\]', '')|map('replace', ' ', ' ')|list}}",
|
||
|
|
"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": false
|
||
|
|
},
|
||
|
|
"lambada_openai": {
|
||
|
|
"task": "lambada_openai",
|
||
|
|
"group": [
|
||
|
|
"lambada",
|
||
|
|
"loglikelihood",
|
||
|
|
"perplexity"
|
||
|
|
],
|
||
|
|
"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}}"
|
||
|
|
},
|
||
|
|
"openbookqa": {
|
||
|
|
"task": "openbookqa",
|
||
|
|
"group": [
|
||
|
|
"multiple_choice"
|
||
|
|
],
|
||
|
|
"dataset_path": "openbookqa",
|
||
|
|
"dataset_name": "main",
|
||
|
|
"training_split": "train",
|
||
|
|
"validation_split": "validation",
|
||
|
|
"test_split": "test",
|
||
|
|
"doc_to_text": "question_stem",
|
||
|
|
"doc_to_target": "{{choices.label.index(answerKey.lstrip())}}",
|
||
|
|
"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_stem"
|
||
|
|
},
|
||
|
|
"piqa": {
|
||
|
|
"task": "piqa",
|
||
|
|
"group": [
|
||
|
|
"multiple_choice"
|
||
|
|
],
|
||
|
|
"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",
|
||
|
|
"group": [
|
||
|
|
"multiple_choice"
|
||
|
|
],
|
||
|
|
"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}}"
|
||
|
|
},
|
||
|
|
"wikitext": {
|
||
|
|
"task": "wikitext",
|
||
|
|
"group": [
|
||
|
|
"perplexity",
|
||
|
|
"loglikelihood_rolling"
|
||
|
|
],
|
||
|
|
"dataset_path": "EleutherAI/wikitext_document_level",
|
||
|
|
"dataset_name": "wikitext-2-raw-v1",
|
||
|
|
"training_split": "train",
|
||
|
|
"validation_split": "validation",
|
||
|
|
"test_split": "test",
|
||
|
|
"doc_to_text": "",
|
||
|
|
"doc_to_target": "<function wikitext_detokenizer at 0x7ff390401120>",
|
||
|
|
"description": "",
|
||
|
|
"target_delimiter": " ",
|
||
|
|
"fewshot_delimiter": "\n\n",
|
||
|
|
"num_fewshot": 0,
|
||
|
|
"metric_list": [
|
||
|
|
{
|
||
|
|
"metric": "word_perplexity"
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"metric": "byte_perplexity"
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"metric": "bits_per_byte"
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"output_type": "loglikelihood_rolling",
|
||
|
|
"repeats": 1,
|
||
|
|
"should_decontaminate": true,
|
||
|
|
"doc_to_decontamination_query": "{{page}}"
|
||
|
|
},
|
||
|
|
"winogrande": {
|
||
|
|
"task": "winogrande",
|
||
|
|
"dataset_path": "winogrande",
|
||
|
|
"dataset_name": "winogrande_xl",
|
||
|
|
"training_split": "train",
|
||
|
|
"validation_split": "validation",
|
||
|
|
"doc_to_text": "<function doc_to_text at 0x7ff390401360>",
|
||
|
|
"doc_to_target": "<function doc_to_target at 0x7ff3904016c0>",
|
||
|
|
"doc_to_choice": "<function doc_to_choice at 0x7ff390401a20>",
|
||
|
|
"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"
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"versions": {
|
||
|
|
"arc_challenge": "Yaml",
|
||
|
|
"arc_easy": "Yaml",
|
||
|
|
"boolq": "Yaml",
|
||
|
|
"hellaswag": "Yaml",
|
||
|
|
"lambada_openai": "Yaml",
|
||
|
|
"openbookqa": "Yaml",
|
||
|
|
"piqa": "Yaml",
|
||
|
|
"sciq": "Yaml",
|
||
|
|
"wikitext": "Yaml",
|
||
|
|
"winogrande": "Yaml"
|
||
|
|
},
|
||
|
|
"config": {
|
||
|
|
"model": "hf",
|
||
|
|
"model_args": "pretrained=EleutherAI/pythia-6.9b",
|
||
|
|
"batch_size": "4",
|
||
|
|
"batch_sizes": [],
|
||
|
|
"device": null,
|
||
|
|
"use_cache": null,
|
||
|
|
"limit": null,
|
||
|
|
"bootstrap_iters": 100000
|
||
|
|
},
|
||
|
|
"git_hash": "d1a44c8"
|
||
|
|
}
|