open-llm-leaderboard-old/details_instructkr__lynn-7b-alpha
收藏数据集概述
该数据集是在评估模型instructkr/lynn-7b-alpha在Open LLM Leaderboard上的运行过程中自动创建的。
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,每个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。
- 额外的配置"results"存储所有聚合的运行结果,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_instructkr__lynn-7b-alpha", "harness_winogrande_5", split="train")
最新结果
以下是2024-04-04T20:00:20.469937运行的最新结果: python { "all": { "acc": 0.6244332916260128, "acc_stderr": 0.03278365538347851, "acc_norm": 0.6304221161202934, "acc_norm_stderr": 0.033454447449031026, "mc1": 0.2998776009791922, "mc1_stderr": 0.016040352966713623, "mc2": 0.45420633450837655, "mc2_stderr": 0.014804075164141449 }, "harness|arc:challenge|25": { "acc": 0.5802047781569966, "acc_stderr": 0.014422181226303026, "acc_norm": 0.6220136518771331, "acc_norm_stderr": 0.0141696645203031 }, "harness|hellaswag|10": { "acc": 0.6464847639912368, "acc_stderr": 0.004770838678356039, "acc_norm": 0.8417645887273452, "acc_norm_stderr": 0.003642157166162359 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.33, "acc_stderr": 0.04725815626252606, "acc_norm": 0.33, "acc_norm_stderr": 0.04725815626252606 }, # 其他任务的结果... }
配置详情
-
config_name: harness_arc_challenge_25
- 数据文件:
- 分割: 2024_04_04T20_00_20.469937
- 路径: **/details_harness|arc:challenge|25_2024-04-04T20-00-20.469937.parquet
- 分割: latest
- 路径: **/details_harness|arc:challenge|25_2024-04-04T20-00-20.469937.parquet
- 分割: 2024_04_04T20_00_20.469937
- 数据文件:
-
config_name: harness_gsm8k_5
- 数据文件:
- 分割: 2024_04_04T20_00_20.469937
- 路径: **/details_harness|gsm8k|5_2024-04-04T20-00-20.469937.parquet
- 分割: latest
- 路径: **/details_harness|gsm8k|5_2024-04-04T20-00-20.469937.parquet
- 分割: 2024_04_04T20_00_20.469937
- 数据文件:
-
config_name: harness_hellaswag_10
- 数据文件:
- 分割: 2024_04_04T20_00_20.469937
- 路径: **/details_harness|hellaswag|10_2024-04-04T20-00-20.469937.parquet
- 分割: latest
- 路径: **/details_harness|hellaswag|10_2024-04-04T20-00-20.469937.parquet
- 分割: 2024_04_04T20_00_20.469937
- 数据文件:
-
config_name: harness_hendrycksTest_5
- 数据文件:
- 分割: 2024_04_04T20_00_20.469937
- 路径:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-04-04T20-00-20.469937.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2024-04-04T20-00-20.469937.parquet
其他任务的路径...
- 路径:
- 分割: 2024_04_04T20_00_20.469937
- 数据文件:
}



