Upload folder using huggingface_hub
Browse files- arc_challenge_25shot_bs16_fp32.json +25 -0
- config.json +26 -0
- generation_config.json +6 -0
- gsm8k_5shot.json +26 -0
- gsm8k_5shot_bs8_bf32.json +26 -0
- gsm8k_5shot_bs8_fp32.json +26 -0
- hellaswag_10shot_bs16_fp32.json +25 -0
- mmlu_5shot.json +420 -0
- mmlu_5shot_bs4_fp32.json +417 -0
- model-00001-of-00006.safetensors +3 -0
- model-00002-of-00006.safetensors +3 -0
- model-00003-of-00006.safetensors +3 -0
- model-00004-of-00006.safetensors +3 -0
- model-00005-of-00006.safetensors +3 -0
- model-00006-of-00006.safetensors +3 -0
- model.safetensors.index.json +298 -0
- special_tokens_map.json +23 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +41 -0
- truthfulqa_mc_0shot.json +28 -0
- truthfulqa_mc_0shot_bs16_fp32.json +25 -0
- winogrande_5shot_bs16_fp32.json +23 -0
arc_challenge_25shot_bs16_fp32.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"arc_challenge": {
|
| 4 |
+
"acc": 0.4974402730375427,
|
| 5 |
+
"acc_stderr": 0.014611199329843791,
|
| 6 |
+
"acc_norm": 0.5281569965870307,
|
| 7 |
+
"acc_norm_stderr": 0.014588204105102203
|
| 8 |
+
}
|
| 9 |
+
},
|
| 10 |
+
"versions": {
|
| 11 |
+
"arc_challenge": 0
|
| 12 |
+
},
|
| 13 |
+
"config": {
|
| 14 |
+
"model": "sparseml",
|
| 15 |
+
"model_args": "pretrained=/nm/drive1/alexandre/zoo/llama2-7b-ultrachat200k_llama2_pretrain-base/training,dtype=float32",
|
| 16 |
+
"num_fewshot": 25,
|
| 17 |
+
"batch_size": "16",
|
| 18 |
+
"batch_sizes": [],
|
| 19 |
+
"device": "cuda:1",
|
| 20 |
+
"no_cache": true,
|
| 21 |
+
"limit": null,
|
| 22 |
+
"bootstrap_iters": 100000,
|
| 23 |
+
"description_dict": {}
|
| 24 |
+
}
|
| 25 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "meta-llama/Llama-2-7b-hf",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"bos_token_id": 1,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 4096,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 11008,
|
| 13 |
+
"max_position_embeddings": 4096,
|
| 14 |
+
"model_type": "llama",
|
| 15 |
+
"num_attention_heads": 32,
|
| 16 |
+
"num_hidden_layers": 32,
|
| 17 |
+
"num_key_value_heads": 32,
|
| 18 |
+
"pretraining_tp": 1,
|
| 19 |
+
"rms_norm_eps": 1e-05,
|
| 20 |
+
"rope_scaling": null,
|
| 21 |
+
"tie_word_embeddings": false,
|
| 22 |
+
"torch_dtype": "float32",
|
| 23 |
+
"transformers_version": "4.35.2",
|
| 24 |
+
"use_cache": true,
|
| 25 |
+
"vocab_size": 32000
|
| 26 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"transformers_version": "4.35.2"
|
| 6 |
+
}
|
gsm8k_5shot.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"gsm8k": {
|
| 4 |
+
"acc": 0.1243366186504928,
|
| 5 |
+
"acc_stderr": 0.00908888096202845
|
| 6 |
+
}
|
| 7 |
+
},
|
| 8 |
+
"versions": {
|
| 9 |
+
"gsm8k": 0
|
| 10 |
+
},
|
| 11 |
+
"config": {
|
| 12 |
+
"model": "hf-causal-experimental",
|
| 13 |
+
"model_args": {
|
| 14 |
+
"pretrained": "/network/alexandre/research/llama2_7b_ultrachat/dense/dense_finetuning/dense_LR1e-4_E1/training",
|
| 15 |
+
"trust_remote_code": true
|
| 16 |
+
},
|
| 17 |
+
"num_fewshot": 5,
|
| 18 |
+
"batch_size": "8",
|
| 19 |
+
"batch_sizes": [],
|
| 20 |
+
"device": "cuda:3",
|
| 21 |
+
"no_cache": true,
|
| 22 |
+
"limit": null,
|
| 23 |
+
"bootstrap_iters": 100000,
|
| 24 |
+
"description_dict": {}
|
| 25 |
+
}
|
| 26 |
+
}
|
gsm8k_5shot_bs8_bf32.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"gsm8k": {
|
| 4 |
+
"acc": 0.1243366186504928,
|
| 5 |
+
"acc_stderr": 0.00908888096202845
|
| 6 |
+
}
|
| 7 |
+
},
|
| 8 |
+
"versions": {
|
| 9 |
+
"gsm8k": 0
|
| 10 |
+
},
|
| 11 |
+
"config": {
|
| 12 |
+
"model": "hf-causal-experimental",
|
| 13 |
+
"model_args": {
|
| 14 |
+
"pretrained": "/nm/drive1/alexandre/zoo/llama2-7b-ultrachat200k_llama2_pretrain-base/training",
|
| 15 |
+
"dtype": "float32"
|
| 16 |
+
},
|
| 17 |
+
"num_fewshot": 5,
|
| 18 |
+
"batch_size": "8",
|
| 19 |
+
"batch_sizes": [],
|
| 20 |
+
"device": "cuda:7",
|
| 21 |
+
"no_cache": true,
|
| 22 |
+
"limit": null,
|
| 23 |
+
"bootstrap_iters": 100000,
|
| 24 |
+
"description_dict": {}
|
| 25 |
+
}
|
| 26 |
+
}
|
gsm8k_5shot_bs8_fp32.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"gsm8k": {
|
| 4 |
+
"acc": 0.1243366186504928,
|
| 5 |
+
"acc_stderr": 0.00908888096202845
|
| 6 |
+
}
|
| 7 |
+
},
|
| 8 |
+
"versions": {
|
| 9 |
+
"gsm8k": 0
|
| 10 |
+
},
|
| 11 |
+
"config": {
|
| 12 |
+
"model": "hf-causal-experimental",
|
| 13 |
+
"model_args": {
|
| 14 |
+
"pretrained": "/nm/drive1/alexandre/zoo/llama2-7b-ultrachat200k_llama2_pretrain-base/training",
|
| 15 |
+
"dtype": "float32"
|
| 16 |
+
},
|
| 17 |
+
"num_fewshot": 5,
|
| 18 |
+
"batch_size": "8",
|
| 19 |
+
"batch_sizes": [],
|
| 20 |
+
"device": "cuda:7",
|
| 21 |
+
"no_cache": true,
|
| 22 |
+
"limit": null,
|
| 23 |
+
"bootstrap_iters": 100000,
|
| 24 |
+
"description_dict": {}
|
| 25 |
+
}
|
| 26 |
+
}
|
hellaswag_10shot_bs16_fp32.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"hellaswag": {
|
| 4 |
+
"acc": 0.5690101573391755,
|
| 5 |
+
"acc_stderr": 0.004942026200279588,
|
| 6 |
+
"acc_norm": 0.7590121489743079,
|
| 7 |
+
"acc_norm_stderr": 0.004268088879039827
|
| 8 |
+
}
|
| 9 |
+
},
|
| 10 |
+
"versions": {
|
| 11 |
+
"hellaswag": 0
|
| 12 |
+
},
|
| 13 |
+
"config": {
|
| 14 |
+
"model": "sparseml",
|
| 15 |
+
"model_args": "pretrained=/nm/drive1/alexandre/zoo/llama2-7b-ultrachat200k_llama2_pretrain-base/training,dtype=float32",
|
| 16 |
+
"num_fewshot": 10,
|
| 17 |
+
"batch_size": "16",
|
| 18 |
+
"batch_sizes": [],
|
| 19 |
+
"device": "cuda:4",
|
| 20 |
+
"no_cache": true,
|
| 21 |
+
"limit": null,
|
| 22 |
+
"bootstrap_iters": 100000,
|
| 23 |
+
"description_dict": {}
|
| 24 |
+
}
|
| 25 |
+
}
|
mmlu_5shot.json
ADDED
|
@@ -0,0 +1,420 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"hendrycksTest-abstract_algebra": {
|
| 4 |
+
"acc": 0.31,
|
| 5 |
+
"acc_stderr": 0.04648231987117316,
|
| 6 |
+
"acc_norm": 0.31,
|
| 7 |
+
"acc_norm_stderr": 0.04648231987117316
|
| 8 |
+
},
|
| 9 |
+
"hendrycksTest-anatomy": {
|
| 10 |
+
"acc": 0.43703703703703706,
|
| 11 |
+
"acc_stderr": 0.04284958639753399,
|
| 12 |
+
"acc_norm": 0.43703703703703706,
|
| 13 |
+
"acc_norm_stderr": 0.04284958639753399
|
| 14 |
+
},
|
| 15 |
+
"hendrycksTest-astronomy": {
|
| 16 |
+
"acc": 0.40789473684210525,
|
| 17 |
+
"acc_stderr": 0.03999309712777471,
|
| 18 |
+
"acc_norm": 0.40789473684210525,
|
| 19 |
+
"acc_norm_stderr": 0.03999309712777471
|
| 20 |
+
},
|
| 21 |
+
"hendrycksTest-business_ethics": {
|
| 22 |
+
"acc": 0.46,
|
| 23 |
+
"acc_stderr": 0.05009082659620332,
|
| 24 |
+
"acc_norm": 0.46,
|
| 25 |
+
"acc_norm_stderr": 0.05009082659620332
|
| 26 |
+
},
|
| 27 |
+
"hendrycksTest-clinical_knowledge": {
|
| 28 |
+
"acc": 0.4830188679245283,
|
| 29 |
+
"acc_stderr": 0.030755120364119905,
|
| 30 |
+
"acc_norm": 0.4830188679245283,
|
| 31 |
+
"acc_norm_stderr": 0.030755120364119905
|
| 32 |
+
},
|
| 33 |
+
"hendrycksTest-college_biology": {
|
| 34 |
+
"acc": 0.4652777777777778,
|
| 35 |
+
"acc_stderr": 0.04171115858181618,
|
| 36 |
+
"acc_norm": 0.4652777777777778,
|
| 37 |
+
"acc_norm_stderr": 0.04171115858181618
|
| 38 |
+
},
|
| 39 |
+
"hendrycksTest-college_chemistry": {
|
| 40 |
+
"acc": 0.27,
|
| 41 |
+
"acc_stderr": 0.044619604333847394,
|
| 42 |
+
"acc_norm": 0.27,
|
| 43 |
+
"acc_norm_stderr": 0.044619604333847394
|
| 44 |
+
},
|
| 45 |
+
"hendrycksTest-college_computer_science": {
|
| 46 |
+
"acc": 0.39,
|
| 47 |
+
"acc_stderr": 0.04902071300001975,
|
| 48 |
+
"acc_norm": 0.39,
|
| 49 |
+
"acc_norm_stderr": 0.04902071300001975
|
| 50 |
+
},
|
| 51 |
+
"hendrycksTest-college_mathematics": {
|
| 52 |
+
"acc": 0.29,
|
| 53 |
+
"acc_stderr": 0.045604802157206845,
|
| 54 |
+
"acc_norm": 0.29,
|
| 55 |
+
"acc_norm_stderr": 0.045604802157206845
|
| 56 |
+
},
|
| 57 |
+
"hendrycksTest-college_medicine": {
|
| 58 |
+
"acc": 0.36416184971098264,
|
| 59 |
+
"acc_stderr": 0.03669072477416907,
|
| 60 |
+
"acc_norm": 0.36416184971098264,
|
| 61 |
+
"acc_norm_stderr": 0.03669072477416907
|
| 62 |
+
},
|
| 63 |
+
"hendrycksTest-college_physics": {
|
| 64 |
+
"acc": 0.2549019607843137,
|
| 65 |
+
"acc_stderr": 0.043364327079931785,
|
| 66 |
+
"acc_norm": 0.2549019607843137,
|
| 67 |
+
"acc_norm_stderr": 0.043364327079931785
|
| 68 |
+
},
|
| 69 |
+
"hendrycksTest-computer_security": {
|
| 70 |
+
"acc": 0.55,
|
| 71 |
+
"acc_stderr": 0.05,
|
| 72 |
+
"acc_norm": 0.55,
|
| 73 |
+
"acc_norm_stderr": 0.05
|
| 74 |
+
},
|
| 75 |
+
"hendrycksTest-conceptual_physics": {
|
| 76 |
+
"acc": 0.4085106382978723,
|
| 77 |
+
"acc_stderr": 0.03213418026701576,
|
| 78 |
+
"acc_norm": 0.4085106382978723,
|
| 79 |
+
