five

open-llm-leaderboard-old/details_ValiantLabs__Fireplace-13b

收藏
Hugging Face2024-01-18 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_ValiantLabs__Fireplace-13b
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ValiantLabs/Fireplace-13b在Open LLM Leaderboard上的评估运行期间自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集是从一个运行中创建的,每个运行作为特定分割存储,分割名称使用运行的时间戳命名。train分割始终指向最新的结果。此外,results配置存储了运行的所有聚合结果,用于在Leaderboard上计算和显示聚合指标。数据集包含多种任务,如arc_challenge、gsm8k、hellaswag以及hendrycksTest下的多个测试,涵盖了广泛的主题。README还提供了一个Python示例,说明如何加载数据集。

该数据集是在模型ValiantLabs/Fireplace-13b在Open LLM Leaderboard上的评估运行期间自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集是从一个运行中创建的,每个运行作为特定分割存储,分割名称使用运行的时间戳命名。train分割始终指向最新的结果。此外,results配置存储了运行的所有聚合结果,用于在Leaderboard上计算和显示聚合指标。数据集包含多种任务,如arc_challenge、gsm8k、hellaswag以及hendrycksTest下的多个测试,涵盖了广泛的主题。README还提供了一个Python示例,说明如何加载数据集。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

数据集基本信息

  • 数据集名称: Evaluation run of ValiantLabs/Fireplace-13b
  • 数据集来源: 自动创建于模型 ValiantLabs/Fireplace-13bOpen LLM Leaderboard 的评估运行中。
  • 数据集结构: 包含63个配置,每个配置对应一个评估任务。
  • 数据集创建: 从1次运行中创建,每次运行的详细结果存储在特定的分割中,分割名称使用运行的时间戳。"train" 分割始终指向最新的结果。

数据集加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_ValiantLabs__Fireplace-13b", "harness_winogrande_5", split="train")

最新结果

  • 最新结果时间戳: 2024-01-18T22:29:29.742832
  • 结果示例: python { "all": { "acc": 0.4379234507249955, "acc_stderr": 0.034366003843291505, "acc_norm": 0.44072254848921605, "acc_norm_stderr": 0.03509834671600673, "mc1": 0.3182374541003672, "mc1_stderr": 0.01630598864892061, "mc2": 0.48240026363735705, "mc2_stderr": 0.015361662513373581 }, "harness|arc:challenge|25": { "acc": 0.4325938566552901, "acc_stderr": 0.014478005694182528, "acc_norm": 0.47696245733788395, "acc_norm_stderr": 0.01459587320535827 }, "harness|hellaswag|10": { "acc": 0.5217088229436367, "acc_stderr": 0.004985076094464753, "acc_norm": 0.6960764787890859, "acc_norm_stderr": 0.0045901000501988335 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.28, "acc_stderr": 0.04512608598542129, "acc_norm": 0.28, "acc_norm_stderr": 0.04512608598542129 },

    其他任务结果省略...

}

配置信息

  • 配置名称: harness_arc_challenge_25

    • 数据文件:
      • 分割: 2024_01_18T22_29_29.742832
        • 路径: **/details_harness|arc:challenge|25_2024-01-18T22-29-29.742832.parquet
      • 分割: latest
        • 路径: **/details_harness|arc:challenge|25_2024-01-18T22-29-29.742832.parquet
  • 配置名称: harness_gsm8k_5

    • 数据文件:
      • 分割: 2024_01_18T22_29_29.742832
        • 路径: **/details_harness|gsm8k|5_2024-01-18T22-29-29.742832.parquet
      • 分割: latest
        • 路径: **/details_harness|gsm8k|5_2024-01-18T22-29-29.742832.parquet
  • 配置名称: harness_hellaswag_10

    • 数据文件:
      • 分割: 2024_01_18T22_29_29.742832
        • 路径: **/details_harness|hellaswag|10_2024-01-18T22-29-29.742832.parquet
      • 分割: latest
        • 路径: **/details_harness|hellaswag|10_2024-01-18T22-29-29.742832.parquet
  • 配置名称: harness_hendrycksTest_5

    • 数据文件:
      • 分割: 2024_01_18T22_29_29.742832
        • 路径:
          • **/details_harness|hendrycksTest-abstract_algebra|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-anatomy|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-astronomy|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-business_ethics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-college_biology|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-college_chemistry|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-college_computer_science|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-college_mathematics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-college_medicine|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-college_physics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-computer_security|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-conceptual_physics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-econometrics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-electrical_engineering|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-formal_logic|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-global_facts|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_biology|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_chemistry|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_computer_science|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_european_history|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_geography|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_mathematics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_microeconomics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_physics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_psychology|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_statistics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_us_history|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-high_school_world_history|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-human_aging|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-human_sexuality|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-international_law|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-jurisprudence|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-logical_fallacies|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-machine_learning|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-management|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-marketing|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-medical_genetics|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-miscellaneous|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-moral_disputes|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-moral_scenarios|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-nutrition|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-philosophy|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-prehistory|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-professional_accounting|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-professional_law|5_2024-01-18T22-29-29.742832.parquet
          • **/details_harness|hendrycksTest-professional_medicine|5_2024-01-18T22-29-29.742832.parquet
          • `**/details_harness|hendrycksTest-professional_psychology|5_20
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