five

open-llm-leaderboard/details_mosaicml__mpt-7b-storywriter

收藏
Hugging Face2023-10-16 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_mosaicml__mpt-7b-storywriter
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mosaicml/mpt-7b-storywriter在Open LLM Leaderboard上的评估运行期间自动创建的。数据集由64个配置组成,每个配置对应一个评估任务。数据集是从4次运行中生成的,每次运行在每个配置中表示为特定的分割,分割名称使用运行的时间戳。train分割始终指向最新的结果。一个名为results的额外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。README还提供了如何使用`datasets`库中的`load_dataset`函数加载运行中的详细信息的示例。
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 mosaicml/mpt-7b-storywriterOpen LLM Leaderboard 上的评估运行期间自动创建的。

数据集组成

  • 数据集包含 64 个配置,每个配置对应一个评估任务。
  • 数据集由 4 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_mosaicml__mpt-7b-storywriter", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-16T08:53:05.263222 运行的最新结果: python { "all": { "em": 0.0006291946308724832, "em_stderr": 0.00025680027497237983, "f1": 0.0032026006711409396, "f1_stderr": 0.0005040610386397096, "acc": 0.2557221783741121, "acc_stderr": 0.0070244020999296625 }, "harness|drop|3": { "em": 0.0006291946308724832, "em_stderr": 0.00025680027497237983, "f1": 0.0032026006711409396, "f1_stderr": 0.0005040610386397096 }, "harness|gsm8k|5": { "acc": 0.0, "acc_stderr": 0.0 }, "harness|winogrande|5": { "acc": 0.5114443567482242, "acc_stderr": 0.014048804199859325 } }

配置详情

  • harness_arc_challenge_25

    • 分割:2023_07_20T10_23_53.118062
      • 路径:**/details_harness|arc:challenge|25_2023-07-20T10:23:53.118062.parquet
    • 分割:2023_10_03T22_53_23.133729
      • 路径:**/details_harness|arc:challenge|25_2023-10-03T22-53-23.133729.parquet
    • 分割:latest
      • 路径:**/details_harness|arc:challenge|25_2023-10-03T22-53-23.133729.parquet
  • harness_drop_3

    • 分割:2023_09_22T15_18_47.960530
      • 路径:**/details_harness|drop|3_2023-09-22T15-18-47.960530.parquet
    • 分割:2023_10_16T08_53_05.263222
      • 路径:**/details_harness|drop|3_2023-10-16T08-53-05.263222.parquet
    • 分割:latest
      • 路径:**/details_harness|drop|3_2023-10-16T08-53-05.263222.parquet
  • harness_gsm8k_5

    • 分割:2023_09_22T15_18_47.960530
      • 路径:**/details_harness|gsm8k|5_2023-09-22T15-18-47.960530.parquet
    • 分割:2023_10_16T08_53_05.263222
      • 路径:**/details_harness|gsm8k|5_2023-10-16T08-53-05.263222.parquet
    • 分割:latest
      • 路径:**/details_harness|gsm8k|5_2023-10-16T08-53-05.263222.parquet
  • harness_hellaswag_10

    • 分割:2023_07_20T10_23_53.118062
      • 路径:**/details_harness|hellaswag|10_2023-07-20T10:23:53.118062.parquet
    • 分割:2023_10_03T22_53_23.133729
      • 路径:**/details_harness|hellaswag|10_2023-10-03T22-53-23.133729.parquet
    • 分割:latest
      • 路径:**/details_harness|hellaswag|10_2023-10-03T22-53-23.133729.parquet
  • harness_hendrycksTest_5

    • 分割:2023_07_20T10_23_53.118062
      • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-anatomy|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-astronomy|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-business_ethics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-college_biology|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-college_chemistry|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-college_computer_science|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-college_mathematics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-college_medicine|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-college_physics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-computer_security|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-econometrics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-formal_logic|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-global_facts|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_biology|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_geography|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_physics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-human_aging|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-human_sexuality|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-international_law|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-jurisprudence|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-machine_learning|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-management|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-marketing|5_2023-07-20T10:23:53.118062.parquet
      • 路径:**/details_harness|hendrycksTest-medical_genetics|5_2023-07-20T10:23:53.118062.parquet
      • 路径:`**/details_harness|
二维码
社区交流群
二维码
科研交流群
商业服务