open-llm-leaderboard/details_breadlicker45__dough-instruct-base-001
收藏数据集概述
该数据集是在对模型 breadlicker45/dough-instruct-base-001 进行评估运行期间自动创建的。数据集包含 64 个配置,每个配置对应一个评估任务。数据集从 2 次运行中创建,每次运行可以在每个配置中找到一个特定的分割,分割名称使用运行的时间戳。"train" 分割始终指向最新的结果。
数据集结构
配置列表
-
harness_arc_challenge_25
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|arc:challenge|25_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_drop_3
- 分割:
2023_09_16T20_08_34.854052,latest - 路径:
**/details_harness|drop|3_2023-09-16T20-08-34.854052.parquet
- 分割:
-
harness_gsm8k_5
- 分割:
2023_09_16T20_08_34.854052,latest - 路径:
**/details_harness|gsm8k|5_2023-09-16T20-08-34.854052.parquet
- 分割:
-
harness_hellaswag_10
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hellaswag|10_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径: 多个文件,例如
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_abstract_algebra_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_anatomy_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_astronomy_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_business_ethics_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_clinical_knowledge_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_college_biology_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_college_chemistry_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_college_computer_science_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_college_mathematics_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_college_medicine_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_college_physics_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_computer_security_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_conceptual_physics_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_econometrics_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_electrical_engineering_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-02T14:53:33.597217.parquet
- 分割:
-
harness_hendrycksTest_elementary_mathematics_5
- 分割:
2023_08_02T14_53_33.597217,latest - 路径:
**/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-02T14:53:33.597217.parquet
- 分割:
最新结果
以下是来自 2023-09-16T20:08:34.854052 运行的最新结果:
python { "all": { "em": 0.0, "em_stderr": 0.0, "f1": 0.0029163171140939564, "f1_stderr": 0.00019355490209304062, "acc": 0.255327545382794, "acc_stderr": 0.007024647268145198 }, "harness|drop|3": { "em": 0.0, "em_stderr": 0.0, "f1": 0.0029163171140939564, "f1_stderr": 0.00019355490209304062 }, "harness|gsm8k|5": { "acc": 0.0, "acc_stderr": 0.0 }, "harness|winogrande|5": { "acc": 0.510655090765588, "acc_stderr": 0.014049294536290396 } }



