five

open-llm-leaderboard-old/details_mlabonne__NeuralPipe-7B-slerp

收藏
Hugging Face2024-01-05 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_mlabonne__NeuralPipe-7B-slerp
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mlabonne/NeuralPipe-7B-slerp时自动生成的,评估过程在Open LLM Leaderboard上进行。数据集包含63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,每次运行的结果作为一个特定的分割存储在配置中,分割名称使用运行的时间戳。train分割始终指向最新的结果。此外,还有一个名为results的配置,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。

该数据集是在评估模型mlabonne/NeuralPipe-7B-slerp时自动生成的,评估过程在Open LLM Leaderboard上进行。数据集包含63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,每次运行的结果作为一个特定的分割存储在配置中,分割名称使用运行的时间戳。train分割始终指向最新的结果。此外,还有一个名为results的配置,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。
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mlabonne/NeuralPipe-7B-slerpOpen LLM Leaderboard上的运行过程中自动创建的。

数据集结构

  • 配置数量:63个配置,每个配置对应一个评估任务。
  • 运行次数:数据集从2次运行中创建。每个运行在每个配置中作为一个特定的分割存在,分割名称使用运行的时间戳。
  • 训练分割:"train"分割始终指向最新的结果。
  • 结果配置: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_mlabonne__NeuralPipe-7B-slerp", "harness_winogrande_5", split="train")

最新结果

以下是2024-01-05T12:33:01.505276运行的最新结果:

python { "all": { "acc": 0.6445269708058093, "acc_stderr": 0.03218714474134609, "acc_norm": 0.6449418405596148, "acc_norm_stderr": 0.03284511879516387, "mc1": 0.4283965728274174, "mc1_stderr": 0.017323088597314754, "mc2": 0.598408044881861, "mc2_stderr": 0.015149948573522944 }, "harness|arc:challenge|25": { "acc": 0.6476109215017065, "acc_stderr": 0.013960142600598675, "acc_norm": 0.6757679180887372, "acc_norm_stderr": 0.013678810399518829 }, "harness|hellaswag|10": { "acc": 0.6701852220673172, "acc_stderr": 0.0046918486653990685, "acc_norm": 0.8616809400517825, "acc_norm_stderr": 0.003445289925011734 }, # 其他任务结果省略... }

配置详情

  • harness_arc_challenge_25

    • 分割:2023_12_29T17_44_55.770154, 2024_01_05T12_33_01.505276, latest
    • 路径:**/details_harness|arc:challenge|25_2023-12-29T17-44-55.770154.parquet, **/details_harness|arc:challenge|25_2024-01-05T12-33-01.505276.parquet
  • harness_gsm8k_5

    • 分割:2023_12_29T17_44_55.770154, 2024_01_05T12_33_01.505276, latest
    • 路径:**/details_harness|gsm8k|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|gsm8k|5_2024-01-05T12-33-01.505276.parquet
  • harness_hellaswag_10

    • 分割:2023_12_29T17_44_55.770154, 2024_01_05T12_33_01.505276, latest
    • 路径:**/details_harness|hellaswag|10_2023-12-29T17-44-55.770154.parquet, **/details_harness|hellaswag|10_2024-01-05T12-33-01.505276.parquet
  • harness_hendrycksTest_5

    • 分割:2023_12_29T17_44_55.770154, 2024_01_05T12_33_01.505276, latest
    • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-anatomy|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-astronomy|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-business_ethics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-clinical_knowledge|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-college_biology|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-college_chemistry|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-college_computer_science|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-college_mathematics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-college_medicine|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-college_physics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-computer_security|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-conceptual_physics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-econometrics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-electrical_engineering|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-elementary_mathematics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-formal_logic|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-global_facts|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_biology|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_chemistry|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_computer_science|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_european_history|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_geography|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_mathematics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_physics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_psychology|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_statistics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_us_history|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-high_school_world_history|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-human_aging|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-human_sexuality|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-international_law|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-jurisprudence|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-logical_fallacies|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-machine_learning|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-management|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-marketing|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-medical_genetics|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-miscellaneous|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-moral_disputes|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-moral_scenarios|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-nutrition|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-philosophy|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-prehistory|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-professional_accounting|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-professional_law|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-professional_medicine|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-professional_psychology|5_2023-12-29T17-44-55.770154.parquet, **/details_harness|hendrycksTest-public_relations|5_2023-12-29T17-44-55.770154.parquet, `**/details_harness|hendrycksTest-security_studies|5_
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