five

open-llm-leaderboard-old/details_Sao10K__BrainDerp

收藏
Hugging Face2023-10-24 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_Sao10K__BrainDerp
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Sao10K/BrainDerp时自动创建的,评估是在Open LLM Leaderboard上进行的。数据集包含64个配置,每个配置对应一个评估任务。数据集由2次运行创建,每次运行可以在每个配置中找到特定的分割,分割以运行的时间戳命名。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,并用于计算和显示Open LLM Leaderboard上的聚合指标。

该数据集是在评估模型Sao10K/BrainDerp时自动创建的,评估是在Open LLM Leaderboard上进行的。数据集包含64个配置,每个配置对应一个评估任务。数据集由2次运行创建,每次运行可以在每个配置中找到特定的分割,分割以运行的时间戳命名。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,并用于计算和显示Open LLM Leaderboard上的聚合指标。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

该数据集是在模型 Sao10K/BrainDerpOpen LLM Leaderboard 上的评估运行期间自动创建的。

数据集组成

  • 数据集包含 64 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,每次运行在每个配置中作为一个特定的分割存在,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_Sao10K__BrainDerp", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-24T00:03:05.483075 运行 的最新结果: python { "all": { "em": 0.031774328859060404, "em_stderr": 0.00179624735213122, "f1": 0.15481229026845572, "f1_stderr": 0.002609735138062246, "acc": 0.4205614858545465, "acc_stderr": 0.009806536501080007 }, "harness|drop|3": { "em": 0.031774328859060404, "em_stderr": 0.00179624735213122, "f1": 0.15481229026845572, "f1_stderr": 0.002609735138062246 }, "harness|gsm8k|5": { "acc": 0.08263836239575435, "acc_stderr": 0.00758408922014814 }, "harness|winogrande|5": { "acc": 0.7584846093133386, "acc_stderr": 0.012028983782011874 } }

配置详情

配置列表

  • harness_arc_challenge_25
  • harness_drop_3
  • harness_gsm8k_5
  • harness_hellaswag_10
  • harness_hendrycksTest_5
  • harness_hendrycksTest_abstract_algebra_5
  • harness_hendrycksTest_anatomy_5
  • harness_hendrycksTest_astronomy_5
  • harness_hendrycksTest_business_ethics_5
  • harness_hendrycksTest_clinical_knowledge_5
  • harness_hendrycksTest_college_biology_5
  • harness_hendrycksTest_college_chemistry_5
  • harness_hendrycksTest_college_computer_science_5
  • harness_hendrycksTest_college_mathematics_5
  • harness_hendrycksTest_college_medicine_5
  • harness_hendrycksTest_college_physics_5
  • harness_hendrycksTest_computer_security_5
  • harness_hendrycksTest_conceptual_physics_5
  • harness_hendrycksTest_econometrics_5
  • harness_hendrycksTest_electrical_engineering_5
  • harness_hendrycksTest_elementary_mathematics_5

数据文件路径

每个配置包含多个分割,每个分割对应一个特定的时间戳,路径如下:

  • harness_arc_challenge_25
    • 2023_10_04T06_59_16.770544: **/details_harness|arc:challenge|25_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|arc:challenge|25_2023-10-04T06-59-16.770544.parquet
  • harness_drop_3
    • 2023_10_24T00_03_05.483075: **/details_harness|drop|3_2023-10-24T00-03-05.483075.parquet
    • latest: **/details_harness|drop|3_2023-10-24T00-03-05.483075.parquet
  • harness_gsm8k_5
    • 2023_10_24T00_03_05.483075: **/details_harness|gsm8k|5_2023-10-24T00-03-05.483075.parquet
    • latest: **/details_harness|gsm8k|5_2023-10-24T00-03-05.483075.parquet
  • harness_hellaswag_10
    • 2023_10_04T06_59_16.770544: **/details_harness|hellaswag|10_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hellaswag|10_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_5
    • 2023_10_04T06_59_16.770544: 多个路径
    • latest: 多个路径
  • harness_hendrycksTest_abstract_algebra_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_anatomy_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-anatomy|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-anatomy|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_astronomy_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-astronomy|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-astronomy|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_business_ethics_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-business_ethics|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-business_ethics|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_clinical_knowledge_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_college_biology_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-college_biology|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-college_biology|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_college_chemistry_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_college_computer_science_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_college_mathematics_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_college_medicine_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-college_medicine|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-college_medicine|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_college_physics_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-college_physics|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-college_physics|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_computer_security_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-computer_security|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-computer_security|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_conceptual_physics_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_econometrics_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-econometrics|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-econometrics|5_2023-10-04T06-59-16.770544.parquet
  • harness_hendrycksTest_electrical_engineering_5
    • 2023_10_04T06_59_16.770544: **/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T06-59-16.770544.parquet
    • latest: **/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T06-59-16.770544.parquet
  • `harness_hendrycksTest_elementary_mathematics
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