{ "config_general": { "lighteval_sha": "?", "num_fewshot_seeds": 1, "override_batch_size": 4, "max_samples": null, "job_id": "", "start_time": 1117.895192294, "end_time": 5167.91069733, "total_evaluation_time_secondes": "4050.0155050360004", "model_name": "abhishek/autotrain-mixtral-8x7b-orpo-v1", "model_sha": "a8be37cf01ad767a0c71e0ba3af29c0b3ebcb559", "model_dtype": "torch.bfloat16", "model_size": "87.49 GB", "config": null }, "results": { "leaderboard|hellaswag|10": { "acc": 0.6608245369448317, "acc_stderr": 0.004724619193427587, "acc_norm": 0.8556064528978291, "acc_norm_stderr": 0.003507699935074236 }, "all": { "acc": 0.6608245369448317, "acc_stderr": 0.004724619193427587, "acc_norm": 0.8556064528978291, "acc_norm_stderr": 0.003507699935074236 } }, "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 } }