lomahony/eleuther-pythia6.9b-hh-dpo
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: lomahony/eleuther-pythia6.9b-hh-dpo
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-06-08 22:45:40 +08:00
commit 8f058deb86
15 changed files with 102335 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

35
.gitattributes vendored Normal file
View File

@@ -0,0 +1,35 @@
*.7z filter=lfs diff=lfs merge=lfs -text
*.arrow filter=lfs diff=lfs merge=lfs -text
*.bin filter=lfs diff=lfs merge=lfs -text
*.bz2 filter=lfs diff=lfs merge=lfs -text
*.ckpt filter=lfs diff=lfs merge=lfs -text
*.ftz filter=lfs diff=lfs merge=lfs -text
*.gz filter=lfs diff=lfs merge=lfs -text
*.h5 filter=lfs diff=lfs merge=lfs -text
*.joblib filter=lfs diff=lfs merge=lfs -text
*.lfs.* filter=lfs diff=lfs merge=lfs -text
*.mlmodel filter=lfs diff=lfs merge=lfs -text
*.model filter=lfs diff=lfs merge=lfs -text
*.msgpack filter=lfs diff=lfs merge=lfs -text
*.npy filter=lfs diff=lfs merge=lfs -text
*.npz filter=lfs diff=lfs merge=lfs -text
*.onnx filter=lfs diff=lfs merge=lfs -text
*.ot filter=lfs diff=lfs merge=lfs -text
*.parquet filter=lfs diff=lfs merge=lfs -text
*.pb filter=lfs diff=lfs merge=lfs -text
*.pickle filter=lfs diff=lfs merge=lfs -text
*.pkl filter=lfs diff=lfs merge=lfs -text
*.pt filter=lfs diff=lfs merge=lfs -text
*.pth filter=lfs diff=lfs merge=lfs -text
*.rar filter=lfs diff=lfs merge=lfs -text
*.safetensors filter=lfs diff=lfs merge=lfs -text
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
*.tar.* filter=lfs diff=lfs merge=lfs -text
*.tar filter=lfs diff=lfs merge=lfs -text
*.tflite filter=lfs diff=lfs merge=lfs -text
*.tgz filter=lfs diff=lfs merge=lfs -text
*.wasm filter=lfs diff=lfs merge=lfs -text
*.xz filter=lfs diff=lfs merge=lfs -text
*.zip filter=lfs diff=lfs merge=lfs -text
*.zst filter=lfs diff=lfs merge=lfs -text
*tfevents* filter=lfs diff=lfs merge=lfs -text

19
README.md Normal file
View File

@@ -0,0 +1,19 @@
---
language:
- en
tags:
- pytorch
- causal-lm
- pythia
license: apache-2.0
datasets:
- Anthropic/hh-rlhf
---
[Pythia-6.9b](https://huggingface.co/EleutherAI/pythia-6.9b) supervised finetuned with [Anthropic-hh-rlhf dataset](https://huggingface.co/datasets/Anthropic/hh-rlhf) for 1 epoch [(sft-model)](https://huggingface.co/lomahony/eleuther-pythia6.9b-hh-sft),
before DPO [(paper)](https://arxiv.org/abs/2305.18290) with same dataset for 1 epoch.
[wandb log](https://wandb.ai/pythia_dpo/Pythia_LOM/runs/i53fbcih?workspace=user-lauraomahony)
Benchmark evaluations included in repo done using [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness/tree/big-refactor).
See [Pythia-6.9b](https://huggingface.co/EleutherAI/pythia-6.9b) for original model details [(paper)](https://arxiv.org/abs/2101.00027).