"acc_norm_stderr": 0.03213418026701576
|
| 80 |
+
},
|
| 81 |
+
"hendrycksTest-econometrics": {
|
| 82 |
+
"acc": 0.2894736842105263,
|
| 83 |
+
"acc_stderr": 0.04266339443159394,
|
| 84 |
+
"acc_norm": 0.2894736842105263,
|
| 85 |
+
"acc_norm_stderr": 0.04266339443159394
|
| 86 |
+
},
|
| 87 |
+
"hendrycksTest-electrical_engineering": {
|
| 88 |
+
"acc": 0.4206896551724138,
|
| 89 |
+
"acc_stderr": 0.0411391498118926,
|
| 90 |
+
"acc_norm": 0.4206896551724138,
|
| 91 |
+
"acc_norm_stderr": 0.0411391498118926
|
| 92 |
+
},
|
| 93 |
+
"hendrycksTest-elementary_mathematics": {
|
| 94 |
+
"acc": 0.291005291005291,
|
| 95 |
+
"acc_stderr": 0.02339382650048487,
|
| 96 |
+
"acc_norm": 0.291005291005291,
|
| 97 |
+
"acc_norm_stderr": 0.02339382650048487
|
| 98 |
+
},
|
| 99 |
+
"hendrycksTest-formal_logic": {
|
| 100 |
+
"acc": 0.23015873015873015,
|
| 101 |
+
"acc_stderr": 0.037649508797906045,
|
| 102 |
+
"acc_norm": 0.23015873015873015,
|
| 103 |
+
"acc_norm_stderr": 0.037649508797906045
|
| 104 |
+
},
|
| 105 |
+
"hendrycksTest-global_facts": {
|
| 106 |
+
"acc": 0.34,
|
| 107 |
+
"acc_stderr": 0.04760952285695236,
|
| 108 |
+
"acc_norm": 0.34,
|
| 109 |
+
"acc_norm_stderr": 0.04760952285695236
|
| 110 |
+
},
|
| 111 |
+
"hendrycksTest-high_school_biology": {
|
| 112 |
+
"acc": 0.4870967741935484,
|
| 113 |
+
"acc_stderr": 0.028434533152681855,
|
| 114 |
+
"acc_norm": 0.4870967741935484,
|
| 115 |
+
"acc_norm_stderr": 0.028434533152681855
|
| 116 |
+
},
|
| 117 |
+
"hendrycksTest-high_school_chemistry": {
|
| 118 |
+
"acc": 0.33497536945812806,
|
| 119 |
+
"acc_stderr": 0.033208527423483104,
|
| 120 |
+
"acc_norm": 0.33497536945812806,
|
| 121 |
+
"acc_norm_stderr": 0.033208527423483104
|
| 122 |
+
},
|
| 123 |
+
"hendrycksTest-high_school_computer_science": {
|
| 124 |
+
"acc": 0.44,
|
| 125 |
+
"acc_stderr": 0.04988876515698589,
|
| 126 |
+
"acc_norm": 0.44,
|
| 127 |
+
"acc_norm_stderr": 0.04988876515698589
|
| 128 |
+
},
|
| 129 |
+
"hendrycksTest-high_school_european_history": {
|
| 130 |
+
"acc": 0.6363636363636364,
|
| 131 |
+
"acc_stderr": 0.03756335775187897,
|
| 132 |
+
"acc_norm": 0.6363636363636364,
|
| 133 |
+
"acc_norm_stderr": 0.03756335775187897
|
| 134 |
+
},
|
| 135 |
+
"hendrycksTest-high_school_geography": {
|
| 136 |
+
"acc": 0.5202020202020202,
|
| 137 |
+
"acc_stderr": 0.03559443565563918,
|
| 138 |
+
"acc_norm": 0.5202020202020202,
|
| 139 |
+
"acc_norm_stderr": 0.03559443565563918
|
| 140 |
+
},
|
| 141 |
+
"hendrycksTest-high_school_government_and_politics": {
|
| 142 |
+
"acc": 0.6424870466321243,
|
| 143 |
+
"acc_stderr": 0.034588160421810114,
|
| 144 |
+
"acc_norm": 0.6424870466321243,
|
| 145 |
+
"acc_norm_stderr": 0.034588160421810114
|
| 146 |
+
},
|
| 147 |
+
"hendrycksTest-high_school_macroeconomics": {
|
| 148 |
+
"acc": 0.4025641025641026,
|
| 149 |
+
"acc_stderr": 0.02486499515976775,
|
| 150 |
+
"acc_norm": 0.4025641025641026,
|
| 151 |
+
"acc_norm_stderr": 0.02486499515976775
|
| 152 |
+
},
|
| 153 |
+
"hendrycksTest-high_school_mathematics": {
|
| 154 |
+
"acc": 0.25925925925925924,
|
| 155 |
+
"acc_stderr": 0.026719240783712177,
|
| 156 |
+
"acc_norm": 0.25925925925925924,
|
| 157 |
+
"acc_norm_stderr": 0.026719240783712177
|
| 158 |
+
},
|
| 159 |
+
"hendrycksTest-high_school_microeconomics": {
|
| 160 |
+
"acc": 0.4411764705882353,
|
| 161 |
+
"acc_stderr": 0.0322529423239964,
|
| 162 |
+
"acc_norm": 0.4411764705882353,
|
| 163 |
+
"acc_norm_stderr": 0.0322529423239964
|
| 164 |
+
},
|
| 165 |
+
"hendrycksTest-high_school_physics": {
|
| 166 |
+
"acc": 0.23841059602649006,
|
| 167 |
+
"acc_stderr": 0.03479185572599661,
|
| 168 |
+
"acc_norm": 0.23841059602649006,
|
| 169 |
+
"acc_norm_stderr": 0.03479185572599661
|
| 170 |
+
},
|
| 171 |
+
"hendrycksTest-high_school_psychology": {
|
| 172 |
+
"acc": 0.6165137614678899,
|
| 173 |
+
"acc_stderr": 0.020847156641915984,
|
| 174 |
+
"acc_norm": 0.6165137614678899,
|
| 175 |
+
"acc_norm_stderr": 0.020847156641915984
|
| 176 |
+
},
|
| 177 |
+
"hendrycksTest-high_school_statistics": {
|
| 178 |
+
"acc": 0.27314814814814814,
|
| 179 |
+
"acc_stderr": 0.03038805130167812,
|
| 180 |
+
"acc_norm": 0.27314814814814814,
|
| 181 |
+
"acc_norm_stderr": 0.03038805130167812
|
| 182 |
+
},
|
| 183 |
+
"hendrycksTest-high_school_us_history": {
|
| 184 |
+
"acc": 0.6470588235294118,
|
| 185 |
+
"acc_stderr": 0.03354092437591518,
|
| 186 |
+
"acc_norm": 0.6470588235294118,
|
| 187 |
+
"acc_norm_stderr": 0.03354092437591518
|
| 188 |
+
},
|
| 189 |
+
"hendrycksTest-high_school_world_history": {
|
| 190 |
+
"acc": 0.6455696202531646,
|
| 191 |
+
"acc_stderr": 0.031137304297185805,
|
| 192 |
+
"acc_norm": 0.6455696202531646,
|
| 193 |
+
"acc_norm_stderr": 0.031137304297185805
|
| 194 |
+
},
|
| 195 |
+
"hendrycksTest-human_aging": {
|
| 196 |
+
"acc": 0.547085201793722,
|
| 197 |
+
"acc_stderr": 0.03340867501923324,
|
| 198 |
+
"acc_norm": 0.547085201793722,
|
| 199 |
+
"acc_norm_stderr": 0.03340867501923324
|
| 200 |
+
},
|
| 201 |
+
"hendrycksTest-human_sexuality": {
|
| 202 |
+
"acc": 0.549618320610687,
|
| 203 |
+
"acc_stderr": 0.04363643698524779,
|
| 204 |
+
"acc_norm": 0.549618320610687,
|
| 205 |
+
"acc_norm_stderr": 0.04363643698524779
|
| 206 |
+
},
|
| 207 |
+
"hendrycksTest-international_law": {
|
| 208 |
+
"acc": 0.628099173553719,
|
| 209 |
+
"acc_stderr": 0.04412015806624504,
|
| 210 |
+
"acc_norm": 0.628099173553719,
|
| 211 |
+
"acc_norm_stderr": 0.04412015806624504
|
| 212 |
+
},
|
| 213 |
+
"hendrycksTest-jurisprudence": {
|
| 214 |
+
"acc": 0.49074074074074076,
|
| 215 |
+
"acc_stderr": 0.04832853553437055,
|
| 216 |
+
"acc_norm": 0.49074074074074076,
|
| 217 |
+
"acc_norm_stderr": 0.04832853553437055
|
| 218 |
+
},
|
| 219 |
+
"hendrycksTest-logical_fallacies": {
|
| 220 |
+
"acc": 0.49079754601226994,
|
| 221 |
+
"acc_stderr": 0.03927705600787443,
|
| 222 |
+
"acc_norm": 0.49079754601226994,
|
| 223 |
+
"acc_norm_stderr": 0.03927705600787443
|
| 224 |
+
},
|
| 225 |
+
"hendrycksTest-machine_learning": {
|
| 226 |
+
"acc": 0.42857142857142855,
|
| 227 |
+
"acc_stderr": 0.04697113923010212,
|
| 228 |
+
"acc_norm": 0.42857142857142855,
|
| 229 |
+
"acc_norm_stderr": 0.04697113923010212
|
| 230 |
+
},
|
| 231 |
+
"hendrycksTest-management": {
|
| 232 |
+
"acc": 0.5339805825242718,
|
| 233 |
+
"acc_stderr": 0.0493929144727348,
|
| 234 |
+
"acc_norm": 0.5339805825242718,
|
| 235 |
+
"acc_norm_stderr": 0.0493929144727348
|
| 236 |
+
},
|
| 237 |
+
"hendrycksTest-marketing": {
|
| 238 |
+
"acc": 0.7136752136752137,
|
| 239 |
+
"acc_stderr": 0.02961432369045665,
|
| 240 |
+
"acc_norm": 0.7136752136752137,
|
| 241 |
+
"acc_norm_stderr": 0.02961432369045665
|
| 242 |
+
},
|
| 243 |
+
"hendrycksTest-medical_genetics": {
|
| 244 |
+
"acc": 0.55,
|
| 245 |
+
"acc_stderr": 0.04999999999999999,
|
| 246 |
+
"acc_norm": 0.55,
|
| 247 |
+
"acc_norm_stderr": 0.04999999999999999
|
| 248 |
+
},
|
| 249 |
+
"hendrycksTest-miscellaneous": {
|
| 250 |
+
"acc": 0.6219667943805874,
|
| 251 |
+
"acc_stderr": 0.017339844462104594,
|
| 252 |
+
"acc_norm": 0.6219667943805874,
|
| 253 |
+
"acc_norm_stderr": 0.017339844462104594
|
| 254 |
+
},
|
| 255 |
+
"hendrycksTest-moral_disputes": {
|
| 256 |
+
"acc": 0.5317919075144508,
|
| 257 |
+
"acc_stderr": 0.02686462436675664,
|
| 258 |
+
"acc_norm": 0.5317919075144508,
|
| 259 |
+
"acc_norm_stderr": 0.02686462436675664
|
| 260 |
+
},
|
| 261 |
+
"hendrycksTest-moral_scenarios": {
|
| 262 |
+
"acc": 0.2424581005586592,
|
| 263 |
+
"acc_stderr": 0.014333522059217892,
|
| 264 |
+
"acc_norm": 0.2424581005586592,
|
| 265 |
+
"acc_norm_stderr": 0.014333522059217892
|
| 266 |
+
},
|
| 267 |
+
"hendrycksTest-nutrition": {
|
| 268 |
+
"acc": 0.477124183006536,
|
| 269 |
+
"acc_stderr": 0.028599936776089782,
|
| 270 |
+
"acc_norm": 0.477124183006536,
|
| 271 |
+
"acc_norm_stderr": 0.028599936776089782
|
| 272 |
+
},
|
| 273 |
+
"hendrycksTest-philosophy": {
|
| 274 |
+
"acc": 0.5530546623794212,
|
| 275 |
+
"acc_stderr": 0.028237769422085328,
|
| 276 |
+
"acc_norm": 0.5530546623794212,
|
| 277 |
+
"acc_norm_stderr": 0.028237769422085328
|
| 278 |
+
},
|
| 279 |
+
"hendrycksTest-prehistory": {
|
| 280 |
+
"acc": 0.5308641975308642,
|
| 281 |
+
"acc_stderr": 0.027767689606833925,
|
| 282 |
+
"acc_norm": 0.5308641975308642,
|
| 283 |
+
"acc_norm_stderr": 0.027767689606833925
|
| 284 |
+
},
|
| 285 |
+
"hendrycksTest-professional_accounting": {
|
| 286 |
+
"acc": 0.375886524822695,
|
| 287 |
+
"acc_stderr": 0.028893955412115882,
|
| 288 |
+
"acc_norm": 0.375886524822695,
|
| 289 |
+
"acc_norm_stderr": 0.028893955412115882
|
| 290 |
+
},
|
| 291 |
+
"hendrycksTest-professional_law": {
