open-llm-leaderboard-old/details_ajibawa-2023__Code-13B
收藏数据集概述
数据集简介
该数据集是在对模型 ajibawa-2023/Code-13B 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。
数据集结构
- 配置数量:63个配置,每个配置对应一个评估任务。
- 运行次数:数据集由1次运行创建。每个运行在每个配置中都有一个特定的分割,分割名称使用运行的时间戳。
- 分割:每个配置中的 "train" 分割始终指向最新的结果。
- 结果配置:一个额外的 "results" 配置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_ajibawa-2023__Code-13B", "harness_winogrande_5", split="train")
最新结果
以下是 2023-12-09T19:40:16.694610 运行的最新结果:
python { "all": { "acc": 0.5315302469691541, "acc_stderr": 0.0338171547995471, "acc_norm": 0.5374650243523146, "acc_norm_stderr": 0.034550805778528454, "mc1": 0.2962056303549572, "mc1_stderr": 0.01598359510181139, "mc2": 0.4246156253859874, "mc2_stderr": 0.01586771249517698 }, "harness|arc:challenge|25": { "acc": 0.5511945392491467, "acc_stderr": 0.014534599585097665, "acc_norm": 0.5733788395904437, "acc_norm_stderr": 0.014453185592920293 }, "harness|hellaswag|10": { "acc": 0.6420035849432384, "acc_stderr": 0.004784312972495391, "acc_norm": 0.8328022306313483, "acc_norm_stderr": 0.003723897305645496 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.28, "acc_stderr": 0.045126085985421296, "acc_norm": 0.28, "acc_norm_stderr": 0.045126085985421296 }, # 其他任务的结果... }
配置详情
-
harness_arc_challenge_25
- 分割:2023_12_09T19_40_16.694610, latest
- 路径:
**/details_harness|arc:challenge|25_2023-12-09T19-40-16.694610.parquet
-
harness_gsm8k_5
- 分割:2023_12_09T19_40_16.694610, latest
- 路径:
**/details_harness|gsm8k|5_2023-12-09T19-40-16.694610.parquet
-
harness_hellaswag_10
- 分割:2023_12_09T19_40_16.694610, latest
- 路径:
**/details_harness|hellaswag|10_2023-12-09T19-40-16.694610.parquet
-
harness_hendrycksTest_5
- 分割:2023_12_09T19_40_16.694610, latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-anatomy|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-astronomy|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-business_ethics|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-college_biology|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-college_chemistry|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-college_computer_science|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-college_mathematics|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-college_medicine|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-college_physics|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-computer_security|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-econometrics|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-formal_logic|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-global_facts|5_2023-12-09T19-40-16.694610.parquet**/details_harness|hendrycksTest-high_school_biology|5_2023-12-09T19-40-16.694610.parquet
其他路径...
}