24
base-6.9b-eval-files/EleutherAI-pythia-6.9b-0shot-shelloutput.txt Normal file
View File

@@ -0,0 +1,24 @@
bootstrapping for stddev: perplexity
hf (pretrained=EleutherAI/pythia-6.9b), limit: None, num_fewshot: 0, batch_size: 4
| Task |Version|Filter| Metric | Value | |Stderr|
|--------------|-------|------|---------------|------:|---|-----:|
|arc_challenge |Yaml |none |acc | 0.3140|± |0.0136|
| | |none |acc_norm | 0.3515|± |0.0140|
|arc_easy |Yaml |none |acc | 0.6705|± |0.0096|
| | |none |acc_norm | 0.6128|± |0.0100|
|boolq |Yaml |none |acc | 0.6352|± |0.0084|
|hellaswag |Yaml |none |acc | 0.4798|± |0.0050|
| | |none |acc_norm | 0.6389|± |0.0048|
|lambada_openai|Yaml |none |perplexity | 4.4566|± |0.1000|
| | |none |acc | 0.6713|± |0.0065|
|openbookqa |Yaml |none |acc | 0.2560|± |0.0195|
| | |none |acc_norm | 0.3720|± |0.0216|
|piqa |Yaml |none |acc | 0.7524|± |0.0101|
| | |none |acc_norm | 0.7639|± |0.0099|
|sciq |Yaml |none |acc | 0.8930|± |0.0098|
| | |none |acc_norm | 0.8380|± |0.0117|
|wikitext |Yaml |none |word_perplexity|17.6830| | |
| | |none |byte_perplexity| 1.6057| | |
| | |none |bits_per_byte | 0.6832| | |
|winogrande |Yaml |none |acc | 0.6077|± |0.0137|

404
base-6.9b-eval-files/EleutherAI-pythia-6.9b-0shot/results.json Normal file
View File

@@ -0,0 +1,404 @@
{
"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"
}

24
base-6.9b-eval-files/EleutherAI-pythia-6.9b-5shot-shelloutput.txt Normal file
View File

@@ -0,0 +1,24 @@
bootstrapping for stddev: perplexity
hf (pretrained=EleutherAI/pythia-6.9b), limit: None, num_fewshot: 5, batch_size: 4
| Task |Version|Filter| Metric | Value | |Stderr|
|--------------|-------|------|---------------|------:|---|-----:|
|arc_challenge |Yaml |none |acc | 0.3541|± |0.0140|
| | |none |acc_norm | 0.3891|± |0.0142|
|arc_easy |Yaml |none |acc | 0.6944|± |0.0095|
| | |none |acc_norm | 0.7045|± |0.0094|
|boolq |Yaml |none |acc | 0.6575|± |0.0083|
|hellaswag |Yaml |none |acc | 0.4804|± |0.0050|
| | |none |acc_norm | 0.6509|± |0.0048|
|lambada_openai|Yaml |none |perplexity | 5.6328|± |0.1331|
| | |none |acc | 0.6231|± |0.0068|
|openbookqa |Yaml |none |acc | 0.2800|± |0.0201|
| | |none |acc_norm | 0.3540|± |0.0214|
|piqa |Yaml |none |acc | 0.7644|± |0.0099|
| | |none |acc_norm | 0.7633|± |0.0099|
|sciq |Yaml |none |acc | 0.9470|± |0.0071|
| | |none |acc_norm | 0.9550|± |0.0066|
|wikitext |Yaml |none |word_perplexity|17.6830| | |
| | |none |byte_perplexity| 1.6057| | |
| | |none |bits_per_byte | 0.6832| | |
|winogrande |Yaml |none |acc | 0.6069|± |0.0137|

404
base-6.9b-eval-files/EleutherAI-pythia-6.9b-5shot/results.json Normal file
View File

