open-llm-leaderboard/details_Open-Orca__OpenOrca-Preview1-13B
收藏数据集概述
数据集简介
该数据集是在评估模型 Open-Orca/OpenOrca-Preview1-13B 在 Open LLM Leaderboard 上的自动创建的。数据集包含 64 个配置,每个配置对应一个评估任务。
数据集结构
数据集由 4 次运行创建,每次运行的结果可以在每个配置中找到,以运行的时间戳命名的特定分割形式存储。"train" 分割始终指向最新的结果。
额外配置
有一个额外的配置 "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
以下是加载特定运行详细信息的示例代码: python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_Open-Orca__OpenOrca-Preview1-13B", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-14T20:21:09.766845 运行的最新结果: python { "all": { "em": 0.29068791946308725, "em_stderr": 0.0046502020293331045, "f1": 0.3704488255033573, "f1_stderr": 0.004551116258249772, "acc": 0.37980957088224854, "acc_stderr": 0.009355350731725371 }, "harness|drop|3": { "em": 0.29068791946308725, "em_stderr": 0.0046502020293331045, "f1": 0.3704488255033573, "f1_stderr": 0.004551116258249772 }, "harness|gsm8k|5": { "acc": 0.04927975739196361, "acc_stderr": 0.005962150655812475 }, "harness|winogrande|5": { "acc": 0.7103393843725335, "acc_stderr": 0.012748550807638268 } }
配置详情
以下是数据集的配置详情:
配置列表
-
config_name: harness_arc_challenge_25
- data_files:
- split: 2023_07_19T18_45_02.691094
- path: **/details_harness|arc:challenge|25_2023-07-19T18:45:02.691094.parquet
- split: 2023_07_25T10_13_37.099822
- path: **/details_harness|arc:challenge|25_2023-07-25T10:13:37.099822.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2023-07-25T10:13:37.099822.parquet
- split: 2023_07_19T18_45_02.691094
- data_files:
-
config_name: harness_drop_3
- data_files:
- split: 2023_09_16T21_46_22.998704
- path: **/details_harness|drop|3_2023-09-16T21-46-22.998704.parquet
- split: 2023_10_14T20_21_09.766845
- path: **/details_harness|drop|3_2023-10-14T20-21-09.766845.parquet
- split: latest
- path: **/details_harness|drop|3_2023-10-14T20-21-09.766845.parquet
- split: 2023_09_16T21_46_22.998704
- data_files:
-
config_name: harness_gsm8k_5
- data_files:
- split: 2023_09_16T21_46_22.998704
- path: **/details_harness|gsm8k|5_2023-09-16T21-46-22.998704.parquet
- split: 2023_10_14T20_21_09.766845
- path: **/details_harness|gsm8k|5_2023-10-14T20-21-09.766845.parquet
- split: latest
- path: **/details_harness|gsm8k|5_2023-10-14T20-21-09.766845.parquet
- split: 2023_09_16T21_46_22.998704
- data_files:
-
config_name: harness_hellaswag_10
- data_files:
- split: 2023_07_19T18_45_02.691094
- path: **/details_harness|hellaswag|10_2023-07-19T18:45:02.691094.parquet
- split: 2023_07_25T10_13_37.099822
- path: **/details_harness|hellaswag|10_2023-07-25T10:13:37.099822.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2023-07-25T10:13:37.099822.parquet
- split: 2023_07_19T18_45_02.691094
- data_files:
-
config_name: harness_hendrycksTest_5
- data_files:
- split: 2023_07_19T18_45_02.691094
- path:
- **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-anatomy|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-astronomy|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-college_biology|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-college_physics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-computer_security|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-econometrics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-global_facts|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-human_aging|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-international_law|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-machine_learning|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-management|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-marketing|5_2023-07-19T18:45:02.691094.parquet
- **/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T18:45:02.691094.parquet
- path:
- split: 2023_07_19T18_45_02.691094
- data_files:



