five

open-llm-leaderboard/details_meta-llama__Llama-2-13b-hf

收藏
Hugging Face2023-12-02 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_meta-llama__Llama-2-13b-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meta-llama/Llama-2-13b-hf模型的评估运行期间自动创建的,包含123个配置,每个配置对应一个评估任务。数据集由8次运行创建,每次运行都有特定的时间戳作为分割名称。此外,还有一个名为results的配置,用于存储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meta-llama/Llama-2-13b-hf模型的评估运行期间自动创建的,包含123个配置,每个配置对应一个评估任务。数据集由8次运行创建,每次运行都有特定的时间戳作为分割名称。此外,还有一个名为results的配置,用于存储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
原始信息汇总

数据集概述

数据集来源

数据集结构

  • 配置数量: 123 个配置,每个配置对应一个评估任务。
  • 创建次数: 数据集由 8 次运行创建,每次运行可以在每个配置中找到特定的分割,分割名称使用运行的时间戳。
  • 最新结果: "train" 分割始终指向最新的结果。
  • 汇总结果: 一个额外的配置 "results" 存储所有运行的汇总结果,用于计算和显示在 Open LLM Leaderboard 上的汇总指标。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_meta-llama__Llama-2-13b-hf", "harness_gsm8k_5", split="train")

最新结果示例

  • 最新结果: 以下是 2023-12-02T13:11:49.394544 运行的最新结果: python { "all": { "acc": 0.22820318423047764, "acc_stderr": 0.011559914877317397 }, "harness|gsm8k|5": { "acc": 0.22820318423047764, "acc_stderr": 0.011559914877317397 } }

配置详情

  • 配置名称: harness_arc_challenge_25

    • 分割: 2023_08_19T22_35_38.117975
      • 路径: **/details_harness|arc:challenge|25_2023-08-19T22:35:38.117975.parquet
    • 分割: 2023_08_23T17_28_00.015478
      • 路径: **/details_harness|arc:challenge|25_2023-08-23T17:28:00.015478.parquet
    • 分割: 2023_08_29T22_26_02.660247
      • 路径: **/details_harness|arc:challenge|25_2023-08-29T22:26:02.660247.parquet
    • 分割: latest
      • 路径: **/details_harness|arc:challenge|25_2023-08-29T22:26:02.660247.parquet
  • 配置名称: harness_drop_0

    • 分割: 2023_09_15T14_07_08.353318
      • 路径: **/details_harness|drop|0_2023-09-15T14-07-08.353318.parquet
    • 分割: latest
      • 路径: **/details_harness|drop|0_2023-09-15T14-07-08.353318.parquet
  • 配置名称: harness_drop_3

    • 分割: 2023_09_08T14_32_14.957248
      • 路径: **/details_harness|drop|3_2023-09-08T14-32-14.957248.parquet
    • 分割: 2023_10_14T23_00_26.644553
      • 路径: **/details_harness|drop|3_2023-10-14T23-00-26.644553.parquet
    • 分割: latest
      • 路径: **/details_harness|drop|3_2023-10-14T23-00-26.644553.parquet
  • 配置名称: harness_gsm8k_5

    • 分割: 2023_09_08T14_32_14.957248
      • 路径: **/details_harness|gsm8k|5_2023-09-08T14-32-14.957248.parquet
    • 分割: 2023_10_14T23_00_26.644553
      • 路径: **/details_harness|gsm8k|5_2023-10-14T23-00-26.644553.parquet
    • 分割: 2023_12_02T13_11_49.394544
      • 路径: **/details_harness|gsm8k|5_2023-12-02T13-11-49.394544.parquet
    • 分割: latest
      • 路径: **/details_harness|gsm8k|5_2023-12-02T13-11-49.394544.parquet
  • 配置名称: harness_hellaswag_10

    • 分割: 2023_08_19T22_35_38.117975
      • 路径: **/details_harness|hellaswag|10_2023-08-19T22:35:38.117975.parquet
    • 分割: 2023_08_23T17_28_00.015478
      • 路径: **/details_harness|hellaswag|10_2023-08-23T17:28:00.015478.parquet
    • 分割: 2023_08_29T22_26_02.660247
      • 路径: **/details_harness|hellaswag|10_2023-08-29T22:26:02.660247.parquet
    • 分割: latest
      • 路径: **/details_harness|hellaswag|10_2023-08-29T22:26:02.660247.parquet
  • 配置名称: harness_hendrycksTest_5

    • 分割: 2023_08_19T22_35_38.117975
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-management|5_2023-08-19T22:35:38.117975.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-08-19T22:35:38.1
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