five

open-llm-leaderboard/details_harborwater__open-llama-3b-everything-v2

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_harborwater__open-llama-3b-everything-v2
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 harborwater/open-llama-3b-everything-v2 在 Open LLM Leaderboard 上进行评估时自动生成的。数据集包含 64 个配置,每个配置对应一个被评估的任务。数据集由 2 次运行的数据组成,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。此外,还有一个 results 配置,用于存储聚合结果,这些结果用于计算和显示在排行榜上的指标。README 还提供了如何使用 Hugging Face datasets 库加载数据集的示例。

This dataset was automatically generated when the model harborwater/open-llama-3b-everything-v2 was evaluated on the Open LLM Leaderboard. The dataset contains 64 configurations, each corresponding to one evaluated task. The dataset consists of data from two runs, where each run is represented as a specific split under each configuration. The "train" split always points to the most recent results. Additionally, there is a "results" configuration for storing aggregated results, which are used to calculate and display the metrics shown on the leaderboard. The README also provides examples of how to load the dataset using the Hugging Face Datasets library.
提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

数据集简介

该数据集是在评估模型 harborwater/open-llama-3b-everything-v2Open LLM Leaderboard 上的运行过程中自动创建的。

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。
  • 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_harborwater__open-llama-3b-everything-v2", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-29T00:43:57.732775 运行的最新结果: python { "all": { "em": 0.0020973154362416107, "em_stderr": 0.0004685065030368325, "f1": 0.0560864093959733, "f1_stderr": 0.0013597729822813858, "acc": 0.341030820866541, "acc_stderr": 0.008350924483766176 }, "harness|drop|3": { "em": 0.0020973154362416107, "em_stderr": 0.0004685065030368325, "f1": 0.0560864093959733, "f1_stderr": 0.0013597729822813858 }, "harness|gsm8k|5": { "acc": 0.01592115238817286, "acc_stderr": 0.0034478192723889915 }, "harness|winogrande|5": { "acc": 0.6661404893449092, "acc_stderr": 0.013254029695143358 } }

配置详情

以下是数据集的配置详情:

  • harness_arc_challenge_25

    • 分割:2023_10_12T09_37_10.252705
      • 路径:**/details_harness|arc:challenge|25_2023-10-12T09-37-10.252705.parquet
    • 分割:latest
      • 路径:**/details_harness|arc:challenge|25_2023-10-12T09-37-10.252705.parquet
  • harness_drop_3

    • 分割:2023_10_29T00_43_57.732775
      • 路径:**/details_harness|drop|3_2023-10-29T00-43-57.732775.parquet
    • 分割:latest
      • 路径:**/details_harness|drop|3_2023-10-29T00-43-57.732775.parquet
  • harness_gsm8k_5

    • 分割:2023_10_29T00_43_57.732775
      • 路径:**/details_harness|gsm8k|5_2023-10-29T00-43-57.732775.parquet
    • 分割:latest
      • 路径:**/details_harness|gsm8k|5_2023-10-29T00-43-57.732775.parquet
  • harness_hellaswag_10

    • 分割:2023_10_12T09_37_10.252705
      • 路径:**/details_harness|hellaswag|10_2023-10-12T09-37-10.252705.parquet
    • 分割:latest
      • 路径:**/details_harness|hellaswag|10_2023-10-12T09-37-10.252705.parquet
  • harness_hendrycksTest_5

    • 分割:2023_10_12T09_37_10.252705
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-management|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-prehistory|5_2023-10-12T09-37-10.252705.parquet
        • **/details_harness|hendrycksTest-professional_accounting|5_2023-10-12T09-37-10.252705.parquet
        • `**/details_harness|hendrycksTest-professional_law|5_2023-10-12T09-37-10.252705
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