five

open-llm-leaderboard-old/details_CHIH-HUNG__llama-2-13b-FINETUNE4_3.8w-r4-q_k_v_o_gate_up_down

收藏
Hugging Face2023-10-28 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-FINETUNE4_3.8w-r4-q_k_v_o_gate_up_down
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-FINETUNE4_3.8w-r4-q_k_v_o_gate_up_down]在Open LLM Leaderboard上的评估运行期间自动创建的。它由64个配置组成,每个配置对应一个评估任务。数据集由2次运行创建,每次运行作为每个配置中的一个特定分割,使用运行的时间戳命名。"train"分割始终指向最新结果。一个额外的配置"results"存储了运行的所有聚合结果,用于计算和显示Open LLM Leaderboard上的聚合指标。

该数据集是在模型[CHIH-HUNG/llama-2-13b-FINETUNE4_3.8w-r4-q_k_v_o_gate_up_down]在Open LLM Leaderboard上的评估运行期间自动创建的。它由64个配置组成,每个配置对应一个评估任务。数据集由2次运行创建,每次运行作为每个配置中的一个特定分割,使用运行的时间戳命名。"train"分割始终指向最新结果。一个额外的配置"results"存储了运行的所有聚合结果,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

该数据集是在对模型 CHIH-HUNG/llama-2-13b-FINETUNE4_3.8w-r4-q_k_v_o_gate_up_down 进行评估运行期间自动创建的,用于 Open LLM Leaderboard 上的评估。

数据集结构

  • 配置数量:64个配置,每个配置对应一个评估任务。
  • 数据来源:数据集由2次运行结果组成,每次运行的结果可以在每个配置中找到,以运行的时间戳命名的特定分片形式存在。
  • 最新结果:每个配置的 "train" 分片总是指向最新的结果。
  • 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-FINETUNE4_3.8w-r4-q_k_v_o_gate_up_down", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-28T03:26:24.700404 运行 的最新结果:

python { "all": { "em": 0.3543414429530201, "em_stderr": 0.004898375582961981, "f1": 0.39350461409396, "f1_stderr": 0.004814089237052805, "acc": 0.4362902583993399, "acc_stderr": 0.01014740888692861 }, "harness|drop|3": { "em": 0.3543414429530201, "em_stderr": 0.004898375582961981, "f1": 0.39350461409396, "f1_stderr": 0.004814089237052805 }, "harness|gsm8k|5": { "acc": 0.10462471569370735, "acc_stderr": 0.008430668082029278 }, "harness|winogrande|5": { "acc": 0.7679558011049724, "acc_stderr": 0.011864149691827943 } }

配置详情

以下是数据集的配置详情:

  • harness_arc_challenge_25

    • 分片:2023_10_04T05_10_57.019261
      • 路径:**/details_harness|arc:challenge|25_2023-10-04T05-10-57.019261.parquet
    • 分片:latest
      • 路径:**/details_harness|arc:challenge|25_2023-10-04T05-10-57.019261.parquet
  • harness_drop_3

    • 分片:2023_10_28T03_26_24.700404
      • 路径:**/details_harness|drop|3_2023-10-28T03-26-24.700404.parquet
    • 分片:latest
      • 路径:**/details_harness|drop|3_2023-10-28T03-26-24.700404.parquet
  • harness_gsm8k_5

    • 分片:2023_10_28T03_26_24.700404
      • 路径:**/details_harness|gsm8k|5_2023-10-28T03-26-24.700404.parquet
    • 分片:latest
      • 路径:**/details_harness|gsm8k|5_2023-10-28T03-26-24.700404.parquet
  • harness_hellaswag_10

    • 分片:2023_10_04T05_10_57.019261
      • 路径:**/details_harness|hellaswag|10_2023-10-04T05-10-57.019261.parquet
    • 分片:latest
      • 路径:**/details_harness|hellaswag|10_2023-10-04T05-10-57.019261.parquet
  • harness_hendrycksTest_5

    • 分片:2023_10_04T05_10_57.019261
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-management|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2023-10-04T05-10-57.019261.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2023-10-04T05-10-57.019261.parquet
        • `**/details_harness|hendrycksTest-prehistory|5_2023-10-
二维码
社区交流群
二维码
科研交流群
商业服务