five

open-llm-leaderboard-old/details_cognitivecomputations__openchat-3.5-0106-laser

收藏
Hugging Face2024-01-27 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_cognitivecomputations__openchat-3.5-0106-laser
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cognitivecomputations/openchat-3.5-0106-laser评估运行期间自动创建的,用于Open LLM排行榜。数据集包含63个配置,每个配置对应一个评估任务。数据集来源于1次运行,每次运行作为一个特定分割,以运行的时间戳命名。train分割始终指向最新结果。另外,results配置存储了运行中的所有聚合结果,用于在排行榜上计算和显示聚合指标。README文件还提供了使用HuggingFace的datasets库加载数据集的示例。

该数据集是在模型cognitivecomputations/openchat-3.5-0106-laser评估运行期间自动创建的,用于Open LLM排行榜。数据集包含63个配置,每个配置对应一个评估任务。数据集来源于1次运行,每次运行作为一个特定分割,以运行的时间戳命名。train分割始终指向最新结果。另外,results配置存储了运行中的所有聚合结果,用于在排行榜上计算和显示聚合指标。README文件还提供了使用HuggingFace的datasets库加载数据集的示例。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

该数据集是在对模型 cognitivecomputations/openchat-3.5-0106-laser 进行评估运行期间自动创建的,评估结果发布在 Open LLM Leaderboard 上。

数据集组成

  • 数据集包含 63 个配置,每个配置对应一个评估任务。
  • 数据集由 1 次运行创建,每次运行的详细结果可以在每个配置中找到,并以运行的时间戳命名。
  • "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,用于计算和显示在 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_cognitivecomputations__openchat-3.5-0106-laser", "harness_winogrande_5", split="train")

最新结果

以下是 2024-01-27T06:11:53.971032 运行 的最新结果:

python { "all": { "acc": 0.6536170737766268, "acc_stderr": 0.031877905637757095, "acc_norm": 0.6542883499910643, "acc_norm_stderr": 0.03253360388122567, "mc1": 0.3574051407588739, "mc1_stderr": 0.016776599676729412, "mc2": 0.5207887413270008, "mc2_stderr": 0.01528579867134112 }, "harness|arc:challenge|25": { "acc": 0.6237201365187713, "acc_stderr": 0.014157022555407156, "acc_norm": 0.6604095563139932, "acc_norm_stderr": 0.01383903976282017 }, "harness|hellaswag|10": { "acc": 0.6324437363075085, "acc_stderr": 0.004811543077792714, "acc_norm": 0.8318064130651265, "acc_norm_stderr": 0.0037327367704297182 }, # 其他任务的结果... }

配置详情

  • harness_arc_challenge_25

    • 分割: 2024_01_27T06_11_53.971032, latest
    • 路径: **/details_harness|arc:challenge|25_2024-01-27T06-11-53.971032.parquet
  • harness_gsm8k_5

    • 分割: 2024_01_27T06_11_53.971032, latest
    • 路径: **/details_harness|gsm8k|5_2024-01-27T06-11-53.971032.parquet
  • harness_hellaswag_10

    • 分割: 2024_01_27T06_11_53.971032, latest
    • 路径: **/details_harness|hellaswag|10_2024-01-27T06-11-53.971032.parquet
  • harness_hendrycksTest_5

    • 分割: 2024_01_27T06_11_53.971032, latest
    • 路径:
      • **/details_harness|hendrycksTest-abstract_algebra|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-anatomy|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-astronomy|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-business_ethics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-college_biology|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-college_chemistry|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-college_computer_science|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-college_mathematics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-college_medicine|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-college_physics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-computer_security|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-conceptual_physics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-econometrics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-electrical_engineering|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-formal_logic|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-global_facts|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_biology|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_chemistry|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_computer_science|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_european_history|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_geography|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_mathematics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_microeconomics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_physics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_psychology|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_statistics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_us_history|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-high_school_world_history|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-human_aging|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-human_sexuality|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-international_law|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-jurisprudence|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-logical_fallacies|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-machine_learning|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-management|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-marketing|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-medical_genetics|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-miscellaneous|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-moral_disputes|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-moral_scenarios|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-nutrition|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-philosophy|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-prehistory|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-professional_accounting|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-professional_law|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-professional_medicine|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-professional_psychology|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-public_relations|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-security_studies|5_2024-01-27T06-11-53.971032.parquet
      • **/details_harness|hendrycksTest-sociology|5_2024-01-27T06-11-53.971032.parquet
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