open-llm-leaderboard-old/details_Lazycuber__L2-7b-Guanaco-Uncensored
收藏数据集概述
数据集摘要
该数据集是在评估模型 Lazycuber/L2-7b-Guanaco-Uncensored 在 Open LLM Leaderboard 上的自动创建的。
数据集组成
- 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。
- 一个额外的配置 "results" 存储所有运行结果的聚合,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_Lazycuber__L2-7b-Guanaco-Uncensored", "harness_winogrande_5", split="train")
最新结果
这些是最新结果(来自 2023-10-25T07:21:20.584203 运行)的示例: python { "all": { "em": 0.0014681208053691276, "em_stderr": 0.00039210421902985767, "f1": 0.056828859060402796, "f1_stderr": 0.0013179206618636607, "acc": 0.416677387679193, "acc_stderr": 0.0097821448230569 }, "harness|drop|3": { "em": 0.0014681208053691276, "em_stderr": 0.00039210421902985767, "f1": 0.056828859060402796, "f1_stderr": 0.0013179206618636607 }, "harness|gsm8k|5": { "acc": 0.07960576194086429, "acc_stderr": 0.007455924338676278 }, "harness|winogrande|5": { "acc": 0.7537490134175217, "acc_stderr": 0.012108365307437523 } }
配置详情
-
harness_arc_challenge_25
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|arc:challenge|25_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-09-18T14-24-41.596109.parquet
-
harness_drop_3
- 分割:2023_10_25T07_21_20.584203
- 路径:
**/details_harness|drop|3_2023-10-25T07-21-20.584203.parquet - 分割:latest
- 路径:
**/details_harness|drop|3_2023-10-25T07-21-20.584203.parquet
-
harness_gsm8k_5
- 分割:2023_10_25T07_21_20.584203
- 路径:
**/details_harness|gsm8k|5_2023-10-25T07-21-20.584203.parquet - 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-10-25T07-21-20.584203.parquet
-
harness_hellaswag_10
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hellaswag|10_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-18T14-24-41.596109.parquet等 - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-18T14-24-41.596109.parquet等
-
harness_hendrycksTest_abstract_algebra_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_anatomy_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_astronomy_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_business_ethics_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_clinical_knowledge_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_college_biology_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_college_chemistry_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_college_computer_science_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_college_mathematics_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_college_medicine_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_college_physics_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_computer_security_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_conceptual_physics_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_econometrics_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-09-18T14-24-41.596109.parquet
-
harness_hendrycksTest_electrical_engineering_5
- 分割:2023_09_18T14_24_41.596109
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-18T14-24-41.596109.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-18T14-24-41.596109.parquet
-
**harness_hendrycksTest_elementary_mathematics



