eval_results/alignment-handbook/zephyr-2b-gemma-dpo-mix7-beta-0.05-epoch-6/main/hellaswag/results_2024-03-07T11-05-53.241131.json

{
  "config_general": {
    "lighteval_sha": "?",
    "num_fewshot_seeds": 1,
    "override_batch_size": 1,
    "max_samples": null,
    "job_id": "",
    "start_time": 2313055.015010817,
    "end_time": 2313519.405629995,
    "total_evaluation_time_secondes": "464.3906191778369",
    "model_name": "alignment-handbook/zephyr-2b-gemma-dpo-mix7-beta-0.05-epoch-6",
    "model_sha": "1ecb6ade923a19d209a5ffe161833e4fe42c3580",
    "model_dtype": "torch.bfloat16",
    "model_size": "4.68 GB",
    "config": null
  },
  "results": {
    "lighteval|hellaswag|10": {
      "acc": 0.5403306114319857,
      "acc_stderr": 0.004973522582431218,
      "acc_norm": 0.7210714997012547,
      "acc_norm_stderr": 0.004475557360359704
    }
  },
  "versions": {
    "lighteval|hellaswag|10": 0
  },
  "config_tasks": {
    "lighteval|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": [
        "lighteval"
      ],
      "original_num_docs": 10042,
      "effective_num_docs": 10042
    }
  },
  "summary_tasks": {
    "lighteval|hellaswag|10": {
      "hashes": {
        "hash_examples": "31985c805c3a737e",
        "hash_full_prompts": "48de8fbcf45195b3",
        "hash_input_tokens": "08c832518b7403b1",
        "hash_cont_tokens": "6232829aec74d42d"
      },
      "truncated": 0,
      "non_truncated": 10042,
      "padded": 39905,
      "non_padded": 263,
      "effective_few_shots": 10.0,
      "num_truncated_few_shots": 0
    }
  },
  "summary_general": {
    "hashes": {
      "hash_examples": "63bc2cf8bae03fbc",
      "hash_full_prompts": "20d63c990d3e8572",
      "hash_input_tokens": "e93858038f602554",
      "hash_cont_tokens": "c2b17126697982c3"
    },
    "truncated": 0,
    "non_truncated": 10042,
    "padded": 39905,
    "non_padded": 263,
    "num_truncated_few_shots": 0
  }
}