five

open-llm-leaderboard-old/details_TheBloke__gpt4-x-vicuna-13B-HF

收藏
Hugging Face2023-08-27 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_TheBloke__gpt4-x-vicuna-13B-HF
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在Open LLM Leaderboard上对模型TheBloke/gpt4-x-vicuna-13B-HF进行评估时自动生成的。数据集包含61个配置,每个配置对应一个评估任务。它包含一次运行的结果,每次运行在每个配置中表示为特定的拆分,以运行的时间戳命名。train拆分始终指向最新结果。一个额外的results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。文件还提供了如何使用Python中的datasets库加载数据集的示例,并包含了特定运行的最新结果。

该数据集是在Open LLM Leaderboard上对模型TheBloke/gpt4-x-vicuna-13B-HF进行评估时自动生成的。数据集包含61个配置,每个配置对应一个评估任务。它包含一次运行的结果,每次运行在每个配置中表示为特定的拆分,以运行的时间戳命名。train拆分始终指向最新结果。一个额外的results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。文件还提供了如何使用Python中的datasets库加载数据集的示例,并包含了特定运行的最新结果。
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 TheBloke/gpt4-x-vicuna-13B-HF 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 61 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。每次运行可以在每个配置中找到特定的分割,分割名称使用运行的时间戳。
  • "train" 分割始终指向最新的结果。
  • 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_TheBloke__gpt4-x-vicuna-13B-HF", "harness_truthfulqa_mc_0", split="train")

最新结果

这些是最新结果(来自 2023-07-19T19:01:51.030763 运行)的示例: python { "all": { "acc": 0.5137597162733054, "acc_stderr": 0.03484317305077308, "acc_norm": 0.5174954549900392, "acc_norm_stderr": 0.03482742951911445, "mc1": 0.3635250917992656, "mc1_stderr": 0.016838862883965827, "mc2": 0.5357942440986606, "mc2_stderr": 0.015916184024373756 }, "harness|arc:challenge|25": { "acc": 0.5110921501706485, "acc_stderr": 0.01460779491401305, "acc_norm": 0.5341296928327645, "acc_norm_stderr": 0.014577311315231104 }, "harness|hellaswag|10": { "acc": 0.6038637721569409, "acc_stderr": 0.004880937933163287, "acc_norm": 0.8012348137821151, "acc_norm_stderr": 0.003982553164086259 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.31, "acc_stderr": 0.04648231987117316, "acc_norm": 0.31, "acc_norm_stderr": 0.04648231987117316 }, # 其他任务的结果... }

配置详情

  • config_name: harness_arc_challenge_25

    • split: 2023_07_19T19_01_51.030763
      • 路径: **/details_harness|arc:challenge|25_2023-07-19T19:01:51.030763.parquet
    • split: latest
      • 路径: **/details_harness|arc:challenge|25_2023-07-19T19:01:51.030763.parquet
  • config_name: harness_hellaswag_10

    • split: 2023_07_19T19_01_51.030763
      • 路径: **/details_harness|hellaswag|10_2023-07-19T19:01:51.030763.parquet
    • split: latest
      • 路径: **/details_harness|hellaswag|10_2023-07-19T19:01:51.030763.parquet
  • config_name: harness_hendrycksTest_5

    • split: 2023_07_19T19_01_51.030763
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-management|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-prehistory|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-professional_law|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-professional_medicine|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-professional_psychology|5_2023-07-19T19:01:51.030763.parquet
        • **/details_harness|hendrycksTest-public_relations|5_2023-07-19T19:01:51.030763.parquet
        • `**/details_harness|hendrycksTest-security_studies|5_2023-07-19T19:01:51
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