five

open-llm-leaderboard-old/details_maywell__Synatra-7B-v0.3-RP

收藏
Hugging Face2023-11-18 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_maywell__Synatra-7B-v0.3-RP
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maywell/Synatra-7B-v0.3-RP的评估运行期间自动创建的,用于在Open LLM Leaderboard上进行评估。数据集由64个配置组成,每个配置对应一个评估任务。数据集是从1次运行中创建的,每次运行都可以在特定配置中找到,分割以运行的时间戳命名。train分割始终指向最新结果。此外,results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。

该数据集是在模型maywell/Synatra-7B-v0.3-RP的评估运行期间自动创建的,用于在Open LLM Leaderboard上进行评估。数据集由64个配置组成,每个配置对应一个评估任务。数据集是从1次运行中创建的,每次运行都可以在特定配置中找到,分割以运行的时间戳命名。train分割始终指向最新结果。此外,results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。
提供机构:
open-llm-leaderboard-old
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 maywell/Synatra-7B-v0.3-RP 进行评估运行时自动创建的,用于 Open LLM Leaderboard

数据集结构

  • 配置数量:64个配置,每个配置对应一个评估任务。
  • 创建来源:从1次运行中创建。每个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。
  • 最新结果:"train" 分割始终指向最新结果。
  • 聚合结果:一个额外的配置 "results" 存储所有运行的聚合结果,用于计算和显示聚合指标。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_maywell__Synatra-7B-v0.3-RP_public", "harness_winogrande_5", split="train")

最新结果

这些是最新结果的示例,来自2023-11-18T23:02:29.150817的运行: python { "all": { "acc": 0.6042478184701645, "acc_stderr": 0.03267991789724199, "acc_norm": 0.6118798356357696, "acc_norm_stderr": 0.03337492967666177, "mc1": 0.37209302325581395, "mc1_stderr": 0.016921090118814035, "mc2": 0.5263791321103062, "mc2_stderr": 0.015312628675104242, "em": 0.3953439597315436, "em_stderr": 0.005007043944789993, "f1": 0.46059983221476697, "f1_stderr": 0.00481810685968407 }, "harness|arc:challenge|25": { "acc": 0.5930034129692833, "acc_stderr": 0.014356399418009117, "acc_norm": 0.6220136518771331, "acc_norm_stderr": 0.014169664520303098 }, "harness|hellaswag|10": { "acc": 0.6338378809002191, "acc_stderr": 0.004807699539973411, "acc_norm": 0.8229436367257519, "acc_norm_stderr": 0.003809362761248109 }, # 其他任务的结果... }

配置详情

  • 配置名称:harness_arc_challenge_25

    • 数据文件
      • 分割:2023_11_18T23_02_29.150817
        • 路径:**/details_harness|arc:challenge|25_2023-11-18T23-02-29.150817.parquet
      • 分割:latest
        • 路径:**/details_harness|arc:challenge|25_2023-11-18T23-02-29.150817.parquet
  • 配置名称:harness_drop_3

    • 数据文件
      • 分割:2023_11_18T23_02_29.150817
        • 路径:**/details_harness|drop|3_2023-11-18T23-02-29.150817.parquet
      • 分割:latest
        • 路径:**/details_harness|drop|3_2023-11-18T23-02-29.150817.parquet
  • 配置名称:harness_gsm8k_5

    • 数据文件
      • 分割:2023_11_18T23_02_29.150817
        • 路径:**/details_harness|gsm8k|5_2023-11-18T23-02-29.150817.parquet
      • 分割:latest
        • 路径:**/details_harness|gsm8k|5_2023-11-18T23-02-29.150817.parquet
  • 配置名称:harness_hellaswag_10

    • 数据文件
      • 分割:2023_11_18T23_02_29.150817
        • 路径:**/details_harness|hellaswag|10_2023-11-18T23-02-29.150817.parquet
      • 分割:latest
        • 路径:**/details_harness|hellaswag|10_2023-11-18T23-02-29.150817.parquet
  • 配置名称:harness_hendrycksTest_5

    • 数据文件
      • 分割:2023_11_18T23_02_29.150817
        • 路径:
          • **/details_harness|hendrycksTest-abstract_algebra|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-anatomy|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-astronomy|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-business_ethics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-college_biology|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-college_chemistry|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-college_computer_science|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-college_mathematics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-college_medicine|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-college_physics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-computer_security|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-conceptual_physics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-econometrics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-electrical_engineering|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-formal_logic|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-global_facts|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_biology|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_european_history|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_geography|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_physics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_psychology|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_statistics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_us_history|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-high_school_world_history|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-human_aging|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-human_sexuality|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-international_law|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-jurisprudence|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-logical_fallacies|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-machine_learning|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-management|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-marketing|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-medical_genetics|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-miscellaneous|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-moral_disputes|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-moral_scenarios|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-nutrition|5_2023-11-18T23-02-29.150817.parquet
          • **/details_harness|hendrycksTest-philosophy|5_2023-11-18T23-02-29.150817.parquet
          • `**/details_harness|hendrycksTest-prehistory|5_2023-11-18T23-
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