|
| 292 |
+
"acc": 0.35723598435462844,
|
| 293 |
+
"acc_stderr": 0.012238615750316505,
|
| 294 |
+
"acc_norm": 0.35723598435462844,
|
| 295 |
+
"acc_norm_stderr": 0.012238615750316505
|
| 296 |
+
},
|
| 297 |
+
"hendrycksTest-professional_medicine": {
|
| 298 |
+
"acc": 0.4227941176470588,
|
| 299 |
+
"acc_stderr": 0.03000856284500349,
|
| 300 |
+
"acc_norm": 0.4227941176470588,
|
| 301 |
+
"acc_norm_stderr": 0.03000856284500349
|
| 302 |
+
},
|
| 303 |
+
"hendrycksTest-professional_psychology": {
|
| 304 |
+
"acc": 0.4591503267973856,
|
| 305 |
+
"acc_stderr": 0.020160213617222516,
|
| 306 |
+
"acc_norm": 0.4591503267973856,
|
| 307 |
+
"acc_norm_stderr": 0.020160213617222516
|
| 308 |
+
},
|
| 309 |
+
"hendrycksTest-public_relations": {
|
| 310 |
+
"acc": 0.5454545454545454,
|
| 311 |
+
"acc_stderr": 0.04769300568972744,
|
| 312 |
+
"acc_norm": 0.5454545454545454,
|
| 313 |
+
"acc_norm_stderr": 0.04769300568972744
|
| 314 |
+
},
|
| 315 |
+
"hendrycksTest-security_studies": {
|
| 316 |
+
"acc": 0.49795918367346936,
|
| 317 |
+
"acc_stderr": 0.0320089533497105,
|
| 318 |
+
"acc_norm": 0.49795918367346936,
|
| 319 |
+
"acc_norm_stderr": 0.0320089533497105
|
| 320 |
+
},
|
| 321 |
+
"hendrycksTest-sociology": {
|
| 322 |
+
"acc": 0.6318407960199005,
|
| 323 |
+
"acc_stderr": 0.03410410565495301,
|
| 324 |
+
"acc_norm": 0.6318407960199005,
|
| 325 |
+
"acc_norm_stderr": 0.03410410565495301
|
| 326 |
+
},
|
| 327 |
+
"hendrycksTest-us_foreign_policy": {
|
| 328 |
+
"acc": 0.67,
|
| 329 |
+
"acc_stderr": 0.04725815626252609,
|
| 330 |
+
"acc_norm": 0.67,
|
| 331 |
+
"acc_norm_stderr": 0.04725815626252609
|
| 332 |
+
},
|
| 333 |
+
"hendrycksTest-virology": {
|
| 334 |
+
"acc": 0.41566265060240964,
|
| 335 |
+
"acc_stderr": 0.03836722176598053,
|
| 336 |
+
"acc_norm": 0.41566265060240964,
|
| 337 |
+
"acc_norm_stderr": 0.03836722176598053
|
| 338 |
+
},
|
| 339 |
+
"hendrycksTest-world_religions": {
|
| 340 |
+
"acc": 0.6549707602339181,
|
| 341 |
+
"acc_stderr": 0.036459813773888065,
|
| 342 |
+
"acc_norm": 0.6549707602339181,
|
| 343 |
+
"acc_norm_stderr": 0.036459813773888065
|
| 344 |
+
}
|
| 345 |
+
},
|
| 346 |
+
"versions": {
|
| 347 |
+
"hendrycksTest-abstract_algebra": 1,
|
| 348 |
+
"hendrycksTest-anatomy": 1,
|
| 349 |
+
"hendrycksTest-astronomy": 1,
|
| 350 |
+
"hendrycksTest-business_ethics": 1,
|
| 351 |
+
"hendrycksTest-clinical_knowledge": 1,
|
| 352 |
+
"hendrycksTest-college_biology": 1,
|
| 353 |
+
"hendrycksTest-college_chemistry": 1,
|
| 354 |
+
"hendrycksTest-college_computer_science": 1,
|
| 355 |
+
"hendrycksTest-college_mathematics": 1,
|
| 356 |
+
"hendrycksTest-college_medicine": 1,
|
| 357 |
+
"hendrycksTest-college_physics": 1,
|
| 358 |
+
"hendrycksTest-computer_security": 1,
|
| 359 |
+
"hendrycksTest-conceptual_physics": 1,
|
| 360 |
+
"hendrycksTest-econometrics": 1,
|
| 361 |
+
"hendrycksTest-electrical_engineering": 1,
|
| 362 |
+
"hendrycksTest-elementary_mathematics": 1,
|
| 363 |
+
"hendrycksTest-formal_logic": 1,
|
| 364 |
+
"hendrycksTest-global_facts": 1,
|
| 365 |
+
"hendrycksTest-high_school_biology": 1,
|
| 366 |
+
"hendrycksTest-high_school_chemistry": 1,
|
| 367 |
+
"hendrycksTest-high_school_computer_science": 1,
|
| 368 |
+
"hendrycksTest-high_school_european_history": 1,
|
| 369 |
+
"hendrycksTest-high_school_geography": 1,
|
| 370 |
+
"hendrycksTest-high_school_government_and_politics": 1,
|
| 371 |
+
"hendrycksTest-high_school_macroeconomics": 1,
|
| 372 |
+
"hendrycksTest-high_school_mathematics": 1,
|
| 373 |
+
"hendrycksTest-high_school_microeconomics": 1,
|
| 374 |
+
"hendrycksTest-high_school_physics": 1,
|
| 375 |
+
"hendrycksTest-high_school_psychology": 1,
|
| 376 |
+
"hendrycksTest-high_school_statistics": 1,
|
| 377 |
+
"hendrycksTest-high_school_us_history": 1,
|
| 378 |
+
"hendrycksTest-high_school_world_history": 1,
|
| 379 |
+
"hendrycksTest-human_aging": 1,
|
| 380 |
+
"hendrycksTest-human_sexuality": 1,
|
| 381 |
+
"hendrycksTest-international_law": 1,
|
| 382 |
+
"hendrycksTest-jurisprudence": 1,
|
| 383 |
+
"hendrycksTest-logical_fallacies": 1,
|
| 384 |
+
"hendrycksTest-machine_learning": 1,
|
| 385 |
+
"hendrycksTest-management": 1,
|
| 386 |
+
"hendrycksTest-marketing": 1,
|
| 387 |
+
"hendrycksTest-medical_genetics": 1,
|
| 388 |
+
"hendrycksTest-miscellaneous": 1,
|
| 389 |
+
"hendrycksTest-moral_disputes": 1,
|
| 390 |
+
"hendrycksTest-moral_scenarios": 1,
|
| 391 |
+
"hendrycksTest-nutrition": 1,
|
| 392 |
+
"hendrycksTest-philosophy": 1,
|
| 393 |
+
"hendrycksTest-prehistory": 1,
|
| 394 |
+
"hendrycksTest-professional_accounting": 1,
|
| 395 |
+
"hendrycksTest-professional_law": 1,
|
| 396 |
+
"hendrycksTest-professional_medicine": 1,
|
| 397 |
+
"hendrycksTest-professional_psychology": 1,
|
| 398 |
+
"hendrycksTest-public_relations": 1,
|
| 399 |
+
"hendrycksTest-security_studies": 1,
|
| 400 |
+
"hendrycksTest-sociology": 1,
|
| 401 |
+
"hendrycksTest-us_foreign_policy": 1,
|
| 402 |
+
"hendrycksTest-virology": 1,
|
| 403 |
+
"hendrycksTest-world_religions": 1
|
| 404 |
+
},
|
| 405 |
+
"config": {
|
| 406 |
+
"model": "hf-causal-experimental",
|
| 407 |
+
"model_args": {
|
| 408 |
+
"pretrained": "/network/alexandre/research/llama2_7b_ultrachat/dense/dense_finetuning/dense_LR1e-4_E1/training",
|
| 409 |
+
"trust_remote_code": true
|
| 410 |
+
},
|
| 411 |
+
"num_fewshot": 5,
|
| 412 |
+
"batch_size": "4",
|
| 413 |
+
"batch_sizes": [],
|
| 414 |
+
"device": "cuda:5",
|
| 415 |
+
"no_cache": true,
|
| 416 |
+
"limit": null,
|
| 417 |
+
"bootstrap_iters": 100000,
|
| 418 |
+
"description_dict": {}
|
| 419 |
+
}
|
| 420 |
+
}
|
mmlu_5shot_bs4_fp32.json
ADDED
|
@@ -0,0 +1,417 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"hendrycksTest-abstract_algebra": {
|
| 4 |
+
"acc": 0.32,
|
| 5 |
+
"acc_stderr": 0.046882617226215034,
|
| 6 |
+
"acc_norm": 0.32,
|
| 7 |
+
"acc_norm_stderr": 0.046882617226215034
|
| 8 |
+
},
|
| 9 |
+
"hendrycksTest-anatomy": {
|
| 10 |
+
"acc": 0.4666666666666667,
|
| 11 |
+
"acc_stderr": 0.043097329010363554,
|
| 12 |
+
"acc_norm": 0.4666666666666667,
|
| 13 |
+
"acc_norm_stderr": 0.043097329010363554
|
| 14 |
+
},
|
| 15 |
+
"hendrycksTest-astronomy": {
|
| 16 |
+
"acc": 0.39473684210526316,
|
| 17 |
+
"acc_stderr": 0.039777499346220734,
|
| 18 |
+
"acc_norm": 0.39473684210526316,
|
| 19 |
+
"acc_norm_stderr": 0.039777499346220734
|
| 20 |
+
},
|
| 21 |
+
"hendrycksTest-business_ethics": {
|
| 22 |
+
"acc": 0.49,
|
| 23 |
+
"acc_stderr": 0.05024183937956912,
|
| 24 |
+
"acc_norm": 0.49,
|
| 25 |
+
"acc_norm_stderr": 0.05024183937956912
|
| 26 |
+
},
|
| 27 |
+
"hendrycksTest-clinical_knowledge": {
|
| 28 |
+
"acc": 0.47547169811320755,
|
| 29 |
+
"acc_stderr": 0.030735822206205608,
|
| 30 |
+
"acc_norm": 0.47547169811320755,
|
| 31 |
+
"acc_norm_stderr": 0.030735822206205608
|
| 32 |
+
},
|
| 33 |
+
"hendrycksTest-college_biology": {
|
| 34 |
+
"acc": 0.4583333333333333,
|
| 35 |
+
"acc_stderr": 0.04166666666666666,
|
| 36 |
+
"acc_norm": 0.4583333333333333,
|
| 37 |
+
"acc_norm_stderr": 0.04166666666666666
|
| 38 |
+
},
|
| 39 |
+
"hendrycksTest-college_chemistry": {
|
| 40 |
+
"acc": 0.29,
|
| 41 |
+
"acc_stderr": 0.04560480215720684,
|
| 42 |
+
"acc_norm": 0.29,
|
| 43 |
+
"acc_norm_stderr": 0.04560480215720684
|
| 44 |
+
},
|
| 45 |
+
"hendrycksTest-college_computer_science": {
|
| 46 |
+
"acc": 0.38,
|
| 47 |
+
"acc_stderr": 0.048783173121456316,
|
| 48 |
+
"acc_norm": 0.38,
|
| 49 |
+
"acc_norm_stderr": 0.048783173121456316
|
| 50 |
+
},
|
| 51 |
+
"hendrycksTest-college_mathematics": {
|
| 52 |
+
"acc": 0.25,
|
| 53 |
+
"acc_stderr": 0.04351941398892446,
|
| 54 |
+
"acc_norm": 0.25,
|
| 55 |
+
"acc_norm_stderr": 0.04351941398892446
|
| 56 |
+
},
|
| 57 |
+
"hendrycksTest-college_medicine": {
|
| 58 |
+
"acc": 0.4046242774566474,
|
| 59 |
+
"acc_stderr": 0.03742461193887248,
|
| 60 |
+
"acc_norm": 0.4046242774566474,
|
| 61 |
+
"acc_norm_stderr": 0.03742461193887248
|
| 62 |
+
},
|
| 63 |
+
"hendrycksTest-college_physics": {
|
| 64 |
+
"acc": 0.24509803921568626,
|
| 65 |
+
"acc_stderr": 0.042801058373643966,
|
| 66 |
+
"acc_norm": 0.24509803921568626,
|
| 67 |
+
"acc_norm_stderr": 0.042801058373643966
|
| 68 |
+
},
|
| 69 |
+
"hendrycksTest-computer_security": {
|
| 70 |
+
"acc": 0.54,
|
| 71 |
+
"acc_stderr": 0.05009082659620332,
|
| 72 |
+
"acc_norm": 0.54,
|
| 73 |
+
"acc_norm_stderr": 0.05009082659620332
|
| 74 |
+
},
|
| 75 |
+
"hendrycksTest-conceptual_physics": {
|
| 76 |
+
"acc": 0.40425531914893614,
|
| 77 |
+
"acc_stderr": 0.03208115750788684,
|
| 78 |
+
"acc_norm": 0.40425531914893614,
|
| 79 |
+
