{
  "config_general": {
    "lighteval_sha": "?",
    "num_fewshot_seeds": 1,
    "override_batch_size": null,
    "max_samples": null,
    "job_id": 0,
    "start_time": 48811.364506419,
    "end_time": 49377.553754935,
    "total_evaluation_time_secondes": "566.1892485159988",
    "model_name": "misdelivery/Llamafy-Qwen2.5-0.5B-ck4000",
    "model_sha": "",
    "model_dtype": null,
    "model_size": null
  },
  "results": {
    "helm|mmlu|5": {
      "em": 0.22760290556900725,
      "em_stderr": 0.0035384266475051054,
      "qem": 0.2313060817547358,
      "qem_stderr": 0.0035585348941833826,
      "pem": 0.23657598632673407,
      "pem_stderr": 0.0035864866480669385,
      "pqem": 0.3824241561031192,
      "pqem_stderr": 0.004101274040144867
    },
    "all": {
      "em": 0.22760290556900725,
      "em_stderr": 0.0035384266475051054,
      "qem": 0.2313060817547358,
      "qem_stderr": 0.0035585348941833826,
      "pem": 0.23657598632673407,
      "pem_stderr": 0.0035864866480669385,
      "pqem": 0.3824241561031192,
      "pqem_stderr": 0.004101274040144867
    }
  },
  "versions": {
    "helm|mmlu|5": 0
  },
  "config_tasks": {
    "helm|mmlu": {
      "name": "mmlu",
      "prompt_function": "mmlu_helm",
      "hf_repo": "lighteval/mmlu",
      "hf_subset": "all",
      "metric": [
        {
          "metric_name": "em",
          "higher_is_better": true,
          "category": "3",
          "use_case": "1",
          "sample_level_fn": "compute",
          "corpus_level_fn": "mean"
        },
        {
          "metric_name": "qem",
          "higher_is_better": true,
          "category": "3",
          "use_case": "1",
          "sample_level_fn": "compute",
          "corpus_level_fn": "mean"
        },
        {
          "metric_name": "pem",
          "higher_is_better": true,
          "category": "3",
          "use_case": "1",
          "sample_level_fn": "compute",
          "corpus_level_fn": "mean"
        },
        {
          "metric_name": "pqem",
          "higher_is_better": true,
          "category": "3",
          "use_case": "1",
          "sample_level_fn": "compute",
          "corpus_level_fn": "mean"
        }
      ],
      "hf_revision": null,
      "hf_filter": null,
      "hf_avail_splits": [
        "auxiliary_train",
        "test",
        "validation",
        "dev"
      ],
      "trust_dataset": true,
      "evaluation_splits": [
        "test"
      ],
      "few_shots_split": "dev",
      "few_shots_select": null,
      "generation_size": 5,
      "generation_grammar": null,
      "stop_sequence": [
        "\n"
      ],
      "output_regex": null,
      "num_samples": null,
      "suite": [
        "helm",
        "helm_general"
      ],
      "original_num_docs": 14042,
      "effective_num_docs": 14042,
      "must_remove_duplicate_docs": false,
      "version": 0,
      "frozen": false
    }
  },
  "summary_tasks": {
    "helm|mmlu|5": {
      "hashes": {
        "hash_examples": "2c780174cf34ca88",
        "hash_full_prompts": "561b71727da359f2",
        "hash_input_tokens": "78408df35958ce1d",
        "hash_cont_tokens": "ded0c79257d344c8"
      },
      "truncated": 0,
      "non_truncated": 14042,
      "padded": 0,
      "non_padded": 14042,
      "effective_few_shots": 5.0,
      "num_truncated_few_shots": 0
    }
  },
  "summary_general": {
    "hashes": {
      "hash_examples": "a9bb07d546ec2e61",
      "hash_full_prompts": "9d51ca587bc5fc37",
      "hash_input_tokens": "31b0410ab1422ff1",
      "hash_cont_tokens": "23b01204a340facc"
    },
    "truncated": 0,
    "non_truncated": 14042,
    "padded": 0,
    "non_padded": 14042,
    "num_truncated_few_shots": 0
  }
}