open-llm-leaderboard-old/details_dvruette__llama-13b-pretrained-dropout
收藏数据集概述
该数据集是在评估模型 dvruette/llama-13b-pretrained-dropout 在 Open LLM Leaderboard 上的自动创建的。数据集包含 64 个配置,每个配置对应一个评估任务。
数据集结构
- 配置数量:64 个配置
- 数据来源:从 2 次运行中创建,每次运行对应一个特定的时间戳分割。
- 最新结果:"train" 分割始终指向最新结果。
- 结果汇总:一个额外的配置 "results" 存储所有运行结果的汇总,用于计算和显示在 Open LLM Leaderboard 上的汇总指标。
数据加载示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_dvruette__llama-13b-pretrained-dropout", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-18T13:29:36.249394 运行 的最新结果:
python { "all": { "em": 0.20501258389261745, "em_stderr": 0.0041343766395959035, "f1": 0.2702611157718119, "f1_stderr": 0.004144727885990915, "acc": 0.43522094959648105, "acc_stderr": 0.01051473093615015 }, "harness|drop|3": { "em": 0.20501258389261745, "em_stderr": 0.0041343766395959035, "f1": 0.2702611157718119, "f1_stderr": 0.004144727885990915 }, "harness|gsm8k|5": { "acc": 0.11827141774071266, "acc_stderr": 0.008895075852434953 }, "harness|winogrande|5": { "acc": 0.7521704814522494, "acc_stderr": 0.01213438601986535 } }
配置详情
-
harness_arc_challenge_25
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|arc:challenge|25_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-07-19T19:40:51.054216.parquet
-
harness_drop_3
- 分割:2023_10_18T13_29_36.249394
- 路径:
**/details_harness|drop|3_2023-10-18T13-29-36.249394.parquet - 分割:latest
- 路径:
**/details_harness|drop|3_2023-10-18T13-29-36.249394.parquet
-
harness_gsm8k_5
- 分割:2023_10_18T13_29_36.249394
- 路径:
**/details_harness|gsm8k|5_2023-10-18T13-29-36.249394.parquet - 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-10-18T13-29-36.249394.parquet
-
harness_hellaswag_10
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hellaswag|10_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_5
- 分割:2023_07_19T19_40_51.054216
- 路径:多个文件路径
- 分割:latest
- 路径:多个文件路径
-
harness_hendrycksTest_abstract_algebra_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_anatomy_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_astronomy_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_business_ethics_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_clinical_knowledge_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_college_biology_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_college_chemistry_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_college_computer_science_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_college_mathematics_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_college_medicine_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_college_physics_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_computer_security_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_conceptual_physics_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_econometrics_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_electrical_engineering_5
- 分割:2023_07_19T19_40_51.054216
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T19:40:51.054216.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T19:40:51.054216.parquet
-
harness_hendrycksTest_elementary_mathematics_5
- 分割:2023_07_19T19_40_51.



