five

open-llm-leaderboard-old/details_Changgil__K2S3-Mistral-7b-v1.50

收藏
Hugging Face2024-04-16 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_Changgil__K2S3-Mistral-7b-v1.50
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Changgil/K2S3-Mistral-7b-v1.50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。数据集从1次运行中创建,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,results配置存储了所有运行的聚合结果,并用于计算和显示在Open LLM Leaderboard上的聚合指标。

该数据集是在评估模型Changgil/K2S3-Mistral-7b-v1.50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。数据集从1次运行中创建,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,results配置存储了所有运行的聚合结果,并用于计算和显示在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

该数据集是在评估模型Changgil/K2S3-Mistral-7b-v1.50Open LLM Leaderboard上的运行过程中自动创建的。

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。
  • 数据集由1次运行创建,每个运行可以在每个配置中找到特定的拆分,拆分名称使用运行的时间戳。
  • "train"拆分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_Changgil__K2S3-Mistral-7b-v1.50", "harness_winogrande_5", split="train")

最新结果

这些是最新结果的示例: python { "all": { "acc": 0.6255788994477649, "acc_stderr": 0.03265442737189764, "acc_norm": 0.6278777292603872, "acc_norm_stderr": 0.03331125131060467, "mc1": 0.3463892288861689, "mc1_stderr": 0.01665699710912514, "mc2": 0.4999439345876719, "mc2_stderr": 0.014965961693768017 }, "harness|arc:challenge|25": { "acc": 0.5708191126279863, "acc_stderr": 0.014464085894870655, "acc_norm": 0.6109215017064846, "acc_norm_stderr": 0.014247309976045607 }, "harness|hellaswag|10": { "acc": 0.6271659032065325, "acc_stderr": 0.004825702533920412, "acc_norm": 0.8282214698267277, "acc_norm_stderr": 0.003764169746646177 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.31, "acc_stderr": 0.04648231987117316, "acc_norm": 0.31, "acc_norm_stderr": 0.04648231987117316 }, # 更多任务结果... }

配置详情

  • harness_arc_challenge_25

    • 拆分: 2024_04_16T06_17_54.308769
    • 路径: **/details_harness|arc:challenge|25_2024-04-16T06-17-54.308769.parquet
    • 拆分: latest
    • 路径: **/details_harness|arc:challenge|25_2024-04-16T06-17-54.308769.parquet
  • harness_gsm8k_5

    • 拆分: 2024_04_16T06_17_54.308769
    • 路径: **/details_harness|gsm8k|5_2024-04-16T06-17-54.308769.parquet
    • 拆分: latest
    • 路径: **/details_harness|gsm8k|5_2024-04-16T06-17-54.308769.parquet
  • harness_hellaswag_10

    • 拆分: 2024_04_16T06_17_54.308769
    • 路径: **/details_harness|hellaswag|10_2024-04-16T06-17-54.308769.parquet
    • 拆分: latest
    • 路径: **/details_harness|hellaswag|10_2024-04-16T06-17-54.308769.parquet
  • harness_hendrycksTest_5

    • 拆分: 2024_04_16T06_17_54.308769
    • 路径: **/details_harness|hendrycksTest-abstract_algebra|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-anatomy|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-astronomy|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-business_ethics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-clinical_knowledge|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-college_biology|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-college_chemistry|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-college_computer_science|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-college_mathematics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-college_medicine|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-college_physics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-computer_security|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-conceptual_physics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-econometrics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-electrical_engineering|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-elementary_mathematics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-formal_logic|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-global_facts|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_biology|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_chemistry|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_computer_science|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_european_history|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_geography|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_mathematics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_microeconomics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_physics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_psychology|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_statistics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_us_history|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-high_school_world_history|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-human_aging|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-human_sexuality|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-international_law|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-jurisprudence|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-logical_fallacies|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-machine_learning|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-management|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-marketing|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-medical_genetics|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-miscellaneous|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-moral_disputes|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-moral_scenarios|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-nutrition|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-philosophy|5_2024-04-16T06-17-54.308769.parquet
    • 路径: **/details_harness|hendrycksTest-prehistory|5_2024-04-16T06-17-54.308769.parquet
    • 路径: `**/details_harness|hendrycksTest-professional_accounting|5_2024-04-16T06-17-54.308
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