five

open-llm-leaderboard-old/details_yeontaek__llama-2-13B-ensemble-v3

收藏
Hugging Face2023-08-27 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_yeontaek__llama-2-13B-ensemble-v3
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在Open LLM Leaderboard上对模型yeontaek/llama-2-13B-ensemble-v3进行评估时自动生成的。数据集包含61个配置,每个配置对应一个评估任务。它包含一次或多次运行的结果,每次运行都作为一个特定的分割存储,分割名称由运行的时间戳命名。train分割始终指向最新的结果。一个名为results的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示Open LLM Leaderboard上的指标。README还提供了一个Python代码片段来加载数据集,并列出了特定运行的最新结果。

该数据集是在Open LLM Leaderboard上对模型yeontaek/llama-2-13B-ensemble-v3进行评估时自动生成的。数据集包含61个配置,每个配置对应一个评估任务。它包含一次或多次运行的结果,每次运行都作为一个特定的分割存储,分割名称由运行的时间戳命名。train分割始终指向最新的结果。一个名为results的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示Open LLM Leaderboard上的指标。README还提供了一个Python代码片段来加载数据集,并列出了特定运行的最新结果。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

数据集简介

该数据集是在评估模型yeontaek/llama-2-13B-ensemble-v3Open LLM Leaderboard上的运行过程中自动创建的。

数据集组成

  • 数据集包含61个配置,每个配置对应一个评估任务。
  • 数据集由1次运行创建,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。
  • 额外的配置"results"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_yeontaek__llama-2-13B-ensemble-v3", "harness_truthfulqa_mc_0", split="train")

最新结果

以下是最新结果(来自2023-08-26T00:04:59.687493的运行):

python { "all": { "acc": 0.5779190308106052, "acc_stderr": 0.03411621449047892, "acc_norm": 0.5817150329939268, "acc_norm_stderr": 0.03409598894931763, "mc1": 0.35495716034271724, "mc1_stderr": 0.0167508623813759, "mc2": 0.49782296764839973, "mc2_stderr": 0.015206569782538341 }, "harness|arc:challenge|25": { "acc": 0.5981228668941979, "acc_stderr": 0.014327268614578274, "acc_norm": 0.6237201365187713, "acc_norm_stderr": 0.014157022555407161 }, "harness|hellaswag|10": { "acc": 0.6245767775343557, "acc_stderr": 0.004832423630593182, "acc_norm": 0.8229436367257519, "acc_norm_stderr": 0.0038093627612481094 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.33, "acc_stderr": 0.04725815626252606, "acc_norm": 0.33, "acc_norm_stderr": 0.04725815626252606 }, ... }

配置详情

  • config_name: harness_arc_challenge_25

    • 数据文件:
      • 分割: 2023_08_26T00_04_59.687493
        • 路径: **/details_harness|arc:challenge|25_2023-08-26T00:04:59.687493.parquet
      • 分割: latest
        • 路径: **/details_harness|arc:challenge|25_2023-08-26T00:04:59.687493.parquet
  • config_name: harness_hellaswag_10

    • 数据文件:
      • 分割: 2023_08_26T00_04_59.687493
        • 路径: **/details_harness|hellaswag|10_2023-08-26T00:04:59.687493.parquet
      • 分割: latest
        • 路径: **/details_harness|hellaswag|10_2023-08-26T00:04:59.687493.parquet
  • config_name: harness_hendrycksTest_5

    • 数据文件:
      • 分割: 2023_08_26T00_04_59.687493
        • 路径:
          • **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-anatomy|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-astronomy|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-business_ethics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-college_biology|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-college_chemistry|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-college_computer_science|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-college_mathematics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-college_medicine|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-college_physics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-computer_security|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-econometrics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-formal_logic|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-global_facts|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_biology|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_geography|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_physics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-human_aging|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-human_sexuality|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-international_law|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-jurisprudence|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-machine_learning|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-management|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-marketing|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-medical_genetics|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-miscellaneous|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-moral_disputes|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-nutrition|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-philosophy|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-prehistory|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-professional_accounting|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-professional_law|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-professional_medicine|5_2023-08-26T00:04:59.687493.parquet
          • **/details_harness|hendrycksTest-professional_psychology|5_2023-08-26T00:04:59.687493.parquet
          • `**/details_harness
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作