@@ -0,0 +1,404 @@
{
"results": {
"arc_challenge": {
"acc,none": 0.35409556313993173,
"acc_stderr,none": 0.01397545412275655,
"acc_norm,none": 0.3890784982935154,
"acc_norm_stderr,none": 0.014247309976045609
},
"arc_easy": {
"acc,none": 0.6944444444444444,
"acc_stderr,none": 0.00945218121359346,
"acc_norm,none": 0.7045454545454546,
"acc_norm_stderr,none": 0.009361987126556446
},
"boolq": {
"acc,none": 0.6574923547400612,
"acc_stderr,none": 0.008299903219506771
},
"hellaswag": {
"acc,none": 0.4803823939454292,
"acc_stderr,none": 0.004985939292819588,
"acc_norm,none": 0.650866361282613,
"acc_norm_stderr,none": 0.004757220449283696
},
"lambada_openai": {
"perplexity,none": 5.632830092794682,
"perplexity_stderr,none": 0.13309255473880457,
"acc,none": 0.6231321560256161,
"acc_stderr,none": 0.006751444407117095
},
"openbookqa": {
"acc,none": 0.28,
"acc_stderr,none": 0.020099950647503233,
"acc_norm,none": 0.354,
"acc_norm_stderr,none": 0.021407582047916447
},
"piqa": {
"acc,none": 0.764417845484222,
"acc_stderr,none": 0.009901067586473907,
"acc_norm,none": 0.7633297062023939,
"acc_norm_stderr,none": 0.009916841655042809
},
"sciq": {
"acc,none": 0.947,
"acc_stderr,none": 0.007088105617246445,
"acc_norm,none": 0.955,
"acc_norm_stderr,none": 0.006558812241406067
},
"wikitext": {
"word_perplexity,none": 17.682958079421635,
"byte_perplexity,none": 1.6057045697141277,
"bits_per_byte,none": 0.6832064787735104
},
"winogrande": {
"acc,none": 0.6069455406471981,
"acc_stderr,none": 0.013727276249108442
}
},
"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": 5,
"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": 5,
"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": 5,
"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": 5,
"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": 5,
"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": 5,
"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": 5,
"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": 5,
"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 0x7fd39a70d120>",
"description": "",
"target_delimiter": " ",
"fewshot_delimiter": "\n\n",
"num_fewshot": 5,
"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 0x7fd39a70d360>",
"doc_to_target": "<function doc_to_target at 0x7fd39a70d6c0>",
"doc_to_choice": "<function doc_to_choice at 0x7fd39a70da20>",
"description": "",
"target_delimiter": " ",
"fewshot_delimiter": "\n\n",
"num_fewshot": 5,
"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"
}

24
config.json Normal file
View File

@@ -0,0 +1,24 @@
{
"architectures": [
"GPTNeoXForCausalLM"
],
"bos_token_id": 0,
"eos_token_id": 0,
"hidden_act": "gelu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 16384,
"layer_norm_eps": 1e-05,
"max_position_embeddings": 2048,
"model_type": "gpt_neox",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"rotary_emb_base": 10000,
"rotary_pct": 0.25,
"tie_word_embeddings": false,
"torch_dtype": "float16",
"transformers_version": "4.24.0",
"use_cache": true,
"use_parallel_residual": true,
"vocab_size": 50432
}

24
dpo-6.9b-eval-files/dpo-pythia-6.9b-0shot-shelloutput.txt Normal file
View File

@@ -0,0 +1,24 @@
bootstrapping for stddev: perplexity
hf (pretrained=lomahony/eleuther-pythia6.9b-hh-dpo), limit: None, num_fewshot: 0, batch_size: 4
| Task |Version|Filter| Metric | Value | |Stderr|
|--------------|-------|------|---------------|------:|---|-----:|
|arc_challenge |Yaml |none |acc | 0.3242|± |0.0137|
| | |none |acc_norm | 0.3541|± |0.0140|
|arc_easy |Yaml |none |acc | 0.6860|± |0.0095|
| | |none |acc_norm | 0.6149|± |0.0100|
|boolq |Yaml |none |acc | 0.6688|± |0.0082|
|hellaswag |Yaml |none |acc | 0.4910|± |0.0050|
| | |none |acc_norm | 0.6487|± |0.0048|
|lambada_openai|Yaml |none |perplexity | 3.7085|± |0.0866|
| | |none |acc | 0.6947|± |0.0064|
|openbookqa |Yaml |none |acc | 0.2760|± |0.0200|
| | |none |acc_norm | 0.3860|± |0.0218|
|piqa |Yaml |none |acc | 0.7622|± |0.0099|
| | |none |acc_norm | 0.7704|± |0.0098|
|sciq |Yaml |none |acc | 0.8940|± |0.0097|
| | |none |acc_norm | 0.8300|± |0.0119|
|wikitext |Yaml |none |word_perplexity|18.1988| | |
| | |none |byte_perplexity| 1.6133| | |
| | |none |bits_per_byte | 0.6900| | |
|winogrande |Yaml |none |acc | 0.6235|± |0.0136|

