open-llm-leaderboard-old/details_azale-ai__DukunLM-7B-V1.0-Uncensored
收藏Hugging Face2024-02-04 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_azale-ai__DukunLM-7B-V1.0-Uncensored
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是在评估模型azale-ai/DukunLM-7B-V1.0-Uncensored时自动创建的,包含63个配置,每个配置对应一个评估任务。数据集由1次运行生成,每次运行的结果作为特定配置中的一个分割,分割名称使用运行的时间戳。train分割始终指向最新结果。此外,results配置存储了所有运行的聚合结果,用于在Open LLM Leaderboard上计算和显示聚合指标。
该数据集是在评估模型azale-ai/DukunLM-7B-V1.0-Uncensored时自动创建的,包含63个配置,每个配置对应一个评估任务。数据集由1次运行生成,每次运行的结果作为特定配置中的一个分割,分割名称使用运行的时间戳。train分割始终指向最新结果。此外,results配置存储了所有运行的聚合结果,用于在Open LLM Leaderboard上计算和显示聚合指标。
提供机构:
open-llm-leaderboard-old
原始信息汇总
数据集概述
数据集摘要
该数据集是在评估模型 azale-ai/DukunLM-7B-V1.0-Uncensored 在 Open LLM Leaderboard 上的自动创建的。数据集包含 63 个配置,每个配置对应一个评估任务。数据集从 1 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。
数据集结构
数据集包含多个配置,每个配置对应不同的评估任务。以下是部分配置的示例:
-
harness_arc_challenge_25
- 分割:2024_02_04T03_32_00.345040
- 路径:
**/details_harness|arc:challenge|25_2024-02-04T03-32-00.345040.parquet - 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2024-02-04T03-32-00.345040.parquet
-
harness_gsm8k_5
- 分割:2024_02_04T03_32_00.345040
- 路径:
**/details_harness|gsm8k|5_2024-02-04T03-32-00.345040.parquet - 分割:latest
- 路径:
**/details_harness|gsm8k|5_2024-02-04T03-32-00.345040.parquet
-
harness_hellaswag_10
- 分割:2024_02_04T03_32_00.345040
- 路径:
**/details_harness|hellaswag|10_2024-02-04T03-32-00.345040.parquet - 分割:latest
- 路径:
**/details_harness|hellaswag|10_2024-02-04T03-32-00.345040.parquet
-
harness_hendrycksTest_5
- 分割:2024_02_04T03_32_00.345040
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-anatomy|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-astronomy|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-college_biology|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-college_physics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-computer_security|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-econometrics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-global_facts|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_geography|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_physics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-human_aging|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-human_sexuality|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-international_law|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-jurisprudence|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-machine_learning|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-management|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-marketing|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-medical_genetics|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-miscellaneous|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-moral_disputes|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-nutrition|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-philosophy|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-prehistory|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-professional_accounting|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-professional_law|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-professional_medicine|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-professional_psychology|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-public_relations|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-security_studies|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-sociology|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-us_foreign_policy|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-virology|5_2024-02-04T03-32-00.345040.parquet**/details_harness|hendrycksTest-world_religions|5_2024-02-04T03-32-00.345040.parquet
最新结果
以下是 2024-02-04T03:32:00.345040 运行 的最新结果:
python { "all": { "acc": 0.4018543967795034, "acc_stderr": 0.0342455004465676, "acc_norm": 0.4061924250003242, "acc_norm_stderr": 0.03506469624535442, "mc1": 0.2998776009791922, "mc1_stderr": 0.016040352966713634, "mc2": 0.43947585501681957, "mc2_stderr": 0.015779310526247342 }, "harness|arc:challenge|25": { "acc": 0.4854948805460751, "acc_stderr": 0.0146052410813



