open-llm-leaderboard-old/details_cookinai__CatMacaroni-Slerp
收藏Hugging Face2023-12-20 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_cookinai__CatMacaroni-Slerp
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 cookinai/CatMacaroni-Slerp 在 Open LLM Leaderboard 上进行评估时自动生成的。数据集包含 63 个配置,每个配置对应一个被评估的任务。它包含一次或多次运行的结果,每次运行都作为每个配置中的一个特定分割存储。train 分割始终指向最新结果。一个名为 results 的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示 Open LLM Leaderboard 上的指标。README 还提供了一个 Python 代码片段来加载数据集详细信息,并列出了特定运行的最新结果。
该数据集是在模型 cookinai/CatMacaroni-Slerp 在 Open LLM Leaderboard 上进行评估时自动生成的。数据集包含 63 个配置,每个配置对应一个被评估的任务。它包含一次或多次运行的结果,每次运行都作为每个配置中的一个特定分割存储。train 分割始终指向最新结果。一个名为 results 的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示 Open LLM Leaderboard 上的指标。README 还提供了一个 Python 代码片段来加载数据集详细信息,并列出了特定运行的最新结果。
提供机构:
open-llm-leaderboard-old
原始信息汇总
数据集概述
数据集信息
- 数据集名称: Evaluation run of cookinai/CatMacaroni-Slerp
- 数据集来源: 自动创建于模型 cookinai/CatMacaroni-Slerp 在 Open LLM Leaderboard 的评估运行中。
- 数据集组成: 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: "results" 配置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_cookinai__CatMacaroni-Slerp", "harness_winogrande_5", split="train")
最新结果
- 最新结果来源: 2023-12-20T21:17:23.139479
- 聚合结果: python { "all": { "acc": 0.6549634202662564, "acc_stderr": 0.032063571652802186, "acc_norm": 0.6546746443892243, "acc_norm_stderr": 0.032730217428552345, "mc1": 0.4418604651162791, "mc1_stderr": 0.017384767478986218, "mc2": 0.6102215759974746, "mc2_stderr": 0.015132806306597834 }, "harness|arc:challenge|25": { "acc": 0.6663822525597269, "acc_stderr": 0.013778687054176541, "acc_norm": 0.6928327645051194, "acc_norm_stderr": 0.013481034054980941 }, "harness|hellaswag|10": { "acc": 0.6799442342162916, "acc_stderr": 0.00465544276659947, "acc_norm": 0.8687512447719578, "acc_norm_stderr": 0.0033698210047622503 }, ... }
配置信息
-
配置名称: harness_arc_challenge_25
- 数据文件:
- 分割: 2023_12_20T21_17_23.139479
- 路径:
**/details_harness|arc:challenge|25_2023-12-20T21-17-23.139479.parquet
- 路径:
- 分割: latest
- 路径:
**/details_harness|arc:challenge|25_2023-12-20T21-17-23.139479.parquet
- 路径:
- 分割: 2023_12_20T21_17_23.139479
- 数据文件:
-
配置名称: harness_gsm8k_5
- 数据文件:
- 分割: 2023_12_20T21_17_23.139479
- 路径:
**/details_harness|gsm8k|5_2023-12-20T21-17-23.139479.parquet
- 路径:
- 分割: latest
- 路径:
**/details_harness|gsm8k|5_2023-12-20T21-17-23.139479.parquet
- 路径:
- 分割: 2023_12_20T21_17_23.139479
- 数据文件:
-
配置名称: harness_hellaswag_10
- 数据文件:
- 分割: 2023_12_20T21_17_23.139479
- 路径:
**/details_harness|hellaswag|10_2023-12-20T21-17-23.139479.parquet
- 路径:
- 分割: latest
- 路径:
**/details_harness|hellaswag|10_2023-12-20T21-17-23.139479.parquet
- 路径:
- 分割: 2023_12_20T21_17_23.139479
- 数据文件:
-
配置名称: harness_hendrycksTest_5
- 数据文件:
- 分割: 2023_12_20T21_17_23.139479
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-anatomy|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-astronomy|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-business_ethics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_biology|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_chemistry|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_computer_science|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_mathematics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_medicine|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_physics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-computer_security|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-econometrics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-formal_logic|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-global_facts|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-high_school_biology|5_2023-12-20T21-17-23.139479.parquet- ...
- 路径:
- 分割: latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-anatomy|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-astronomy|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-business_ethics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_biology|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_chemistry|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_computer_science|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_mathematics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_medicine|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-college_physics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-computer_security|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-econometrics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-formal_logic|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-global_facts|5_2023-12-20T21-17-23.139479.parquet**/details_harness|hendrycksTest-high_school_biology|5_2023-12-20T21-17-23.139479.parquet- ...
- 路径:
- 分割: 2023_12_20T21_17_23.139479
- 数据文件:



