lewtun's picture
lewtun HF staff
Upload eval_results/NousResearch/Nous-Hermes-2-Yi-34B/main/arc/results_2024-03-04T20-43-53.422268.json with huggingface_hub
133c4ea verified
raw
history blame
2.32 kB
{
"config_general": {
"lighteval_sha": "?",
"num_fewshot_seeds": 1,
"override_batch_size": 1,
"max_samples": null,
"job_id": "",
"start_time": 3888546.84720616,
"end_time": 3889465.289648732,
"total_evaluation_time_secondes": "918.44244257221",
"model_name": "NousResearch/Nous-Hermes-2-Yi-34B",
"model_sha": "fcb0a8847e76aea14aba9aa44009d4418ad7c18f",
"model_dtype": "torch.bfloat16",
"model_size": "64.17 GB",
"config": null
},
"results": {
"lighteval|arc:challenge|25": {
"acc": 0.6544368600682594,
"acc_stderr": 0.013896938461145675,
"acc_norm": 0.6800341296928327,
"acc_norm_stderr": 0.013631345807016193
}
},
"versions": {
"lighteval|arc:challenge|25": 0
},
"config_tasks": {
"lighteval|arc:challenge": {
"name": "arc:challenge",
"prompt_function": "arc",
"hf_repo": "ai2_arc",
"hf_subset": "ARC-Challenge",
"metric": [
"loglikelihood_acc",
"loglikelihood_acc_norm_nospace"
],
"hf_avail_splits": [
"train",
"test"
],
"evaluation_splits": [
"test"
],
"few_shots_split": null,
"few_shots_select": "random_sampling_from_train",
"generation_size": 1,
"stop_sequence": [
"\n"
],
"output_regex": null,
"frozen": false,
"suite": [
"lighteval",
"arc"
],
"original_num_docs": 1172,
"effective_num_docs": 1172
}
},
"summary_tasks": {
"lighteval|arc:challenge|25": {
"hashes": {
"hash_examples": "17b0cae357c0259e",
"hash_full_prompts": "66498cebc864f4ca",
"hash_input_tokens": "50ff51a31ee7a624",
"hash_cont_tokens": "a3621012b91cd763"
},
"truncated": 0,
"non_truncated": 1172,
"padded": 4660,
"non_padded": 27,
"effective_few_shots": 25.0,
"num_truncated_few_shots": 0
}
},
"summary_general": {
"hashes": {
"hash_examples": "aaa6929c6d3771fb",
"hash_full_prompts": "848db321e3b0d4cd",
"hash_input_tokens": "c67395f2f3e4e07d",
"hash_cont_tokens": "9756e6fe6fe4e9bd"
},
"truncated": 0,
"non_truncated": 1172,
"padded": 4660,
"non_padded": 27,
"num_truncated_few_shots": 0
}
}