five

open-llm-leaderboard/details_TurkuNLP__gpt3-finnish-large

收藏
Hugging Face2023-10-23 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_TurkuNLP__gpt3-finnish-large
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在Open LLM Leaderboard上对TurkuNLP/gpt3-finnish-large模型进行评估时自动生成的。数据集由64个配置组成,每个配置对应一个被评估的任务。数据集由2次运行生成,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,名为results的配置存储了所有运行的聚合结果,用于在Open LLM Leaderboard上计算和显示聚合指标。README还提供了如何使用Python中的datasets库加载运行细节的示例。

This dataset was automatically generated during the evaluation of the TurkuNLP/gpt3-finnish-large model on the Open LLM Leaderboard. It comprises 64 configurations, each corresponding to one evaluated task. The dataset is created 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, a configuration titled `results` stores the aggregated results of all runs, which is used to calculate and display the aggregated metrics on the Open LLM Leaderboard. The accompanying README also provides examples of how to load run details using the `datasets` library in Python.
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 TurkuNLP/gpt3-finnish-largeOpen LLM Leaderboard 上的自动创建的。数据集包含 64 个配置,每个配置对应一个评估任务。

数据集结构

数据集由 2 次运行创建,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。

额外配置

一个额外的配置 "results" 存储了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。

加载数据示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_TurkuNLP__gpt3-finnish-large", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-23T09:44:48.783250 运行的最新结果: python { "all": { "em": 0.029886744966442953, "em_stderr": 0.0017437739254467684, "f1": 0.07029467281879197, "f1_stderr": 0.0020935165533943765, "acc": 0.25769534333070243, "acc_stderr": 0.007022913394891833 }, "harness|drop|3": { "em": 0.029886744966442953, "em_stderr": 0.0017437739254467684, "f1": 0.07029467281879197, "f1_stderr": 0.0020935165533943765 }, "harness|gsm8k|5": { "acc": 0.0, "acc_stderr": 0.0 }, "harness|winogrande|5": { "acc": 0.5153906866614049, "acc_stderr": 0.014045826789783666 } }

配置详情

以下是数据集的配置详情:

配置列表

  • harness_arc_challenge_25

    • 分割: 2023_10_04T06_41_08.773247
      • 路径: **/details_harness|arc:challenge|25_2023-10-04T06-41-08.773247.parquet
    • 分割: latest
      • 路径: **/details_harness|arc:challenge|25_2023-10-04T06-41-08.773247.parquet
  • harness_drop_3

    • 分割: 2023_10_23T09_44_48.783250
      • 路径: **/details_harness|drop|3_2023-10-23T09-44-48.783250.parquet
    • 分割: latest
      • 路径: **/details_harness|drop|3_2023-10-23T09-44-48.783250.parquet
  • harness_gsm8k_5

    • 分割: 2023_10_23T09_44_48.783250
      • 路径: **/details_harness|gsm8k|5_2023-10-23T09-44-48.783250.parquet
    • 分割: latest
      • 路径: **/details_harness|gsm8k|5_2023-10-23T09-44-48.783250.parquet
  • harness_hellaswag_10

    • 分割: 2023_10_04T06_41_08.773247
      • 路径: **/details_harness|hellaswag|10_2023-10-04T06-41-08.773247.parquet
    • 分割: latest
      • 路径: **/details_harness|hellaswag|10_2023-10-04T06-41-08.773247.parquet
  • harness_hendrycksTest_5

    • 分割: 2023_10_04T06_41_08.773247
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-management|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-prehistory|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-professional_accounting|5_2023-10-04T06-41-08.773247.parquet
        • **/details_harness|hendrycksTest-professional_law|5_2023-10-04T06-41-08.773247.parquet
        • `**/details_harness|hendrycksTest-professional_medicine|5_2023-10-04
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