open-llm-leaderboard/details_Undi95__ReMM-L2-13B-PIPPA
收藏数据集概述
数据集来源
该数据集是在评估模型 Undi95/ReMM-L2-13B-PIPPA 在 Open LLM Leaderboard 上的自动创建的。
数据集结构
- 配置数量:64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" 分割总是指向最新的结果。
- 额外配置: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_Undi95__ReMM-L2-13B-PIPPA", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-15T22:47:55.884527 运行的最新结果: python { "all": { "em": 0.3598993288590604, "em_stderr": 0.004915348455608255, "f1": 0.4368917785234919, "f1_stderr": 0.004726186762311207, "acc": 0.3873174710218511, "acc_stderr": 0.008457350051798611 }, "harness|drop|3": { "em": 0.3598993288590604, "em_stderr": 0.004915348455608255, "f1": 0.4368917785234919, "f1_stderr": 0.004726186762311207 }, "harness|gsm8k|5": { "acc": 0.029567854435178165, "acc_stderr": 0.004665893134220799 }, "harness|winogrande|5": { "acc": 0.745067087608524, "acc_stderr": 0.012248806969376422 } }
配置详情
-
harness_arc_challenge_25
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|arc:challenge|25_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-09-05T05:29:49.738166.parquet
-
harness_drop_3
- 分割:2023_10_15T22_47_55.884527
- 路径:
**/details_harness|drop|3_2023-10-15T22-47-55.884527.parquet - 分割:latest
- 路径:
**/details_harness|drop|3_2023-10-15T22-47-55.884527.parquet
-
harness_gsm8k_5
- 分割:2023_10_15T22_47_55.884527
- 路径:
**/details_harness|gsm8k|5_2023-10-15T22-47-55.884527.parquet - 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-10-15T22-47-55.884527.parquet
-
harness_hellaswag_10
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hellaswag|10_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-05T05:29:49.738166.parquet等36个文件 - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-05T05:29:49.738166.parquet等36个文件
-
harness_hendrycksTest_abstract_algebra_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_anatomy_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_astronomy_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_business_ethics_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_clinical_knowledge_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_college_biology_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_college_chemistry_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_college_computer_science_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_college_mathematics_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_college_medicine_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_college_physics_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_computer_security_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_conceptual_physics_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_econometrics_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-09-05T05:29:49.738166.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-09-05T05:29:49.738166.parquet
-
harness_hendrycksTest_electrical_engineering_5
- 分割:2023_09_05T05_29_49.738166
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-05T05:29:49.738166.parquet - 分割:latest



