five

open-llm-leaderboard/details_AtAndDev__ShortKing-3b-v0.3

收藏
Hugging Face2023-10-29 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_AtAndDev__ShortKing-3b-v0.3
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AtAndDev/ShortKing-3b-v0.3的评估运行过程中自动创建的,用于Open LLM Leaderboard的评估。数据集由64个配置组成,每个配置对应一个评估任务。数据集由2次运行生成,每次运行的结果存储在不同的分割中,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,并用于计算和显示Open LLM Leaderboard上的聚合指标。

This dataset was automatically created during the evaluation runs of the AtAndDev/ShortKing-3b-v0.3 model, and is intended for evaluations on the Open LLM Leaderboard. It consists of 64 configurations, each corresponding to one individual evaluation task. The dataset is generated from two separate runs, with the results of each run stored in a distinct dataset split, where the split name is the timestamp of the corresponding run. The `train` split always points to the most recent results. Additionally, there is a configuration named `results` that stores the aggregated results from all runs, and is used to calculate and display the aggregated metrics on the Open LLM Leaderboard.
提供机构:
open-llm-leaderboard
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 AtAndDev/ShortKing-3b-v0.3 进行评估运行期间自动创建的,用于 Open LLM Leaderboard

数据集结构

  • 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 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_AtAndDev__ShortKing-3b-v0.3", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-29T09:17:13.395928 运行 的最新结果: python { "all": { "em": 0.0017827181208053692, "em_stderr": 0.00043200973460391266, "f1": 0.05457843959731554, "f1_stderr": 0.001344563821795035, "acc": 0.34071397754750704, "acc_stderr": 0.008118865064946825 }, "harness|drop|3": { "em": 0.0017827181208053692, "em_stderr": 0.00043200973460391266, "f1": 0.05457843959731554, "f1_stderr": 0.001344563821795035 }, "harness|gsm8k|5": { "acc": 0.012130401819560273, "acc_stderr": 0.0030152942428909517 }, "harness|winogrande|5": { "acc": 0.6692975532754538, "acc_stderr": 0.013222435887002698 } }

配置详情

以下是数据集的部分配置详情:

  • harness_arc_challenge_25

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|arc:challenge|25_2023-10-04T03-04-04.830920.parquet
  • harness_drop_3

    • 分割:2023_10_29T09_17_13.395928, latest
    • 路径:**/details_harness|drop|3_2023-10-29T09-17-13.395928.parquet
  • harness_gsm8k_5

    • 分割:2023_10_29T09_17_13.395928, latest
    • 路径:**/details_harness|gsm8k|5_2023-10-29T09-17-13.395928.parquet
  • harness_hellaswag_10

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hellaswag|10_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:多个路径,例如 **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_abstract_algebra_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_anatomy_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-anatomy|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_astronomy_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-astronomy|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_business_ethics_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-business_ethics|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_clinical_knowledge_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_college_biology_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-college_biology|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_college_chemistry_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_college_computer_science_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_college_mathematics_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_college_medicine_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-college_medicine|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_college_physics_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-college_physics|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_computer_security_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-computer_security|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_conceptual_physics_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_econometrics_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-econometrics|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_electrical_engineering_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T03-04-04.830920.parquet
  • harness_hendrycksTest_elementary_mathematics_5

    • 分割:2023_10_04T03_04_04.830920, latest
    • 路径:**/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-04T03-04-04.830920.parquet
二维码
社区交流群
二维码
科研交流群
商业服务