"acc_norm_stderr": 0.03208115750788684
|
| 80 |
+
},
|
| 81 |
+
"hendrycksTest-econometrics": {
|
| 82 |
+
"acc": 0.3157894736842105,
|
| 83 |
+
"acc_stderr": 0.043727482902780064,
|
| 84 |
+
"acc_norm": 0.3157894736842105,
|
| 85 |
+
"acc_norm_stderr": 0.043727482902780064
|
| 86 |
+
},
|
| 87 |
+
"hendrycksTest-electrical_engineering": {
|
| 88 |
+
"acc": 0.4413793103448276,
|
| 89 |
+
"acc_stderr": 0.04137931034482757,
|
| 90 |
+
"acc_norm": 0.4413793103448276,
|
| 91 |
+
"acc_norm_stderr": 0.04137931034482757
|
| 92 |
+
},
|
| 93 |
+
"hendrycksTest-elementary_mathematics": {
|
| 94 |
+
"acc": 0.2830687830687831,
|
| 95 |
+
"acc_stderr": 0.023201392938194974,
|
| 96 |
+
"acc_norm": 0.2830687830687831,
|
| 97 |
+
"acc_norm_stderr": 0.023201392938194974
|
| 98 |
+
},
|
| 99 |
+
"hendrycksTest-formal_logic": {
|
| 100 |
+
"acc": 0.23015873015873015,
|
| 101 |
+
"acc_stderr": 0.037649508797906045,
|
| 102 |
+
"acc_norm": 0.23015873015873015,
|
| 103 |
+
"acc_norm_stderr": 0.037649508797906045
|
| 104 |
+
},
|
| 105 |
+
"hendrycksTest-global_facts": {
|
| 106 |
+
"acc": 0.34,
|
| 107 |
+
"acc_stderr": 0.04760952285695235,
|
| 108 |
+
"acc_norm": 0.34,
|
| 109 |
+
"acc_norm_stderr": 0.04760952285695235
|
| 110 |
+
},
|
| 111 |
+
"hendrycksTest-high_school_biology": {
|
| 112 |
+
"acc": 0.47419354838709676,
|
| 113 |
+
"acc_stderr": 0.028406095057653326,
|
| 114 |
+
"acc_norm": 0.47419354838709676,
|
| 115 |
+
"acc_norm_stderr": 0.028406095057653326
|
| 116 |
+
},
|
| 117 |
+
"hendrycksTest-high_school_chemistry": {
|
| 118 |
+
"acc": 0.3251231527093596,
|
| 119 |
+
"acc_stderr": 0.032957975663112704,
|
| 120 |
+
"acc_norm": 0.3251231527093596,
|
| 121 |
+
"acc_norm_stderr": 0.032957975663112704
|
| 122 |
+
},
|
| 123 |
+
"hendrycksTest-high_school_computer_science": {
|
| 124 |
+
"acc": 0.44,
|
| 125 |
+
"acc_stderr": 0.04988876515698589,
|
| 126 |
+
"acc_norm": 0.44,
|
| 127 |
+
"acc_norm_stderr": 0.04988876515698589
|
| 128 |
+
},
|
| 129 |
+
"hendrycksTest-high_school_european_history": {
|
| 130 |
+
"acc": 0.6363636363636364,
|
| 131 |
+
"acc_stderr": 0.03756335775187897,
|
| 132 |
+
"acc_norm": 0.6363636363636364,
|
| 133 |
+
"acc_norm_stderr": 0.03756335775187897
|
| 134 |
+
},
|
| 135 |
+
"hendrycksTest-high_school_geography": {
|
| 136 |
+
"acc": 0.5404040404040404,
|
| 137 |
+
"acc_stderr": 0.035507024651313425,
|
| 138 |
+
"acc_norm": 0.5404040404040404,
|
| 139 |
+
"acc_norm_stderr": 0.035507024651313425
|
| 140 |
+
},
|
| 141 |
+
"hendrycksTest-high_school_government_and_politics": {
|
| 142 |
+
"acc": 0.6269430051813472,
|
| 143 |
+
"acc_stderr": 0.03490205592048573,
|
| 144 |
+
"acc_norm": 0.6269430051813472,
|
| 145 |
+
"acc_norm_stderr": 0.03490205592048573
|
| 146 |
+
},
|
| 147 |
+
"hendrycksTest-high_school_macroeconomics": {
|
| 148 |
+
"acc": 0.41025641025641024,
|
| 149 |
+
"acc_stderr": 0.024939313906940784,
|
| 150 |
+
"acc_norm": 0.41025641025641024,
|
| 151 |
+
"acc_norm_stderr": 0.024939313906940784
|
| 152 |
+
},
|
| 153 |
+
"hendrycksTest-high_school_mathematics": {
|
| 154 |
+
"acc": 0.2740740740740741,
|
| 155 |
+
"acc_stderr": 0.027195934804085622,
|
| 156 |
+
"acc_norm": 0.2740740740740741,
|
| 157 |
+
"acc_norm_stderr": 0.027195934804085622
|
| 158 |
+
},
|
| 159 |
+
"hendrycksTest-high_school_microeconomics": {
|
| 160 |
+
"acc": 0.42436974789915966,
|
| 161 |
+
"acc_stderr": 0.03210479051015776,
|
| 162 |
+
"acc_norm": 0.42436974789915966,
|
| 163 |
+
"acc_norm_stderr": 0.03210479051015776
|
| 164 |
+
},
|
| 165 |
+
"hendrycksTest-high_school_physics": {
|
| 166 |
+
"acc": 0.271523178807947,
|
| 167 |
+
"acc_stderr": 0.03631329803969653,
|
| 168 |
+
"acc_norm": 0.271523178807947,
|
| 169 |
+
"acc_norm_stderr": 0.03631329803969653
|
| 170 |
+
},
|
| 171 |
+
"hendrycksTest-high_school_psychology": {
|
| 172 |
+
"acc": 0.618348623853211,
|
| 173 |
+
"acc_stderr": 0.020828148517022582,
|
| 174 |
+
"acc_norm": 0.618348623853211,
|
| 175 |
+
"acc_norm_stderr": 0.020828148517022582
|
| 176 |
+
},
|
| 177 |
+
"hendrycksTest-high_school_statistics": {
|
| 178 |
+
"acc": 0.27314814814814814,
|
| 179 |
+
"acc_stderr": 0.03038805130167812,
|
| 180 |
+
"acc_norm": 0.27314814814814814,
|
| 181 |
+
"acc_norm_stderr": 0.03038805130167812
|
| 182 |
+
},
|
| 183 |
+
"hendrycksTest-high_school_us_history": {
|
| 184 |
+
"acc": 0.6470588235294118,
|
| 185 |
+
"acc_stderr": 0.03354092437591518,
|
| 186 |
+
"acc_norm": 0.6470588235294118,
|
| 187 |
+
"acc_norm_stderr": 0.03354092437591518
|
| 188 |
+
},
|
| 189 |
+
"hendrycksTest-high_school_world_history": {
|
| 190 |
+
"acc": 0.6371308016877637,
|
| 191 |
+
"acc_stderr": 0.03129920825530213,
|
| 192 |
+
"acc_norm": 0.6371308016877637,
|
| 193 |
+
"acc_norm_stderr": 0.03129920825530213
|
| 194 |
+
},
|
| 195 |
+
"hendrycksTest-human_aging": {
|
| 196 |
+
"acc": 0.5560538116591929,
|
| 197 |
+
"acc_stderr": 0.03334625674242728,
|
| 198 |
+
"acc_norm": 0.5560538116591929,
|
| 199 |
+
"acc_norm_stderr": 0.03334625674242728
|
| 200 |
+
},
|
| 201 |
+
"hendrycksTest-human_sexuality": {
|
| 202 |
+
"acc": 0.549618320610687,
|
| 203 |
+
"acc_stderr": 0.04363643698524779,
|
| 204 |
+
"acc_norm": 0.549618320610687,
|
| 205 |
+
"acc_norm_stderr": 0.04363643698524779
|
| 206 |
+
},
|
| 207 |
+
"hendrycksTest-international_law": {
|
| 208 |
+
"acc": 0.628099173553719,
|
| 209 |
+
"acc_stderr": 0.04412015806624504,
|
| 210 |
+
"acc_norm": 0.628099173553719,
|
| 211 |
+
"acc_norm_stderr": 0.04412015806624504
|
| 212 |
+
},
|
| 213 |
+
"hendrycksTest-jurisprudence": {
|
| 214 |
+
"acc": 0.48148148148148145,
|
| 215 |
+
"acc_stderr": 0.04830366024635331,
|
| 216 |
+
"acc_norm": 0.48148148148148145,
|
| 217 |
+
"acc_norm_stderr": 0.04830366024635331
|
| 218 |
+
},
|
| 219 |
+
"hendrycksTest-logical_fallacies": {
|
| 220 |
+
"acc": 0.5214723926380368,
|
| 221 |
+
"acc_stderr": 0.03924746876751129,
|
| 222 |
+
"acc_norm": 0.5214723926380368,
|
| 223 |
+
"acc_norm_stderr": 0.03924746876751129
|
| 224 |
+
},
|
| 225 |
+
"hendrycksTest-machine_learning": {
|
| 226 |
+
"acc": 0.42857142857142855,
|
| 227 |
+
"acc_stderr": 0.04697113923010212,
|
| 228 |
+
"acc_norm": 0.42857142857142855,
|
| 229 |
+
"acc_norm_stderr": 0.04697113923010212
|
| 230 |
+
},
|
| 231 |
+
"hendrycksTest-management": {
|
| 232 |
+
"acc": 0.5242718446601942,
|
| 233 |
+
"acc_stderr": 0.049449010929737795,
|
| 234 |
+
"acc_norm": 0.5242718446601942,
|
| 235 |
+
"acc_norm_stderr": 0.049449010929737795
|
| 236 |
+
},
|
| 237 |
+
"hendrycksTest-marketing": {
|
| 238 |
+
"acc": 0.7094017094017094,
|
| 239 |
+
"acc_stderr": 0.02974504857267408,
|
| 240 |
+
"acc_norm": 0.7094017094017094,
|
| 241 |
+
"acc_norm_stderr": 0.02974504857267408
|
| 242 |
+
},
|
| 243 |
+
"hendrycksTest-medical_genetics": {
|
| 244 |
+
"acc": 0.56,
|
| 245 |
+
"acc_stderr": 0.04988876515698589,
|
| 246 |
+
"acc_norm": 0.56,
|
| 247 |
+
"acc_norm_stderr": 0.04988876515698589
|
| 248 |
+
},
|
| 249 |
+
"hendrycksTest-miscellaneous": {
|
| 250 |
+
"acc": 0.6194125159642401,
|
| 251 |
+
"acc_stderr": 0.017362564126075414,
|
| 252 |
+
"acc_norm": 0.6194125159642401,
|
| 253 |
+
"acc_norm_stderr": 0.017362564126075414
|
| 254 |
+
},
|
| 255 |
+
"hendrycksTest-moral_disputes": {
|
| 256 |
+
"acc": 0.5260115606936416,
|
| 257 |
+
"acc_stderr": 0.026882643434022885,
|
| 258 |
+
"acc_norm": 0.5260115606936416,
|
| 259 |
+
"acc_norm_stderr": 0.026882643434022885
|
| 260 |
+
},
|
| 261 |
+
"hendrycksTest-moral_scenarios": {
|
| 262 |
+
"acc": 0.23798882681564246,
|
| 263 |
+
"acc_stderr": 0.014242630070574911,
|
| 264 |
+
"acc_norm": 0.23798882681564246,
|
| 265 |
+
"acc_norm_stderr": 0.014242630070574911
|
| 266 |
+
},
|
| 267 |
+
"hendrycksTest-nutrition": {
|
| 268 |
+
"acc": 0.4803921568627451,
|
| 269 |
+
"acc_stderr": 0.028607893699576063,
|
| 270 |
+
"acc_norm": 0.4803921568627451,
|
| 271 |
+
"acc_norm_stderr": 0.028607893699576063
|
| 272 |
+
},
|
| 273 |
+
"hendrycksTest-philosophy": {
|
| 274 |
+
"acc": 0.5594855305466238,
|
| 275 |
+
"acc_stderr": 0.028196400574197426,
|
| 276 |
+
"acc_norm": 0.5594855305466238,
|
| 277 |
+
"acc_norm_stderr": 0.028196400574197426
|
| 278 |
+
},
|
| 279 |
+
"hendrycksTest-prehistory": {
|
| 280 |
+
"acc": 0.5462962962962963,
|
| 281 |
+
"acc_stderr": 0.0277012284685426,
|
| 282 |
+
"acc_norm": 0.5462962962962963,
|
| 283 |
+
"acc_norm_stderr": 0.0277012284685426
|
| 284 |
+
},
|
| 285 |
+
"hendrycksTest-professional_accounting": {
|
| 286 |
+
"acc": 0.36524822695035464,
|
| 287 |
+
"acc_stderr": 0.02872386385328128,
|
| 288 |
+
"acc_norm": 0.36524822695035464,
|
| 289 |
+
"acc_norm_stderr": 0.02872386385328128
|
| 290 |
+
},
|
| 291 |
+
