open-llm-leaderboard-old/details_NLUHOPOE__Mistral-7B-loss-100000
收藏数据集概述
数据集基本信息
- 数据集名称: Evaluation run of NLUHOPOE/Mistral-7B-loss-100000
- 创建目的: 该数据集是在对模型 NLUHOPOE/Mistral-7B-loss-100000 进行评估运行期间自动创建的。
- 数据集组成: 包含63个配置，每个配置对应一个评估任务。
- 数据来源: 数据集从2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。
- 最新结果: "train" 分割始终指向最新结果。
- 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 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_NLUHOPOE__Mistral-7B-loss-100000", "harness_winogrande_5", split="train")
最新结果
以下是 最新结果 的摘要： python { "all": { "acc": 0.5369263603517691, "acc_stderr": 0.033848622931727725, "acc_norm": 0.5429723385562029, "acc_norm_stderr": 0.03460142992361602, "mc1": 0.2631578947368421, "mc1_stderr": 0.015415241740237017, "mc2": 0.4092562147054596, "mc2_stderr": 0.014606886822140043 }, "harness|arc:challenge|25": { "acc": 0.47525597269624575, "acc_stderr": 0.01459348769493774, "acc_norm": 0.5179180887372014, "acc_norm_stderr": 0.014602005585490973 }, "harness|hellaswag|10": { "acc": 0.5760804620593507, "acc_stderr": 0.004931679059919374, "acc_norm": 0.7715594503087034, "acc_norm_stderr": 0.004189698894885502 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.29, "acc_stderr": 0.045604802157206845, "acc_norm": 0.29, "acc_norm_stderr": 0.045604802157206845 }, # 其他任务的结果... }
配置信息
-
配置名称: harness_arc_challenge_25
- 数据文件:
- 分割: 2024_01_26T10_00_16.807578
- 路径: **/details_harness|arc:challenge|25_2024-01-26T10-00-16.807578.parquet
- 分割: 2024_01_26T10_12_29.042684
- 路径: **/details_harness|arc:challenge|25_2024-01-26T10-12-29.042684.parquet
- 分割: latest
- 路径: **/details_harness|arc:challenge|25_2024-01-26T10-12-29.042684.parquet
- 分割: 2024_01_26T10_00_16.807578
- 数据文件:
-
配置名称: harness_gsm8k_5
- 数据文件:
- 分割: 2024_01_26T10_00_16.807578
- 路径: **/details_harness|gsm8k|5_2024-01-26T10-00-16.807578.parquet
- 分割: 2024_01_26T10_12_29.042684
- 路径: **/details_harness|gsm8k|5_2024-01-26T10-12-29.042684.parquet
- 分割: latest
- 路径: **/details_harness|gsm8k|5_2024-01-26T10-12-29.042684.parquet
- 分割: 2024_01_26T10_00_16.807578
- 数据文件:
-
配置名称: harness_hellaswag_10
- 数据文件:
- 分割: 2024_01_26T10_00_16.807578
- 路径: **/details_harness|hellaswag|10_2024-01-26T10-00-16.807578.parquet
- 分割: 2024_01_26T10_12_29.042684
- 路径: **/details_harness|hellaswag|10_2024-01-26T10-12-29.042684.parquet
- 分割: latest
- 路径: **/details_harness|hellaswag|10_2024-01-26T10-12-29.042684.parquet
- 分割: 2024_01_26T10_00_16.807578
- 数据文件:
-
配置名称: harness_hendrycksTest_5
- 数据文件:
- 分割: 2024_01_26T10_00_16.807578
- 路径:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-01-26T10-00-16.807578.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-01-26T10-00-16.807578.parquet
其他路径...
- 路径:
- 分割: 2024_01_26T10_00_16.807578
- 数据文件:
}
