five

open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-huangyt_FINETUNE2_3w-q_k_v_o_proj

收藏
Hugging Face2023-10-18 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_CHIH-HUNG__llama-2-13b-huangyt_FINETUNE2_3w-q_k_v_o_proj
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CHIH-HUNG/llama-2-13b-huangyt_FINETUNE2_3w-q_k_v_o_proj时自动创建的,包含64个配置,每个配置对应一个评估任务。数据集由2次运行生成,每次运行的结果存储为特定的分割,分割名称使用运行的时间戳。此外,数据集包含一个名为"results"的配置,用于存储所有运行的聚合结果。
提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

该数据集是在评估模型 CHIH-HUNG/llama-2-13b-huangyt_FINETUNE2_3w-q_k_v_o_projOpen LLM Leaderboard 上的运行过程中自动创建的。

数据集组成

  • 数据集包含 64 个配置,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。
  • 一个额外的配置 "results" 存储所有运行的聚合结果,用于计算和显示在 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-huangyt_FINETUNE2_3w-q_k_v_o_proj", "harness_winogrande_5", split="train")

最新结果

这些是最新的结果,来自 2023-10-18T02:00:22.389736 的运行: python { "all": { "em": 0.3719588926174497, "em_stderr": 0.004949726013193945, "f1": 0.4084679110738261, "f1_stderr": 0.004843145937750956, "acc": 0.4480415851620389, "acc_stderr": 0.010535274120903989 }, "harness|drop|3": { "em": 0.3719588926174497, "em_stderr": 0.004949726013193945, "f1": 0.4084679110738261, "f1_stderr": 0.004843145937750956 }, "harness|gsm8k|5": { "acc": 0.1281273692191054, "acc_stderr": 0.009206398549980031 }, "harness|winogrande|5": { "acc": 0.7679558011049724, "acc_stderr": 0.011864149691827948 } }

配置详情

以下是数据集的配置详情:

  • config_name: harness_arc_challenge_25

    • split: 2023_09_02T11_14_02.105897
      • path: **/details_harness|arc:challenge|25_2023-09-02T11:14:02.105897.parquet
    • split: latest
      • path: **/details_harness|arc:challenge|25_2023-09-02T11:14:02.105897.parquet
  • config_name: harness_drop_3

    • split: 2023_10_18T02_00_22.389736
      • path: **/details_harness|drop|3_2023-10-18T02-00-22.389736.parquet
    • split: latest
      • path: **/details_harness|drop|3_2023-10-18T02-00-22.389736.parquet
  • config_name: harness_gsm8k_5

    • split: 2023_10_18T02_00_22.389736
      • path: **/details_harness|gsm8k|5_2023-10-18T02-00-22.389736.parquet
    • split: latest
      • path: **/details_harness|gsm8k|5_2023-10-18T02-00-22.389736.parquet
  • config_name: harness_hellaswag_10

    • split: 2023_09_02T11_14_02.105897
      • path: **/details_harness|hellaswag|10_2023-09-02T11:14:02.105897.parquet
    • split: latest
      • path: **/details_harness|hellaswag|10_2023-09-02T11:14:02.105897.parquet
  • config_name: harness_hendrycksTest_5

    • split: 2023_09_02T11_14_02.105897
      • path:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-management|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-prehistory|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-professional_accounting|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-professional_law|5_2023-09-02T11:14:02.105897.parquet
        • **/details_harness|hendrycksTest-professional_medicine|5
二维码
社区交流群
二维码
科研交流群
商业服务