open-llm-leaderboard-old/details_akjindal53244__Mistral-7B-v0.1-Open-Platypus
收藏数据集概述
数据集简介
该数据集是在评估模型 akjindal53244/Mistral-7B-v0.1-Open-Platypus 在 Open LLM Leaderboard 上的运行过程中自动创建的。
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。
- 运行次数:数据集从2次运行中创建。每个运行在每个配置中作为一个特定的分割存在,分割名称使用运行的时间戳。
- 训练分割:"train" 分割始终指向最新的结果。
- 结果配置: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 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_akjindal53244__Mistral-7B-v0.1-Open-Platypus", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-25T03:30:37.870273 运行 的最新结果: python { "all": { "em": 0.16128355704697986, "em_stderr": 0.0037665373341562473, "f1": 0.21934249161073788, "f1_stderr": 0.003766121643482467, "acc": 0.47474797642135197, "acc_stderr": 0.011060564905702893 }, "harness|drop|3": { "em": 0.16128355704697986, "em_stderr": 0.0037665373341562473, "f1": 0.21934249161073788, "f1_stderr": 0.003766121643482467 }, "harness|gsm8k|5": { "acc": 0.1728582259287339, "acc_stderr": 0.010415432246200586 }, "harness|winogrande|5": { "acc": 0.77663772691397, "acc_stderr": 0.011705697565205201 } }
配置详情
以下是数据集的配置详情:
配置列表
-
harness_arc_challenge_25
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|arc:challenge|25_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-10-09T12-52-41.880840.parquet
-
harness_drop_3
- 分割:2023_10_25T03_30_37.870273
- 路径:
**/details_harness|drop|3_2023-10-25T03-30-37.870273.parquet - 分割:latest
- 路径:
**/details_harness|drop|3_2023-10-25T03-30-37.870273.parquet
-
harness_gsm8k_5
- 分割:2023_10_25T03_30_37.870273
- 路径:
**/details_harness|gsm8k|5_2023-10-25T03-30-37.870273.parquet - 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-10-25T03-30-37.870273.parquet
-
harness_hellaswag_10
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hellaswag|10_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_abstract_algebra_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_anatomy_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_astronomy_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_business_ethics_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_clinical_knowledge_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_college_biology_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_college_chemistry_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_college_computer_science_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_college_mathematics_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_college_medicine_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_college_physics_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_computer_security_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_conceptual_physics_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_econometrics_5
- 分割:2023_10_09T12_52_41.880840
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-10-09T12-52-41.880840.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-10-09T12-52-41.880840.parquet
-
harness_hendrycksTest_electrical_engineering_5
- 分割:2023_10_09T12_52_41.880840
- 路径:`**/details_har