"hendrycksTest-professional_law": {
|
| 292 |
+
"acc": 0.36571056062581486,
|
| 293 |
+
"acc_stderr": 0.012301028188840567,
|
| 294 |
+
"acc_norm": 0.36571056062581486,
|
| 295 |
+
"acc_norm_stderr": 0.012301028188840567
|
| 296 |
+
},
|
| 297 |
+
"hendrycksTest-professional_medicine": {
|
| 298 |
+
"acc": 0.4411764705882353,
|
| 299 |
+
"acc_stderr": 0.030161911930767105,
|
| 300 |
+
"acc_norm": 0.4411764705882353,
|
| 301 |
+
"acc_norm_stderr": 0.030161911930767105
|
| 302 |
+
},
|
| 303 |
+
"hendrycksTest-professional_psychology": {
|
| 304 |
+
"acc": 0.44281045751633985,
|
| 305 |
+
"acc_stderr": 0.020095083154577347,
|
| 306 |
+
"acc_norm": 0.44281045751633985,
|
| 307 |
+
"acc_norm_stderr": 0.020095083154577347
|
| 308 |
+
},
|
| 309 |
+
"hendrycksTest-public_relations": {
|
| 310 |
+
"acc": 0.5727272727272728,
|
| 311 |
+
"acc_stderr": 0.04738198703545483,
|
| 312 |
+
"acc_norm": 0.5727272727272728,
|
| 313 |
+
"acc_norm_stderr": 0.04738198703545483
|
| 314 |
+
},
|
| 315 |
+
"hendrycksTest-security_studies": {
|
| 316 |
+
"acc": 0.49795918367346936,
|
| 317 |
+
"acc_stderr": 0.0320089533497105,
|
| 318 |
+
"acc_norm": 0.49795918367346936,
|
| 319 |
+
"acc_norm_stderr": 0.0320089533497105
|
| 320 |
+
},
|
| 321 |
+
"hendrycksTest-sociology": {
|
| 322 |
+
"acc": 0.6318407960199005,
|
| 323 |
+
"acc_stderr": 0.03410410565495301,
|
| 324 |
+
"acc_norm": 0.6318407960199005,
|
| 325 |
+
"acc_norm_stderr": 0.03410410565495301
|
| 326 |
+
},
|
| 327 |
+
"hendrycksTest-us_foreign_policy": {
|
| 328 |
+
"acc": 0.66,
|
| 329 |
+
"acc_stderr": 0.04760952285695237,
|
| 330 |
+
"acc_norm": 0.66,
|
| 331 |
+
"acc_norm_stderr": 0.04760952285695237
|
| 332 |
+
},
|
| 333 |
+
"hendrycksTest-virology": {
|
| 334 |
+
"acc": 0.4036144578313253,
|
| 335 |
+
"acc_stderr": 0.038194861407583984,
|
| 336 |
+
"acc_norm": 0.4036144578313253,
|
| 337 |
+
"acc_norm_stderr": 0.038194861407583984
|
| 338 |
+
},
|
| 339 |
+
"hendrycksTest-world_religions": {
|
| 340 |
+
"acc": 0.6432748538011696,
|
| 341 |
+
"acc_stderr": 0.03674013002860954,
|
| 342 |
+
"acc_norm": 0.6432748538011696,
|
| 343 |
+
"acc_norm_stderr": 0.03674013002860954
|
| 344 |
+
}
|
| 345 |
+
},
|
| 346 |
+
"versions": {
|
| 347 |
+
"hendrycksTest-abstract_algebra": 1,
|
| 348 |
+
"hendrycksTest-anatomy": 1,
|
| 349 |
+
"hendrycksTest-astronomy": 1,
|
| 350 |
+
"hendrycksTest-business_ethics": 1,
|
| 351 |
+
"hendrycksTest-clinical_knowledge": 1,
|
| 352 |
+
"hendrycksTest-college_biology": 1,
|
| 353 |
+
"hendrycksTest-college_chemistry": 1,
|
| 354 |
+
"hendrycksTest-college_computer_science": 1,
|
| 355 |
+
"hendrycksTest-college_mathematics": 1,
|
| 356 |
+
"hendrycksTest-college_medicine": 1,
|
| 357 |
+
"hendrycksTest-college_physics": 1,
|
| 358 |
+
"hendrycksTest-computer_security": 1,
|
| 359 |
+
"hendrycksTest-conceptual_physics": 1,
|
| 360 |
+
"hendrycksTest-econometrics": 1,
|
| 361 |
+
"hendrycksTest-electrical_engineering": 1,
|
| 362 |
+
"hendrycksTest-elementary_mathematics": 1,
|
| 363 |
+
"hendrycksTest-formal_logic": 1,
|
| 364 |
+
"hendrycksTest-global_facts": 1,
|
| 365 |
+
"hendrycksTest-high_school_biology": 1,
|
| 366 |
+
"hendrycksTest-high_school_chemistry": 1,
|
| 367 |
+
"hendrycksTest-high_school_computer_science": 1,
|
| 368 |
+
"hendrycksTest-high_school_european_history": 1,
|
| 369 |
+
"hendrycksTest-high_school_geography": 1,
|
| 370 |
+
"hendrycksTest-high_school_government_and_politics": 1,
|
| 371 |
+
"hendrycksTest-high_school_macroeconomics": 1,
|
| 372 |
+
"hendrycksTest-high_school_mathematics": 1,
|
| 373 |
+
"hendrycksTest-high_school_microeconomics": 1,
|
| 374 |
+
"hendrycksTest-high_school_physics": 1,
|
| 375 |
+
"hendrycksTest-high_school_psychology": 1,
|
| 376 |
+
"hendrycksTest-high_school_statistics": 1,
|
| 377 |
+
"hendrycksTest-high_school_us_history": 1,
|
| 378 |
+
"hendrycksTest-high_school_world_history": 1,
|
| 379 |
+
"hendrycksTest-human_aging": 1,
|
| 380 |
+
"hendrycksTest-human_sexuality": 1,
|
| 381 |
+
"hendrycksTest-international_law": 1,
|
| 382 |
+
"hendrycksTest-jurisprudence": 1,
|
| 383 |
+
"hendrycksTest-logical_fallacies": 1,
|
| 384 |
+
"hendrycksTest-machine_learning": 1,
|
| 385 |
+
"hendrycksTest-management": 1,
|
| 386 |
+
"hendrycksTest-marketing": 1,
|
| 387 |
+
"hendrycksTest-medical_genetics": 1,
|
| 388 |
+
"hendrycksTest-miscellaneous": 1,
|
| 389 |
+
"hendrycksTest-moral_disputes": 1,
|
| 390 |
+
"hendrycksTest-moral_scenarios": 1,
|
| 391 |
+
"hendrycksTest-nutrition": 1,
|
| 392 |
+
"hendrycksTest-philosophy": 1,
|
| 393 |
+
"hendrycksTest-prehistory": 1,
|
| 394 |
+
"hendrycksTest-professional_accounting": 1,
|
| 395 |
+
"hendrycksTest-professional_law": 1,
|
| 396 |
+
"hendrycksTest-professional_medicine": 1,
|
| 397 |
+
"hendrycksTest-professional_psychology": 1,
|
| 398 |
+
"hendrycksTest-public_relations": 1,
|
| 399 |
+
"hendrycksTest-security_studies": 1,
|
| 400 |
+
"hendrycksTest-sociology": 1,
|
| 401 |
+
"hendrycksTest-us_foreign_policy": 1,
|
| 402 |
+
"hendrycksTest-virology": 1,
|
| 403 |
+
"hendrycksTest-world_religions": 1
|
| 404 |
+
},
|
| 405 |
+
"config": {
|
| 406 |
+
"model": "sparseml",
|
| 407 |
+
"model_args": "pretrained=/nm/drive1/alexandre/zoo/llama2-7b-ultrachat200k_llama2_pretrain-base/training,dtype=float32",
|
| 408 |
+
"num_fewshot": 5,
|
| 409 |
+
"batch_size": "4",
|
| 410 |
+
"batch_sizes": [],
|
| 411 |
+
"device": "cuda:7",
|
| 412 |
+
"no_cache": true,
|
| 413 |
+
"limit": null,
|
| 414 |
+
"bootstrap_iters": 100000,
|
| 415 |
+
"description_dict": {}
|
| 416 |
+
}
|
| 417 |
+
}
|
model-00001-of-00006.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:89b5ac24b1fa215ba4b512b9923ef472e21846071a8f7277fe1a369118df9b0d
|
| 3 |
+
size 4840396416
|
model-00002-of-00006.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7be4d87ccfa5004aaeed6c4cf0a0783861b31c67a505513612fe800e656ddb7
|
| 3 |
+
size 4857206856
|
model-00003-of-00006.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b33a28684ca744688457a35481999c24405542c52d06014d7714c0df9020b403
|
| 3 |
+
size 4857206904
|
model-00004-of-00006.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5dc4996a448e3032c5bb441cacfc2d4b0f8a80ade0fe29be0c24cfeedc4377e9
|
| 3 |
+
size 4857206904
|
model-00005-of-00006.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:417e67eb6b030cec9c2ba38d31fe76537c9a44290599ed8d0848d480456e00da
|
| 3 |
+
size 4857206904
|
model-00006-of-00006.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a1eb2333082a5e14cf1b05a76620a9752ca8db2488434b9d5e198c0f48513de2
|
| 3 |
+
size 2684472112
|
model.safetensors.index.json
ADDED
|
@@ -0,0 +1,298 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metadata": {
|
| 3 |
+
"total_size": 26953662464
|
| 4 |
+
},
|
| 5 |
+
"weight_map": {
|
| 6 |
+
"lm_head.weight": "model-00006-of-00006.safetensors",
|
| 7 |
+
"model.embed_tokens.weight": "model-00001-of-00006.safetensors",
|
| 8 |
+
"model.layers.0.input_layernorm.weight": "model-00001-of-00006.safetensors",
|
| 9 |
+
"model.layers.0.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
|
| 10 |
+
"model.layers.0.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
|
| 11 |
+
"model.layers.0.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
|
| 12 |
+
"model.layers.0.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
|
| 13 |
+
"model.layers.0.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
|
| 14 |
+
"model.layers.0.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
|
| 15 |
+
"model.layers.0.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
|
| 16 |
+
"model.layers.0.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
|
| 17 |
+
"model.layers.1.input_layernorm.weight": "model-00001-of-00006.safetensors",
|
| 18 |
+
"model.layers.1.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
|
| 19 |
+
"model.layers.1.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
|
| 20 |
+
"model.layers.1.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
|
| 21 |
+
"model.layers.1.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
|
| 22 |
+
"model.layers.1.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
|
| 23 |
+
"model.layers.1.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
|
| 24 |
+
"model.layers.1.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
|
| 25 |
+
"model.layers.1.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
|
| 26 |
+
"model.layers.10.input_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 27 |
+
"model.layers.10.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
|
| 28 |
+
"model.layers.10.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
|
| 29 |
+
