five

open-llm-leaderboard/details_chargoddard__llama2-22b-blocktriangular

收藏
Hugging Face2023-10-18 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_chargoddard__llama2-22b-blocktriangular
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在评估模型chargoddard/llama2-22b-blocktriangular时自动创建的,用于Open LLM Leaderboard的评估。数据集包含64个配置,每个配置对应一个评估任务。数据集由3次运行生成,每次运行的结果作为特定的分割存储在配置中,分割名称使用运行的时间戳。train分割始终指向最新的结果。此外,还有一个名为results的配置,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。
提供机构:
open-llm-leaderboard
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 chargoddard/llama2-22b-blocktriangularOpen LLM Leaderboard 上的自动创建的。数据集包含 64 个配置,每个配置对应一个评估任务。

数据集结构

数据集由 3 次运行结果组成,每次运行的结果可以在每个配置中找到,并以运行的时间戳命名。"train" 分割始终指向最新的结果。

额外配置

有一个额外的配置 "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_chargoddard__llama2-22b-blocktriangular", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-18T10:02:48.850156 运行的最新结果: python { "all": { "em": 0.002202181208053691, "em_stderr": 0.000480051081661935, "f1": 0.06165897651006692, "f1_stderr": 0.0013848407345463738, "acc": 0.4357400460634537, "acc_stderr": 0.010354651175233286 }, "harness|drop|3": { "em": 0.002202181208053691, "em_stderr": 0.000480051081661935, "f1": 0.06165897651006692, "f1_stderr": 0.0013848407345463738 }, "harness|gsm8k|5": { "acc": 0.11220621683093253, "acc_stderr": 0.008693743138242383 }, "harness|winogrande|5": { "acc": 0.7592738752959748, "acc_stderr": 0.012015559212224186 } }

配置详情

以下是数据集的配置详情:

配置 harness_arc_challenge_25

  • 分割 2023_08_17T16_15_19.075132
    • 路径: **/details_harness|arc:challenge|25_2023-08-17T16:15:19.075132.parquet
  • 分割 latest
    • 路径: **/details_harness|arc:challenge|25_2023-08-17T16:15:19.075132.parquet

配置 harness_drop_3

  • 分割 2023_10_18T05_53_32.762527
    • 路径: **/details_harness|drop|3_2023-10-18T05-53-32.762527.parquet
  • 分割 2023_10_18T10_02_48.850156
    • 路径: **/details_harness|drop|3_2023-10-18T10-02-48.850156.parquet
  • 分割 latest
    • 路径: **/details_harness|drop|3_2023-10-18T10-02-48.850156.parquet

配置 harness_gsm8k_5

  • 分割 2023_10_18T05_53_32.762527
    • 路径: **/details_harness|gsm8k|5_2023-10-18T05-53-32.762527.parquet
  • 分割 2023_10_18T10_02_48.850156
    • 路径: **/details_harness|gsm8k|5_2023-10-18T10-02-48.850156.parquet
  • 分割 latest
    • 路径: **/details_harness|gsm8k|5_2023-10-18T10-02-48.850156.parquet

配置 harness_hellaswag_10

  • 分割 2023_08_17T16_15_19.075132
    • 路径: **/details_harness|hellaswag|10_2023-08-17T16:15:19.075132.parquet
  • 分割 latest
    • 路径: **/details_harness|hellaswag|10_2023-08-17T16:15:19.075132.parquet

配置 harness_hendrycksTest_5

  • 分割 2023_08_17T16_15_19.075132
    • 路径:
      • **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-anatomy|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-astronomy|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-business_ethics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-college_biology|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-college_chemistry|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-college_computer_science|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-college_mathematics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-college_medicine|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-college_physics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-computer_security|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-econometrics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-formal_logic|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-global_facts|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_biology|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_geography|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_physics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-human_aging|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-human_sexuality|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-international_law|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-jurisprudence|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-machine_learning|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-management|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-marketing|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-medical_genetics|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-miscellaneous|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-moral_disputes|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-nutrition|5_2023-08-17T16:15:19.075132.parquet
      • **/details_harness|hendrycksTest-philosophy|5_2023-08-17T16:15:19.075132.parquet
      • `**/details_harness|hend
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