open-llm-leaderboard-old/details_binbi__Ein-72B-v0.1

该数据集是在模型 binbi/Ein-72B-v0.1 在 Open LLM Leaderboard 上的评估运行期间自动创建的。数据集由 63 个配置组成，每个配置对应一个被评估的任务。数据集是从 2 次运行中生成的，每次运行在每个配置中表示为特定的分割。train 分割始终指向最新的结果。此外，还有一个 results 配置存储了所有运行的聚合结果，用于计算和显示 Open LLM Leaderboard 上的聚合指标。README 还提供了如何使用 datasets 库加载数据集的示例，并包含了特定运行的最新结果。

原始信息汇总

数据集概述

该数据集是在对模型 binbi/Ein-72B-v0.1 进行评估运行期间自动创建的，用于 Open LLM Leaderboard。数据集包含 63 个配置，每个配置对应一个评估任务。数据集从 2 次运行中创建，每次运行可以在每个配置中找到特定的分割，分割名称使用运行的时间戳。"train" 分割始终指向最新的结果。

数据集结构

数据集包含多个配置，每个配置对应不同的评估任务。以下是部分配置的详细信息：

配置示例

• 配置名称: harness_arc_challenge_25

  • 分割:
    • 2024_02_05T13_53_38.190866
    • 2024_02_05T14_07_01.496774
    • latest
  • 路径:
    • **/details_harness|arc:challenge|25_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|arc:challenge|25_2024-02-05T14-07-01.496774.parquet

• 配置名称: harness_gsm8k_5

  • 分割:
    • 2024_02_05T13_53_38.190866
    • 2024_02_05T14_07_01.496774
    • latest
  • 路径:
    • **/details_harness|gsm8k|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|gsm8k|5_2024-02-05T14-07-01.496774.parquet

• 配置名称: harness_hellaswag_10

  • 分割:
    • 2024_02_05T13_53_38.190866
    • 2024_02_05T14_07_01.496774
    • latest
  • 路径:
    • **/details_harness|hellaswag|10_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hellaswag|10_2024-02-05T14-07-01.496774.parquet

• 配置名称: harness_hendrycksTest_5

  • 分割:
    • 2024_02_05T13_53_38.190866
    • 2024_02_05T14_07_01.496774
    • latest
  • 路径:
    • **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-anatomy|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-astronomy|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-business_ethics|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-college_biology|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-college_chemistry|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-college_computer_science|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-college_mathematics|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-college_medicine|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-college_physics|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-computer_security|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-econometrics|5_2024-02-05T13-53-38.190866.parquet
    • **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-05T13-53-38.190866.parquet

最新结果

以下是 最新结果 的摘要：

python { "all": { "acc": 0.7714774481143158, "acc_stderr": 0.028033284472043257, "acc_norm": 0.7730249049429745, "acc_norm_stderr": 0.028591667108629, "mc1": 0.6646266829865362, "mc1_stderr": 0.016527534039668987, "mc2": 0.7846937103895764, "mc2_stderr": 0.013816866993252642 }, "harness|arc:challenge|25": { "acc": 0.7448805460750854, "acc_stderr": 0.012739038695202104, "acc_norm": 0.765358361774744, "acc_norm_stderr": 0.012383873560768675 }, "harness|hellaswag|10": { "acc": 0.729735112527385, "acc_stderr": 0.00443188978363381, "acc_norm": 0.8919537940649273, "acc_norm_stderr": 0.0030980431017758325 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.42, "acc_stderr": 0.049604496374885836, "acc_norm": 0.42, "acc_norm_stderr": 0.049604496374885836 }, "harness|hendrycksTest-anatomy|5": { "acc": 0.7185185185185186, "acc_stderr": 0.038850042458002526, "acc_norm": 0.7185185185185186, "acc_norm_stderr": 0.038850042458002526 }, "harness|hendrycksTest-astronomy|5": { "acc": 0.881578947368421, "acc_stderr": 0.026293995855474928, "acc_norm": 0.881578947368421, "acc_norm_stderr": 0.026293995855474928 }, "harness|hendrycksTest-business_ethics|5": { "acc": 0.81, "acc_stderr": 0.03942772444036623, "acc_norm": 0.81, "acc_norm_stderr": 0.03942772444036623 }, "harness|hendrycksTest-clinical_knowledge|5": { "acc": 0.8415094339622642, "acc_stderr": 0.02247652871016772, "acc_norm": 0.8415094339622642, "acc_norm_stderr": 0.02247652871016772 }, "harness|hendrycksTest-college_biology|5": { "acc": 0.9305555555555556, "acc_stderr": 0.021257974822832048, "acc_norm": 0.9305555555555556, "acc_norm_stderr": 0.021257974822832048 }, "harness|hendrycksTest-college_chemistry|5": { "acc": 0.54, "acc_stderr": 0.05009082659620332, "acc_norm": 0.54, "acc_norm_stderr": 0.05009082659620332 }, "harness|hendrycksTest-college_computer_science|5": { "acc": 0.62, "acc_stderr": 0.04878317312145633, "acc_norm": 0.62, "acc_norm_stderr": 0.04878317312145633 }, "harness|hendrycksTest-college_mathematics|5": { "acc": 0.52, "acc_stderr": 0.05021167315686779, "acc_norm": 0.52, "acc_norm_stderr": 0.05021167315686779 }, "harness|hendrycksTest-college_medicine|5": { "acc": 0.7514450867052023, "acc_stderr": 0.03295304696818317, "acc_norm": 0.7514450867052023, "acc_norm_stderr": 0.03295304696818317 }, "harness|hendrycksTest-college_physics|5": { "acc": 0.5588235294117647, "acc_stderr": 0.049406356306056595, "acc_norm": 0.5588235294117647, "acc_norm_stderr": 0.049406356306056595 }, "harness|hendrycksTest-computer_security|5": { "acc": 0.81, "acc_stderr": 0.03942772444036622, "acc_norm": 0.81, "acc_norm_stderr": 0.03942772444036622 }, "harness|hendrycksTest-conceptual_physics|5": { "acc": 0.7872340425531915, "acc_stderr": 0.026754391348039766, "acc_norm": 0.7872340425531915, "acc_norm_stderr": 0.026754391348039766 }, "harness|hendrycksTest-econometrics|5": { "acc": 0.5964912280701754, "acc_stderr": 0.04615186962583707, "acc_norm": 0.5964912280701754, "acc_norm_stderr": 0.04615186962583707 }, "harness|hendrycksTest-electrical_engineering|5": { "acc": 0.7862068965517242, "acc_stderr": 0.034165204477475494, "acc_norm": 0.7862068965517242, "acc_norm_stderr": 0.034165204477475494 }, "harness|hendrycksTest-elementary_mathematics|5": { "acc": 0.6825396825396826, "acc_stderr": 0.023973861998992072, "acc_norm": 0.6825396825396826, "acc_norm_stderr