five

open-llm-leaderboard-old/details_M4-ai__TinyMistral-6x248M-Instruct

收藏
Hugging Face2024-02-02 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_M4-ai__TinyMistral-6x248M-Instruct
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M4-ai/TinyMistral-6x248M-Instruct时自动创建的,评估在Open LLM Leaderboard上进行。数据集由63个配置组成,每个配置对应一个评估任务。数据集由1次运行生成,每次运行的结果作为特定配置中的一个分割,分割名称使用运行的时间戳。train分割始终指向最新的结果。此外,results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M4-ai/TinyMistral-6x248M-Instruct时自动创建的,评估在Open LLM Leaderboard上进行。数据集由63个配置组成,每个配置对应一个评估任务。数据集由1次运行生成,每次运行的结果作为特定配置中的一个分割,分割名称使用运行的时间戳。train分割始终指向最新的结果。此外,results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

数据集简介

该数据集是在对模型M4-ai/TinyMistral-6x248M-Instruct进行评估运行期间自动创建的,用于Open LLM Leaderboard

数据集组成

  • 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1次运行中创建,每个运行可以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_M4-ai__TinyMistral-6x248M-Instruct", "harness_winogrande_5", split="train")

最新结果

以下是2024-02-02T00:43:37.562563运行的最新结果:

python { "all": { "acc": 0.24095884646371188, "acc_stderr": 0.03008991362961084, "acc_norm": 0.24148314901574003, "acc_norm_stderr": 0.03089156878834565, "mc1": 0.2460220318237454, "mc1_stderr": 0.015077219200662595, "mc2": 0.4316446283639332, "mc2_stderr": 0.015551246607562378 }, "harness|arc:challenge|25": { "acc": 0.17235494880546076, "acc_stderr": 0.011037113093461295, "acc_norm": 0.22440273037542663, "acc_norm_stderr": 0.012191404938603836 }, "harness|hellaswag|10": { "acc": 0.2672774347739494, "acc_stderr": 0.004416339450436124, "acc_norm": 0.27016530571599284, "acc_norm_stderr": 0.004431375549911368 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.22, "acc_stderr": 0.04163331998932268, "acc_norm": 0.22, "acc_norm_stderr": 0.04163331998932268 }, ... }

配置详情

  • harness_arc_challenge_25

    • 分割: 2024_02_02T00_43_37.562563
      • 路径: **/details_harness|arc:challenge|25_2024-02-02T00-43-37.562563.parquet
    • 分割: latest
      • 路径: **/details_harness|arc:challenge|25_2024-02-02T00-43-37.562563.parquet
  • harness_gsm8k_5

    • 分割: 2024_02_02T00_43_37.562563
      • 路径: **/details_harness|gsm8k|5_2024-02-02T00-43-37.562563.parquet
    • 分割: latest
      • 路径: **/details_harness|gsm8k|5_2024-02-02T00-43-37.562563.parquet
  • harness_hellaswag_10

    • 分割: 2024_02_02T00_43_37.562563
      • 路径: **/details_harness|hellaswag|10_2024-02-02T00-43-37.562563.parquet
    • 分割: latest
      • 路径: **/details_harness|hellaswag|10_2024-02-02T00-43-37.562563.parquet
  • harness_hendrycksTest_5

    • 分割: 2024_02_02T00_43_37.562563
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2024-02-02T00-43-37.562563.parquet
    • 分割: latest
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2024-02-02T00-43-37.562563.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2024-02-02T00-43-37.562563.parquet
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