404
dpo-6.9b-eval-files/dpo-pythia-6.9b-0shot/results.json Normal file
View File

@@ -0,0 +1,404 @@
{
"results": {
"arc_challenge": {
"acc,none": 0.3242320819112628,
"acc_stderr,none": 0.013678810399518826,
"acc_norm,none": 0.35409556313993173,
"acc_norm_stderr,none": 0.013975454122756555
},
"arc_easy": {
"acc,none": 0.686026936026936,
"acc_stderr,none": 0.009523245335215506,
"acc_norm,none": 0.61489898989899,
"acc_norm_stderr,none": 0.00998521479873725
},
"boolq": {
"acc,none": 0.6688073394495413,
"acc_stderr,none": 0.008231583858517827
},
"hellaswag": {
"acc,none": 0.4910376419040032,
"acc_stderr,none": 0.004988979750014446,
"acc_norm,none": 0.6486755626369249,
"acc_norm_stderr,none": 0.00476408459717691
},
"lambada_openai": {
"perplexity,none": 3.708528636318625,
"perplexity_stderr,none": 0.08658115943815492,
"acc,none": 0.6947409276149815,
"acc_stderr,none": 0.006415903230922688
},
"openbookqa": {
"acc,none": 0.276,
"acc_stderr,none": 0.02001121929807353,
"acc_norm,none": 0.386,
"acc_norm_stderr,none": 0.021793529219281165
},
"piqa": {
"acc,none": 0.7622415669205659,
"acc_stderr,none": 0.009932525779525485,
"acc_norm,none": 0.7704026115342764,
"acc_norm_stderr,none": 0.009812682950815187
},
"sciq": {
"acc,none": 0.894,
"acc_stderr,none": 0.00973955126578514,
"acc_norm,none": 0.83,
"acc_norm_stderr,none": 0.011884495834541663
},
"wikitext": {
"word_perplexity,none": 18.198752319892222,
"byte_perplexity,none": 1.6133334471336491,
"bits_per_byte,none": 0.6900446485426469
},
"winogrande": {
"acc,none": 0.6235201262825573,
"acc_stderr,none": 0.013616931960667182
}
},
"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 0x7fed0a7c5120>",
"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 0x7fed0a7c5360>",
"doc_to_target": "<function doc_to_target at 0x7fed0a7c56c0>",
"doc_to_choice": "<function doc_to_choice at 0x7fed0a7c5a20>",
"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=lomahony/eleuther-pythia6.9b-hh-dpo",
"batch_size": "4",
"batch_sizes": [],
"device": null,
"use_cache": null,
"limit": null,
"bootstrap_iters": 100000
},
"git_hash": "d1a44c8"
}

24
dpo-6.9b-eval-files/dpo-pythia-6.9b-5shot-shelloutput.txt Normal file
View File

@@ -0,0 +1,24 @@
bootstrapping for stddev: perplexity
hf (pretrained=lomahony/eleuther-pythia6.9b-hh-dpo), limit: None, num_fewshot: 5, batch_size: 4
| Task |Version|Filter| Metric | Value | |Stderr|
|--------------|-------|------|---------------|------:|---|-----:|
|arc_challenge |Yaml |none |acc | 0.3746|± |0.0141|
| | |none |acc_norm | 0.4036|± |0.0143|
|arc_easy |Yaml |none |acc | 0.7193|± |0.0092|
| | |none |acc_norm | 0.7197|± |0.0092|
|boolq |Yaml |none |acc | 0.6694|± |0.0082|
|hellaswag |Yaml |none |acc | 0.4890|± |0.0050|
| | |none |acc_norm | 0.6550|± |0.0047|
|lambada_openai|Yaml |none |perplexity | 4.7574|± |0.1175|
| | |none |acc | 0.6487|± |0.0067|
|openbookqa |Yaml |none |acc | 0.2960|± |0.0204|
| | |none |acc_norm | 0.3820|± |0.0218|
|piqa |Yaml |none |acc | 0.7726|± |0.0098|
| | |none |acc_norm | 0.7704|± |0.0098|
|sciq |Yaml |none |acc | 0.9490|± |0.0070|
| | |none |acc_norm | 0.9530|± |0.0067|
|wikitext |Yaml |none |word_perplexity|18.1988| | |
| | |none |byte_perplexity| 1.6133| | |
| | |none |bits_per_byte | 0.6900| | |
|winogrande |Yaml |none |acc | 0.6235|± |0.0136|

