{ "config_general": { "lighteval_sha": "?", "num_fewshot_seeds": 1, "override_batch_size": 1, "max_samples": null, "job_id": "", "start_time": 485337.70357421, "end_time": 491945.351292691, "total_evaluation_time_secondes": "6607.647718481021", "model_name": "edbeeching/mixtral-8x7b-instruct-v0.1_merged", "model_sha": "02e5e37da870d6382b35fd16e9545bbab51dd218", "model_dtype": "torch.bfloat16", "model_size": "87.49 GB", "config": null }, "results": { "leaderboard|hellaswag|10": { "acc": 0.5386377215694085, "acc_stderr": 0.004974860878464429, "acc_norm": 0.7305317665803625, "acc_norm_stderr": 0.0044277679963016375 }, "all": { "acc": 0.5386377215694085, "acc_stderr": 0.004974860878464429, "acc_norm": 0.7305317665803625, "acc_norm_stderr": 0.0044277679963016375 } }, "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": "a3c5cf19ab983927", "hash_input_tokens": "f340cb1674dbe5ed", "hash_cont_tokens": "cf1fd30525a7ce71" }, "truncated": 0, "non_truncated": 10042, "padded": 40011, "non_padded": 157, "effective_few_shots": 10.0, "num_truncated_few_shots": 0 } }, "summary_general": { "hashes": { "hash_examples": "63bc2cf8bae03fbc", "hash_full_prompts": "af0716edcec0366b", "hash_input_tokens": "4ac9ac0520577724", "hash_cont_tokens": "605832a579f190e3" }, "truncated": 0, "non_truncated": 10042, "padded": 40011, "non_padded": 157, "num_truncated_few_shots": 0 } }