open-llm-leaderboard-old/details_TheBloke__Genz-70b-GPTQ

该数据集是在模型TheBloke/Genz-70b-GPTQ在Open LLM Leaderboard上的评估运行期间自动创建的。数据集由61个配置组成，每个配置对应一个评估任务。数据集是从1次运行中创建的，每次运行在每个配置中作为一个特定的分割找到，分割名称使用运行的时间戳。train分割始终指向最新的结果。一个额外的配置results存储了运行的所有聚合结果，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Python代码加载运行中的详细信息的示例，并包含了特定运行的最新结果。

原始信息汇总

数据集概述

数据集简介

该数据集是在对模型 TheBloke/Genz-70b-GPTQ 进行评估运行期间自动创建的，用于 Open LLM Leaderboard。

数据集结构

• 配置数量：61个配置，每个配置对应一个评估任务。
• 创建来源：从1次运行中创建。每个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。
• 额外配置："results" 配置存储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。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_TheBloke__Genz-70b-GPTQ", "harness_truthfulqa_mc_0", split="train")

最新结果

这些是最新结果（来自2023-08-31T00:30:34.342002运行）的示例： python { "all": { "acc": 0.7017249416277331, "acc_stderr": 0.030832772804323012, "acc_norm": 0.70569345061239, "acc_norm_stderr": 0.03080075128019408, "mc1": 0.4320685434516524, "mc1_stderr": 0.01734120239498826, "mc2": 0.6228267270427654, "mc2_stderr": 0.014836432877772263 }, "harness|arc:challenge|25": { "acc": 0.6638225255972696, "acc_stderr": 0.013804855026205763, "acc_norm": 0.7107508532423208, "acc_norm_stderr": 0.013250012579393443 }, "harness|hellaswag|10": { "acc": 0.689205337582155, "acc_stderr": 0.004618730353217047, "acc_norm": 0.8764190400318662, "acc_norm_stderr": 0.0032843028764223 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.34, "acc_stderr": 0.04760952285695235, "acc_norm": 0.34, "acc_norm_stderr": 0.04760952285695235 }, # 更多任务结果... }

配置详情

• 配置名称：harness_arc_challenge_25

  • 数据文件：
    • 分割：2023_08_31T00_30_34.342002
      • 路径：**/details_harness|arc:challenge|25_2023-08-31T00:30:34.342002.parquet
    • 分割：latest
      • 路径：**/details_harness|arc:challenge|25_2023-08-31T00:30:34.342002.parquet

• 配置名称：harness_hellaswag_10

  • 数据文件：
    • 分割：2023_08_31T00_30_34.342002
      • 路径：**/details_harness|hellaswag|10_2023-08-31T00:30:34.342002.parquet
    • 分割：latest
      • 路径：**/details_harness|hellaswag|10_2023-08-31T00:30:34.342002.parquet

• 配置名称：harness_hendrycksTest_5

  • 数据文件：
    • 分割：2023_08_31T00_30_34.342002
      • 路径：
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-management|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-prehistory|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-professional_accounting|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-professional_law|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-professional_medicine|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-professional_psychology|5_2023-08-31T00:30:34.342002.parquet
        • **/details_harness|hendrycksTest-public_relations|5_2023-08-31T00:30:34.342002.parquet
        • `**/details_harness|hendrycksTest-security_studies|5_2023-08-31T00:30:34.342