eval_results/alignment-handbook/zephyr-2b-gemma-dpo-hermes-beta-0.2-epoch-2/main/arc/results_2024-03-05T20-26-38.488915.json

{
  "config_general": {
    "lighteval_sha": "?",
    "num_fewshot_seeds": 1,
    "override_batch_size": 1,
    "max_samples": null,
    "job_id": "",
    "start_time": 3974734.891065999,
    "end_time": 3974830.377311283,
    "total_evaluation_time_secondes": "95.48624528385699",
    "model_name": "alignment-handbook/zephyr-2b-gemma-dpo-hermes-beta-0.2-epoch-2",
    "model_sha": "12843c7d670fab1aeb7d040d8fa406bae52f656b",
    "model_dtype": "torch.bfloat16",
    "model_size": "4.68 GB",
    "config": null
  },
  "results": {
    "lighteval|arc:challenge|25": {
      "acc": 0.4778156996587031,
      "acc_stderr": 0.014597001927076138,
      "acc_norm": 0.48208191126279865,
      "acc_norm_stderr": 0.01460200558549098
    }
  },
  "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": "7ebbd32d4124c208",
        "hash_cont_tokens": "3bb85634c18ea13e"
      },
      "truncated": 0,
      "non_truncated": 1172,
      "padded": 4647,
      "non_padded": 40,
      "effective_few_shots": 25.0,
      "num_truncated_few_shots": 0
    }
  },
  "summary_general": {
    "hashes": {
      "hash_examples": "aaa6929c6d3771fb",
      "hash_full_prompts": "848db321e3b0d4cd",
      "hash_input_tokens": "6ceef404485c21dc",
      "hash_cont_tokens": "274bfa3146be24a9"
    },
    "truncated": 0,
    "non_truncated": 1172,
    "padded": 4647,
    "non_padded": 40,
    "num_truncated_few_shots": 0
  }
}