open-llm-leaderboard-old/details_mindy-labs__mindy-7b-v2
收藏数据集概述
该数据集是在对模型 mindy-labs/mindy-7b-v2 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。
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。
- 数据集从 1 次运行中创建,每个运行可以在每个配置中作为一个特定的分割找到,分割名称使用运行的时间戳。
- "train" 分割始终指向最新的结果。
- 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_mindy-labs__mindy-7b-v2", "harness_winogrande_5", split="train")
最新结果
这些是最新结果的示例,来自 2023-12-21T18:22:51.264759 的运行: python { "all": { "acc": 0.6558321041397203, "acc_stderr": 0.03207006697624872, "acc_norm": 0.6560363290954173, "acc_norm_stderr": 0.0327312814050994, "mc1": 0.44063647490820074, "mc1_stderr": 0.017379697555437446, "mc2": 0.6016405207483612, "mc2_stderr": 0.015192119540299543 }, "harness|arc:challenge|25": { "acc": 0.6535836177474402, "acc_stderr": 0.013905011180063235, "acc_norm": 0.6868600682593856, "acc_norm_stderr": 0.013552671543623492 }, "harness|hellaswag|10": { "acc": 0.678550089623581, "acc_stderr": 0.004660785616933756, "acc_norm": 0.8658633738299144, "acc_norm_stderr": 0.0034010255178737263 }, "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
- 分割: 2023_12_21T18_22_51.264759
- 路径: **/details_harness|arc:challenge|25_2023-12-21T18-22-51.264759.parquet
- 分割: latest
- 路径: **/details_harness|arc:challenge|25_2023-12-21T18-22-51.264759.parquet
- 分割: 2023_12_21T18_22_51.264759
-
config_name: harness_gsm8k_5
- 分割: 2023_12_21T18_22_51.264759
- 路径: **/details_harness|gsm8k|5_2023-12-21T18-22-51.264759.parquet
- 分割: latest
- 路径: **/details_harness|gsm8k|5_2023-12-21T18-22-51.264759.parquet
- 分割: 2023_12_21T18_22_51.264759
-
config_name: harness_hellaswag_10
- 分割: 2023_12_21T18_22_51.264759
- 路径: **/details_harness|hellaswag|10_2023-12-21T18-22-51.264759.parquet
- 分割: latest
- 路径: **/details_harness|hellaswag|10_2023-12-21T18-22-51.264759.parquet
- 分割: 2023_12_21T18_22_51.264759
-
config_name: harness_hendrycksTest_5
- 分割: 2023_12_21T18_22_51.264759
- 路径:
- **/details_harness|hendrycksTest-abstract_algebra|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-anatomy|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-astronomy|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-college_biology|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-college_physics|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-computer_security|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-econometrics|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-global_facts|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2023-12-21T18-22-51.264759.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2023-12-21T18-22-51.264759.parquet
其他路径...
- 路径:
- 分割: 2023_12_21T18_22_51.264759
}



