five

open-llm-leaderboard-old/details_CHIH-HUNG__llama-2-13b-FINETUNE3_3.3w-r4-q_k_v_o

收藏
Hugging Face2023-10-29 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_CHIH-HUNG__llama-2-13b-FINETUNE3_3.3w-r4-q_k_v_o
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在模型CHIH-HUNG/llama-2-13b-FINETUNE3_3.3w-r4-q_k_v_o在Open LLM Leaderboard上的评估运行期间自动创建的。数据集由64个配置组成,每个配置对应一个被评估的任务。数据集是从2次运行中创建的,每次运行在每个配置中表示为特定的分割。train分割始终指向最新的结果。一个名为results的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示Open LLM Leaderboard上的聚合指标。README还提供了一个如何使用Python中的datasets库加载运行细节的示例。

该数据集是在模型CHIH-HUNG/llama-2-13b-FINETUNE3_3.3w-r4-q_k_v_o在Open LLM Leaderboard上的评估运行期间自动创建的。数据集由64个配置组成,每个配置对应一个被评估的任务。数据集是从2次运行中创建的,每次运行在每个配置中表示为特定的分割。train分割始终指向最新的结果。一个名为results的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示Open LLM Leaderboard上的聚合指标。README还提供了一个如何使用Python中的datasets库加载运行细节的示例。
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CHIH-HUNG/llama-2-13b-FINETUNE3_3.3w-r4-q_k_v_oOpen LLM Leaderboard上的运行过程中自动创建的。

数据集结构

  • 数据集由64个配置组成,每个配置对应一个评估任务。
  • 数据集从2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_CHIH-HUNG__llama-2-13b-FINETUNE3_3.3w-r4-q_k_v_o", "harness_winogrande_5", split="train")

最新结果

以下是2023-10-29T06:18:40.870535运行的最新结果: python { "all": { "em": 0.3716442953020134, "em_stderr": 0.004948871392755147, "f1": 0.4110413171140945, "f1_stderr": 0.004850802412390029, "acc": 0.4419141525144315, "acc_stderr": 0.010419855537406255 }, "harness|drop|3": { "em": 0.3716442953020134, "em_stderr": 0.004948871392755147, "f1": 0.4110413171140945, "f1_stderr": 0.004850802412390029 }, "harness|gsm8k|5": { "acc": 0.11902956785443518, "acc_stderr": 0.008919702911161637 }, "harness|winogrande|5": { "acc": 0.7647987371744278, "acc_stderr": 0.011920008163650873 } }

配置详情

以下是数据集的配置详情:

配置列表

  • harness_arc_challenge_25

    • 分割: 2023_10_01T14_24_56.870950
      • 路径: **/details_harness|arc:challenge|25_2023-10-01T14-24-56.870950.parquet
    • 分割: latest
      • 路径: **/details_harness|arc:challenge|25_2023-10-01T14-24-56.870950.parquet
  • harness_drop_3

    • 分割: 2023_10_29T06_18_40.870535
      • 路径: **/details_harness|drop|3_2023-10-29T06-18-40.870535.parquet
    • 分割: latest
      • 路径: **/details_harness|drop|3_2023-10-29T06-18-40.870535.parquet
  • harness_gsm8k_5

    • 分割: 2023_10_29T06_18_40.870535
      • 路径: **/details_harness|gsm8k|5_2023-10-29T06-18-40.870535.parquet
    • 分割: latest
      • 路径: **/details_harness|gsm8k|5_2023-10-29T06-18-40.870535.parquet
  • harness_hellaswag_10

    • 分割: 2023_10_01T14_24_56.870950
      • 路径: **/details_harness|hellaswag|10_2023-10-01T14-24-56.870950.parquet
    • 分割: latest
      • 路径: **/details_harness|hellaswag|10_2023-10-01T14-24-56.870950.parquet
  • harness_hendrycksTest_5

    • 分割: 2023_10_01T14_24_56.870950
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-management|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2023-10-01T14-24-56.870950.parquet
        • **/details_harness|hendrycksTest-prehistory|5_2023-10-01T14-24-56.870950.parquet
        • `**/details_harness|hendrycksTest-professional_accounting|5_2023-10
二维码
社区交流群
二维码
科研交流群
商业服务