404
dpo-6.9b-eval-files/dpo-pythia-6.9b-5shot/results.json Normal file
View File

@@ -0,0 +1,404 @@
{
"results": {
"arc_challenge": {
"acc,none": 0.37457337883959047,
"acc_stderr,none": 0.01414419347189344,
"acc_norm,none": 0.4035836177474403,
"acc_norm_stderr,none": 0.014337158914268447
},
"arc_easy": {
"acc,none": 0.7192760942760943,
"acc_stderr,none": 0.009220526174711353,
"acc_norm,none": 0.7196969696969697,
"acc_norm_stderr,none": 0.009216306864088041
},
"boolq": {
"acc,none": 0.6694189602446483,
"acc_stderr,none": 0.008227739156121659
},
"hellaswag": {
"acc,none": 0.48904600677155946,
"acc_stderr,none": 0.0049885838203099185,
"acc_norm,none": 0.6550487950607449,
"acc_norm_stderr,none": 0.004743808792037855
},
"lambada_openai": {
"perplexity,none": 4.757413961755731,
"perplexity_stderr,none": 0.11748014516027318,
"acc,none": 0.6487483019600233,
"acc_stderr,none": 0.006650578225573465
},
"openbookqa": {
"acc,none": 0.296,
"acc_stderr,none": 0.020435342091896135,
"acc_norm,none": 0.382,
"acc_norm_stderr,none": 0.021750820591250837
},
"piqa": {
"acc,none": 0.7725788900979326,
"acc_stderr,none": 0.009779850767847252,
"acc_norm,none": 0.7704026115342764,
"acc_norm_stderr,none": 0.009812682950815181
},
"sciq": {
"acc,none": 0.949,
"acc_stderr,none": 0.006960420062571422,
"acc_norm,none": 0.953,
"acc_norm_stderr,none": 0.006695956678163039
},
"wikitext": {
"word_perplexity,none": 18.198752319892222,
"byte_perplexity,none": 1.6133334471336491,
"bits_per_byte,none": 0.6900446485426469
},
"winogrande": {
"acc,none": 0.6235201262825573,
"acc_stderr,none": 0.013616931960667182
}
},
"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": 5,
"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": 5,
"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": 5,
"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": 5,
"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": 5,
"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": 5,
"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": 5,
"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": 5,
"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 0x7fac47ff5120>",
"description": "",
"target_delimiter": " ",
"fewshot_delimiter": "\n\n",
"num_fewshot": 5,
"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 0x7fac47ff5360>",
"doc_to_target": "<function doc_to_target at 0x7fac47ff56c0>",
"doc_to_choice": "<function doc_to_choice at 0x7fac47ff5a20>",
"description": "",
"target_delimiter": " ",
"fewshot_delimiter": "\n\n",
"num_fewshot": 5,
"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=lomahony/eleuther-pythia6.9b-hh-dpo",
"batch_size": "4",
"batch_sizes": [],
"device": null,
"use_cache": null,
"limit": null,
"bootstrap_iters": 100000
},
"git_hash": "d1a44c8"
}

3
pytorch_model.bin Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:979cb981a43853cd0d8ca28823eef2b7fd762f7003a3d139fefe70e0bd645bed
size 13848993085

5
special_tokens_map.json Normal file
View File

@@ -0,0 +1,5 @@
{
"bos_token": "<|endoftext|>",
"eos_token": "<|endoftext|>",
"unk_token": "<|endoftext|>"
}

100528
tokenizer.json Normal file
View File

File diff suppressed because it is too large Load Diff

9
tokenizer_config.json Normal file
View File

@@ -0,0 +1,9 @@
{
"add_prefix_space": false,
"bos_token": "<|endoftext|>",
"eos_token": "<|endoftext|>",
"name_or_path": "EleutherAI/gpt-neox-20b",
"special_tokens_map_file": "/admin/home-hailey/.cache/huggingface/hub/models--EleutherAI--gpt-neox-20b/snapshots/4e49eadb5d14bd22f314ec3f45b69a87b88c7691/special_tokens_map.json",
"tokenizer_class": "GPTNeoXTokenizer",
"unk_token": "<|endoftext|>"
}