"model.layers.10.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
|
| 30 |
+
"model.layers.10.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 31 |
+
"model.layers.10.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
|
| 32 |
+
"model.layers.10.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
|
| 33 |
+
"model.layers.10.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
|
| 34 |
+
"model.layers.10.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
|
| 35 |
+
"model.layers.11.input_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 36 |
+
"model.layers.11.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
|
| 37 |
+
"model.layers.11.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
|
| 38 |
+
"model.layers.11.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
|
| 39 |
+
"model.layers.11.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 40 |
+
"model.layers.11.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
|
| 41 |
+
"model.layers.11.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
|
| 42 |
+
"model.layers.11.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
|
| 43 |
+
"model.layers.11.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
|
| 44 |
+
"model.layers.12.input_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 45 |
+
"model.layers.12.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
|
| 46 |
+
"model.layers.12.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
|
| 47 |
+
"model.layers.12.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
|
| 48 |
+
"model.layers.12.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 49 |
+
"model.layers.12.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
|
| 50 |
+
"model.layers.12.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
|
| 51 |
+
"model.layers.12.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
|
| 52 |
+
"model.layers.12.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
|
| 53 |
+
"model.layers.13.input_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 54 |
+
"model.layers.13.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
|
| 55 |
+
"model.layers.13.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
|
| 56 |
+
"model.layers.13.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
|
| 57 |
+
"model.layers.13.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 58 |
+
"model.layers.13.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
|
| 59 |
+
"model.layers.13.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
|
| 60 |
+
"model.layers.13.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
|
| 61 |
+
"model.layers.13.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
|
| 62 |
+
"model.layers.14.input_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 63 |
+
"model.layers.14.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
|
| 64 |
+
"model.layers.14.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
|
| 65 |
+
"model.layers.14.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
|
| 66 |
+
"model.layers.14.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 67 |
+
"model.layers.14.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
|
| 68 |
+
"model.layers.14.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
|
| 69 |
+
"model.layers.14.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
|
| 70 |
+
"model.layers.14.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
|
| 71 |
+
"model.layers.15.input_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 72 |
+
"model.layers.15.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
|
| 73 |
+
"model.layers.15.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
|
| 74 |
+
"model.layers.15.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
|
| 75 |
+
"model.layers.15.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 76 |
+
"model.layers.15.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
|
| 77 |
+
"model.layers.15.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
|
| 78 |
+
"model.layers.15.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
|
| 79 |
+
"model.layers.15.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
|
| 80 |
+
"model.layers.16.input_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 81 |
+
"model.layers.16.mlp.down_proj.weight": "model-00003-of-00006.safetensors",
|
| 82 |
+
"model.layers.16.mlp.gate_proj.weight": "model-00003-of-00006.safetensors",
|
| 83 |
+
"model.layers.16.mlp.up_proj.weight": "model-00003-of-00006.safetensors",
|
| 84 |
+
"model.layers.16.post_attention_layernorm.weight": "model-00003-of-00006.safetensors",
|
| 85 |
+
"model.layers.16.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
|
| 86 |
+
"model.layers.16.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
|
| 87 |
+
"model.layers.16.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
|
| 88 |
+
"model.layers.16.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
|
| 89 |
+
"model.layers.17.input_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 90 |
+
"model.layers.17.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
|
| 91 |
+
"model.layers.17.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
|
| 92 |
+
"model.layers.17.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
|
| 93 |
+
"model.layers.17.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 94 |
+
"model.layers.17.self_attn.k_proj.weight": "model-00003-of-00006.safetensors",
|
| 95 |
+
"model.layers.17.self_attn.o_proj.weight": "model-00003-of-00006.safetensors",
|
| 96 |
+
"model.layers.17.self_attn.q_proj.weight": "model-00003-of-00006.safetensors",
|
| 97 |
+
"model.layers.17.self_attn.v_proj.weight": "model-00003-of-00006.safetensors",
|
| 98 |
+
"model.layers.18.input_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 99 |
+
"model.layers.18.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
|
| 100 |
+
"model.layers.18.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
|
| 101 |
+
"model.layers.18.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
|
| 102 |
+
"model.layers.18.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 103 |
+
"model.layers.18.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
|
| 104 |
+
"model.layers.18.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
|
| 105 |
+
"model.layers.18.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
|
| 106 |
+
"model.layers.18.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
|
| 107 |
+
"model.layers.19.input_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 108 |
+
"model.layers.19.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
|
| 109 |
+
"model.layers.19.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
|
| 110 |
+
"model.layers.19.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
|
| 111 |
+
"model.layers.19.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 112 |
+
"model.layers.19.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
|
| 113 |
+
"model.layers.19.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
|
| 114 |
+
"model.layers.19.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
|
| 115 |
+
"model.layers.19.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
|
| 116 |
+
"model.layers.2.input_layernorm.weight": "model-00001-of-00006.safetensors",
|
| 117 |
+
"model.layers.2.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
|
| 118 |
+
"model.layers.2.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
|
| 119 |
+
"model.layers.2.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
|
| 120 |
+
"model.layers.2.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
|
| 121 |
+
"model.layers.2.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
|
| 122 |
+
"model.layers.2.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
|
| 123 |
+
"model.layers.2.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
|
| 124 |
+
"model.layers.2.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
|
| 125 |
+
"model.layers.20.input_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 126 |
+
"model.layers.20.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
|
| 127 |
+
"model.layers.20.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
|
| 128 |
+
"model.layers.20.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
|
| 129 |
+
"model.layers.20.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 130 |
+
"model.layers.20.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
|
| 131 |
+
"model.layers.20.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
|
| 132 |
+
"model.layers.20.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
|
| 133 |
+
