five

open-llm-leaderboard-old/details_dfurman__HermesBagel-34B-v0.1

收藏
Hugging Face2024-01-14 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_dfurman__HermesBagel-34B-v0.1
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dfurman/HermesBagel-34B-v0.1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的特定分割中找到,分割以运行的时间戳命名。train分割始终指向最新结果。此外,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。

该数据集是在评估模型dfurman/HermesBagel-34B-v0.1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的特定分割中找到,分割以运行的时间戳命名。train分割始终指向最新结果。此外,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-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 dfurman/HermesBagel-34B-v0.1 的过程中自动创建的,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。

额外配置

一个额外的配置 "results" 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。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_dfurman__HermesBagel-34B-v0.1", "harness_winogrande_5", split="train")

最新结果

以下是 2024-01-14T03:53:56.861170 运行的最新结果

python { "all": { "acc": 0.7695763625614322, "acc_stderr": 0.02793431209028075, "acc_norm": 0.7740465788313311, "acc_norm_stderr": 0.028460203996252778, "mc1": 0.5006119951040392, "mc1_stderr": 0.01750348793889251, "mc2": 0.673352473186811, "mc2_stderr": 0.014617965588559495 }, "harness|arc:challenge|25": { "acc": 0.6757679180887372, "acc_stderr": 0.013678810399518822, "acc_norm": 0.7056313993174061, "acc_norm_stderr": 0.01331852846053942 }, "harness|hellaswag|10": { "acc": 0.6638119896434973, "acc_stderr": 0.004714386376337134, "acc_norm": 0.8573989245170285, "acc_norm_stderr": 0.003489509493001622 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.5, "acc_stderr": 0.050251890762960605, "acc_norm": 0.5, "acc_norm_stderr": 0.050251890762960605 }, ... }

配置详情

  • config_name: harness_arc_challenge_25

    • data_files:
      • split: 2024_01_14T03_53_56.861170
        • path: **/details_harness|arc:challenge|25_2024-01-14T03-53-56.861170.parquet
      • split: latest
        • path: **/details_harness|arc:challenge|25_2024-01-14T03-53-56.861170.parquet
  • config_name: harness_gsm8k_5

    • data_files:
      • split: 2024_01_14T03_53_56.861170
        • path: **/details_harness|gsm8k|5_2024-01-14T03-53-56.861170.parquet
      • split: latest
        • path: **/details_harness|gsm8k|5_2024-01-14T03-53-56.861170.parquet
  • config_name: harness_hellaswag_10

    • data_files:
      • split: 2024_01_14T03_53_56.861170
        • path: **/details_harness|hellaswag|10_2024-01-14T03-53-56.861170.parquet
      • split: latest
        • path: **/details_harness|hellaswag|10_2024-01-14T03-53-56.861170.parquet
  • config_name: harness_hendrycksTest_5

    • data_files:
      • split: 2024_01_14T03_53_56.861170
        • path:
          • **/details_harness|hendrycksTest-abstract_algebra|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-anatomy|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-astronomy|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-business_ethics|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-college_biology|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-college_chemistry|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-college_computer_science|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-college_mathematics|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-college_medicine|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-college_physics|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-computer_security|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-conceptual_physics|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-econometrics|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-electrical_engineering|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-formal_logic|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-global_facts|5_2024-01-14T03-53-56.861170.parquet
          • **/details_harness|hendrycksTest-high_school_biology|5_2024-01-14T03-53-56.861170.parquet
          • ... }
二维码
社区交流群
二维码
科研交流群
商业服务