open-llm-leaderboard/details_ashercn97__manatee-7b
收藏数据集概述
该数据集是在评估模型 ashercn97/manatee-7b 的过程中自动创建的,用于 Open LLM Leaderboard。数据集包含 64 个配置,每个配置对应一个评估任务。
数据集结构
数据集由 2 次运行结果组成,每次运行的结果可以在每个配置中找到,并以运行的时间戳命名。"train" 分割始终指向最新的结果。此外,一个额外的配置 "results" 存储了所有运行结果的聚合数据,用于计算和显示聚合指标。
数据加载示例
以下是加载特定运行结果的示例代码:
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_ashercn97__manatee-7b", "harness_winogrande_5", split="train")
最新结果
以下是来自 2023-09-17T18:42:42.384089 运行的最新结果:
python { "all": { "em": 0.0030411073825503355, "em_stderr": 0.0005638896908753201, "f1": 0.059899328859060456, "f1_stderr": 0.001397556369094792, "acc": 0.4077875240923591, "acc_stderr": 0.009650175391680019 }, "harness|drop|3": { "em": 0.0030411073825503355, "em_stderr": 0.0005638896908753201, "f1": 0.059899328859060456, "f1_stderr": 0.001397556369094792 }, "harness|gsm8k|5": { "acc": 0.07050796057619409, "acc_stderr": 0.0070515438139836135 }, "harness|winogrande|5": { "acc": 0.745067087608524, "acc_stderr": 0.012248806969376422 } }
配置详情
数据集包含多个配置,每个配置对应不同的任务和运行结果。以下是部分配置的详细信息:
-
harness_arc_challenge_25
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|arc:challenge|25_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-08-02T16:08:56.879142.parquet
-
harness_drop_3
- 分割:2023_09_17T18_42_42.384089
- 路径:
**/details_harness|drop|3_2023-09-17T18-42-42.384089.parquet - 分割:latest
- 路径:
**/details_harness|drop|3_2023-09-17T18-42-42.384089.parquet
-
harness_gsm8k_5
- 分割:2023_09_17T18_42_42.384089
- 路径:
**/details_harness|gsm8k|5_2023-09-17T18-42-42.384089.parquet - 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-09-17T18-42-42.384089.parquet
-
harness_hellaswag_10
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hellaswag|10_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-02T16:08:56.879142.parquet等 - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-02T16:08:56.879142.parquet等
-
harness_hendrycksTest_abstract_algebra_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_anatomy_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_astronomy_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_business_ethics_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_clinical_knowledge_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_college_biology_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_college_chemistry_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_college_computer_science_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_college_mathematics_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_college_medicine_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_college_physics_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_computer_security_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_conceptual_physics_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_econometrics_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_electrical_engineering_5
- 分割:2023_08_02T16_08_56.879142
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-02T16:08:56.879142.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-02T16:08:56.879142.parquet
-
harness_hendrycksTest_elementary_mathematics_5
- 分割:2023_08_02T16_08_56.879142
- 路径:`**/details_harness|hendry