"model.layers.20.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
|
| 134 |
+
"model.layers.21.input_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 135 |
+
"model.layers.21.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
|
| 136 |
+
"model.layers.21.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
|
| 137 |
+
"model.layers.21.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
|
| 138 |
+
"model.layers.21.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 139 |
+
"model.layers.21.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
|
| 140 |
+
"model.layers.21.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
|
| 141 |
+
"model.layers.21.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
|
| 142 |
+
"model.layers.21.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
|
| 143 |
+
"model.layers.22.input_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 144 |
+
"model.layers.22.mlp.down_proj.weight": "model-00004-of-00006.safetensors",
|
| 145 |
+
"model.layers.22.mlp.gate_proj.weight": "model-00004-of-00006.safetensors",
|
| 146 |
+
"model.layers.22.mlp.up_proj.weight": "model-00004-of-00006.safetensors",
|
| 147 |
+
"model.layers.22.post_attention_layernorm.weight": "model-00004-of-00006.safetensors",
|
| 148 |
+
"model.layers.22.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
|
| 149 |
+
"model.layers.22.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
|
| 150 |
+
"model.layers.22.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
|
| 151 |
+
"model.layers.22.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
|
| 152 |
+
"model.layers.23.input_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 153 |
+
"model.layers.23.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
|
| 154 |
+
"model.layers.23.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
|
| 155 |
+
"model.layers.23.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
|
| 156 |
+
"model.layers.23.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 157 |
+
"model.layers.23.self_attn.k_proj.weight": "model-00004-of-00006.safetensors",
|
| 158 |
+
"model.layers.23.self_attn.o_proj.weight": "model-00004-of-00006.safetensors",
|
| 159 |
+
"model.layers.23.self_attn.q_proj.weight": "model-00004-of-00006.safetensors",
|
| 160 |
+
"model.layers.23.self_attn.v_proj.weight": "model-00004-of-00006.safetensors",
|
| 161 |
+
"model.layers.24.input_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 162 |
+
"model.layers.24.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
|
| 163 |
+
"model.layers.24.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
|
| 164 |
+
"model.layers.24.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
|
| 165 |
+
"model.layers.24.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 166 |
+
"model.layers.24.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
|
| 167 |
+
"model.layers.24.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
|
| 168 |
+
"model.layers.24.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
|
| 169 |
+
"model.layers.24.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
|
| 170 |
+
"model.layers.25.input_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 171 |
+
"model.layers.25.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
|
| 172 |
+
"model.layers.25.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
|
| 173 |
+
"model.layers.25.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
|
| 174 |
+
"model.layers.25.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 175 |
+
"model.layers.25.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
|
| 176 |
+
"model.layers.25.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
|
| 177 |
+
"model.layers.25.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
|
| 178 |
+
"model.layers.25.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
|
| 179 |
+
"model.layers.26.input_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 180 |
+
"model.layers.26.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
|
| 181 |
+
"model.layers.26.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
|
| 182 |
+
"model.layers.26.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
|
| 183 |
+
"model.layers.26.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 184 |
+
"model.layers.26.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
|
| 185 |
+
"model.layers.26.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
|
| 186 |
+
"model.layers.26.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
|
| 187 |
+
"model.layers.26.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
|
| 188 |
+
"model.layers.27.input_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 189 |
+
"model.layers.27.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
|
| 190 |
+
"model.layers.27.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
|
| 191 |
+
"model.layers.27.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
|
| 192 |
+
"model.layers.27.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 193 |
+
"model.layers.27.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
|
| 194 |
+
"model.layers.27.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
|
| 195 |
+
"model.layers.27.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
|
| 196 |
+
"model.layers.27.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
|
| 197 |
+
"model.layers.28.input_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 198 |
+
"model.layers.28.mlp.down_proj.weight": "model-00005-of-00006.safetensors",
|
| 199 |
+
"model.layers.28.mlp.gate_proj.weight": "model-00005-of-00006.safetensors",
|
| 200 |
+
"model.layers.28.mlp.up_proj.weight": "model-00005-of-00006.safetensors",
|
| 201 |
+
"model.layers.28.post_attention_layernorm.weight": "model-00005-of-00006.safetensors",
|
| 202 |
+
"model.layers.28.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
|
| 203 |
+
"model.layers.28.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
|
| 204 |
+
"model.layers.28.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
|
| 205 |
+
"model.layers.28.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
|
| 206 |
+
"model.layers.29.input_layernorm.weight": "model-00006-of-00006.safetensors",
|
| 207 |
+
"model.layers.29.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
|
| 208 |
+
"model.layers.29.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
|
| 209 |
+
"model.layers.29.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
|
| 210 |
+
"model.layers.29.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
|
| 211 |
+
"model.layers.29.self_attn.k_proj.weight": "model-00005-of-00006.safetensors",
|
| 212 |
+
"model.layers.29.self_attn.o_proj.weight": "model-00005-of-00006.safetensors",
|
| 213 |
+
"model.layers.29.self_attn.q_proj.weight": "model-00005-of-00006.safetensors",
|
| 214 |
+
"model.layers.29.self_attn.v_proj.weight": "model-00005-of-00006.safetensors",
|
| 215 |
+
"model.layers.3.input_layernorm.weight": "model-00001-of-00006.safetensors",
|
| 216 |
+
"model.layers.3.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
|
| 217 |
+
"model.layers.3.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
|
| 218 |
+
"model.layers.3.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
|
| 219 |
+
"model.layers.3.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
|
| 220 |
+
"model.layers.3.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
|
| 221 |
+
"model.layers.3.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
|
| 222 |
+
"model.layers.3.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
|
| 223 |
+
"model.layers.3.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
|
| 224 |
+
"model.layers.30.input_layernorm.weight": "model-00006-of-00006.safetensors",
|
| 225 |
+
"model.layers.30.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
|
| 226 |
+
"model.layers.30.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
|
| 227 |
+
"model.layers.30.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
|
| 228 |
+
"model.layers.30.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
|
| 229 |
+
"model.layers.30.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
|
| 230 |
+
"model.layers.30.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
|
| 231 |
+
"model.layers.30.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
|
| 232 |
+
"model.layers.30.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
|
| 233 |
+
"model.layers.31.input_layernorm.weight": "model-00006-of-00006.safetensors",
|
| 234 |
+
"model.layers.31.mlp.down_proj.weight": "model-00006-of-00006.safetensors",
|
| 235 |
+
