abhishek's picture
Upload eval_results/abhishek/autotrain-mixtral-8x7b-orpo-v2/main/hellaswag/results_2024-05-02T06-41-19.551430.json with huggingface_hub
de87573 verified
raw
history blame
2.59 kB
{
"config_general": {
"lighteval_sha": "?",
"num_fewshot_seeds": 1,
"override_batch_size": 4,
"max_samples": null,
"job_id": "",
"start_time": 162588.026645642,
"end_time": 166408.715521214,
"total_evaluation_time_secondes": "3820.6888755719992",
"model_name": "abhishek/autotrain-mixtral-8x7b-orpo-v2",
"model_sha": "dcc2c4efe12c26fc1ec211f68cc139d61de5ce84",
"model_dtype": "torch.bfloat16",
"model_size": "87.49 GB",
"config": null
},
"results": {
"leaderboard|hellaswag|10": {
"acc": 0.648177653853814,
"acc_stderr": 0.004765629263643524,
"acc_norm": 0.8376817367058355,
"acc_norm_stderr": 0.0036798891253998164
},
"all": {
"acc": 0.648177653853814,
"acc_stderr": 0.004765629263643524,
"acc_norm": 0.8376817367058355,
"acc_norm_stderr": 0.0036798891253998164
}
},
"versions": {
"leaderboard|hellaswag|10": 0
},
"config_tasks": {
"leaderboard|hellaswag": {
"name": "hellaswag",
"prompt_function": "hellaswag_harness",
"hf_repo": "hellaswag",
"hf_subset": "default",
"metric": [
"loglikelihood_acc",
"loglikelihood_acc_norm"
],
"hf_avail_splits": [
"train",
"test",
"validation"
],
"evaluation_splits": [
"validation"
],
"few_shots_split": null,
"few_shots_select": "random_sampling_from_train",
"generation_size": -1,
"stop_sequence": [
"\n"
],
"output_regex": null,
"frozen": false,
"suite": [
"leaderboard"
],
"original_num_docs": 10042,
"effective_num_docs": 10042,
"trust_dataset": true,
"must_remove_duplicate_docs": null
}
},
"summary_tasks": {
"leaderboard|hellaswag|10": {
"hashes": {
"hash_examples": "31985c805c3a737e",
"hash_full_prompts": "451e22e8c3aaeeb6",
"hash_input_tokens": "211987b65c444c03",
"hash_cont_tokens": "01fbe12ab59c35bc"
},
"truncated": 0,
"non_truncated": 10042,
"padded": 40095,
"non_padded": 73,
"effective_few_shots": 10.0,
"num_truncated_few_shots": 0
}
},
"summary_general": {
"hashes": {
"hash_examples": "63bc2cf8bae03fbc",
"hash_full_prompts": "b2452b7405d0326a",
"hash_input_tokens": "c32a410c27b9a570",
"hash_cont_tokens": "ff85c47746c926c3"
},
"truncated": 0,
"non_truncated": 10042,
"padded": 40095,
"non_padded": 73,
"num_truncated_few_shots": 0
}
}