"model.layers.31.mlp.gate_proj.weight": "model-00006-of-00006.safetensors",
|
| 236 |
+
"model.layers.31.mlp.up_proj.weight": "model-00006-of-00006.safetensors",
|
| 237 |
+
"model.layers.31.post_attention_layernorm.weight": "model-00006-of-00006.safetensors",
|
| 238 |
+
"model.layers.31.self_attn.k_proj.weight": "model-00006-of-00006.safetensors",
|
| 239 |
+
"model.layers.31.self_attn.o_proj.weight": "model-00006-of-00006.safetensors",
|
| 240 |
+
"model.layers.31.self_attn.q_proj.weight": "model-00006-of-00006.safetensors",
|
| 241 |
+
"model.layers.31.self_attn.v_proj.weight": "model-00006-of-00006.safetensors",
|
| 242 |
+
"model.layers.4.input_layernorm.weight": "model-00001-of-00006.safetensors",
|
| 243 |
+
"model.layers.4.mlp.down_proj.weight": "model-00001-of-00006.safetensors",
|
| 244 |
+
"model.layers.4.mlp.gate_proj.weight": "model-00001-of-00006.safetensors",
|
| 245 |
+
"model.layers.4.mlp.up_proj.weight": "model-00001-of-00006.safetensors",
|
| 246 |
+
"model.layers.4.post_attention_layernorm.weight": "model-00001-of-00006.safetensors",
|
| 247 |
+
"model.layers.4.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
|
| 248 |
+
"model.layers.4.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
|
| 249 |
+
"model.layers.4.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
|
| 250 |
+
"model.layers.4.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
|
| 251 |
+
"model.layers.5.input_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 252 |
+
"model.layers.5.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
|
| 253 |
+
"model.layers.5.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
|
| 254 |
+
"model.layers.5.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
|
| 255 |
+
"model.layers.5.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 256 |
+
"model.layers.5.self_attn.k_proj.weight": "model-00001-of-00006.safetensors",
|
| 257 |
+
"model.layers.5.self_attn.o_proj.weight": "model-00001-of-00006.safetensors",
|
| 258 |
+
"model.layers.5.self_attn.q_proj.weight": "model-00001-of-00006.safetensors",
|
| 259 |
+
"model.layers.5.self_attn.v_proj.weight": "model-00001-of-00006.safetensors",
|
| 260 |
+
"model.layers.6.input_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 261 |
+
"model.layers.6.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
|
| 262 |
+
"model.layers.6.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
|
| 263 |
+
"model.layers.6.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
|
| 264 |
+
"model.layers.6.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 265 |
+
"model.layers.6.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
|
| 266 |
+
"model.layers.6.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
|
| 267 |
+
"model.layers.6.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
|
| 268 |
+
"model.layers.6.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
|
| 269 |
+
"model.layers.7.input_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 270 |
+
"model.layers.7.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
|
| 271 |
+
"model.layers.7.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
|
| 272 |
+
"model.layers.7.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
|
| 273 |
+
"model.layers.7.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 274 |
+
"model.layers.7.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
|
| 275 |
+
"model.layers.7.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
|
| 276 |
+
"model.layers.7.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
|
| 277 |
+
"model.layers.7.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
|
| 278 |
+
"model.layers.8.input_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 279 |
+
"model.layers.8.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
|
| 280 |
+
"model.layers.8.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
|
| 281 |
+
"model.layers.8.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
|
| 282 |
+
"model.layers.8.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 283 |
+
"model.layers.8.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
|
| 284 |
+
"model.layers.8.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
|
| 285 |
+
"model.layers.8.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
|
| 286 |
+
"model.layers.8.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
|
| 287 |
+
"model.layers.9.input_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 288 |
+
"model.layers.9.mlp.down_proj.weight": "model-00002-of-00006.safetensors",
|
| 289 |
+
"model.layers.9.mlp.gate_proj.weight": "model-00002-of-00006.safetensors",
|
| 290 |
+
"model.layers.9.mlp.up_proj.weight": "model-00002-of-00006.safetensors",
|
| 291 |
+
"model.layers.9.post_attention_layernorm.weight": "model-00002-of-00006.safetensors",
|
| 292 |
+
"model.layers.9.self_attn.k_proj.weight": "model-00002-of-00006.safetensors",
|
| 293 |
+
"model.layers.9.self_attn.o_proj.weight": "model-00002-of-00006.safetensors",
|
| 294 |
+
"model.layers.9.self_attn.q_proj.weight": "model-00002-of-00006.safetensors",
|
| 295 |
+
"model.layers.9.self_attn.v_proj.weight": "model-00002-of-00006.safetensors",
|
| 296 |
+
"model.norm.weight": "model-00006-of-00006.safetensors"
|
| 297 |
+
}
|
| 298 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"unk_token": {
|
| 17 |
+
"content": "<unk>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
}
|
| 23 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"0": {
|
| 6 |
+
"content": "<unk>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"1": {
|
| 14 |
+
"content": "<s>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"2": {
|
| 22 |
+
"content": "</s>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
"bos_token": "<s>",
|
| 31 |
+
"clean_up_tokenization_spaces": false,
|
| 32 |
+
"eos_token": "</s>",
|
| 33 |
+
"legacy": false,
|
| 34 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 35 |
+
"pad_token": null,
|
| 36 |
+
"padding_side": "right",
|
| 37 |
+
"sp_model_kwargs": {},
|
| 38 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 39 |
+
"unk_token": "<unk>",
|
| 40 |
+
"use_default_system_prompt": false
|
| 41 |
+
}
|
truthfulqa_mc_0shot.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"truthfulqa_mc": {
|
| 4 |
+
"mc1": 0.31211750305997554,
|
| 5 |
+
"mc1_stderr": 0.016220756769520926,
|
| 6 |
+
"mc2": 0.44918598673219645,
|
| 7 |
+
"mc2_stderr": 0.014886808872838312
|
| 8 |
+
}
|
| 9 |
+
},
|
| 10 |
+
"versions": {
|
| 11 |
+
"truthfulqa_mc": 1
|
| 12 |
+
},
|
| 13 |
+
"config": {
|
| 14 |
+
"model": "hf-causal-experimental",
|
| 15 |
+
"model_args": {
|
| 16 |
+
"pretrained": "/network/alexandre/research/llama2_7b_ultrachat/dense/dense_finetuning/dense_LR1e-4_E1/training",
|
| 17 |
+
"trust_remote_code": true
|
| 18 |
+
},
|
| 19 |
+
"num_fewshot": 0,
|
| 20 |
+
"batch_size": "8",
|
| 21 |
+
"batch_sizes": [],
|
| 22 |
+
"device": "cuda:5",
|
| 23 |
+
"no_cache": true,
|
| 24 |
+
"limit": null,
|
| 25 |
+
"bootstrap_iters": 100000,
|
| 26 |
+
"description_dict": {}
|
| 27 |
+
}
|
| 28 |
+
}
|
truthfulqa_mc_0shot_bs16_fp32.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"truthfulqa_mc": {
|
| 4 |
+
"mc1": 0.3108935128518972,
|
| 5 |
+
"mc1_stderr": 0.016203316673559693,
|
| 6 |
+
"mc2": 0.44831514010134543,
|
| 7 |
+
"mc2_stderr": 0.014913413696247348
|
| 8 |
+
}
|
| 9 |
+
},
|
| 10 |
+
"versions": {
|
| 11 |
+
"truthfulqa_mc": 1
|
| 12 |
+
},
|
| 13 |
+
"config": {
|
| 14 |
+
"model": "sparseml",
|
| 15 |
+
"model_args": "pretrained=/nm/drive1/alexandre/zoo/llama2-7b-ultrachat200k_llama2_pretrain-base/training,dtype=float32",
|
| 16 |
+
"num_fewshot": 0,
|
| 17 |
+
"batch_size": "16",
|
| 18 |
+
"batch_sizes": [],
|
| 19 |
+
"device": "cuda:3",
|
| 20 |
+
"no_cache": true,
|
| 21 |
+
"limit": null,
|
| 22 |
+
"bootstrap_iters": 100000,
|
| 23 |
+
"description_dict": {}
|
| 24 |
+
}
|
| 25 |
+
}
|
winogrande_5shot_bs16_fp32.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"winogrande": {
|
| 4 |
+
"acc": 0.7261247040252565,
|
| 5 |
+
"acc_stderr": 0.012533292732620297
|
| 6 |
+
}
|
| 7 |
+
},
|
| 8 |
+
"versions": {
|
| 9 |
+
"winogrande": 0
|
| 10 |
+
},
|
| 11 |
+
"config": {
|
| 12 |
+
"model": "sparseml",
|
| 13 |
+
"model_args": "pretrained=/nm/drive1/alexandre/zoo/llama2-7b-ultrachat200k_llama2_pretrain-base/training,dtype=float32",
|
| 14 |
+
"num_fewshot": 5,
|
| 15 |
+
"batch_size": "16",
|
| 16 |
+
"batch_sizes": [],
|
| 17 |
+
"device": "cuda:3",
|
| 18 |
+
"no_cache": true,
|
| 19 |
+
"limit": null,
|
| 20 |
+
"bootstrap_iters": 100000,
|
| 21 |
+
"description_dict": {}
|
| 22 |
+
}
|
| 23 |
+
}
|