open-llm-leaderboard-old/details_stabilityai__stablelm-2-zephyr-1_6b
收藏数据集概述
数据集摘要
该数据集是在对模型 stabilityai/stablelm-2-zephyr-1_6b 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。数据集包含 63 个配置,每个配置对应一个评估任务。
数据集结构
- 配置数量:63 个配置
- 数据来源:从 1 次运行中创建,每个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。
- 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" 存储所有运行的汇总结果,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_stabilityai__stablelm-2-zephyr-1_6b", "harness_winogrande_5", split="train")
最新结果
以下是来自 2024-03-01T08:52:40.899023 运行的最新结果: python { "all": { "acc": 0.42474242460022255, "acc_stderr": 0.03446072937977788, "acc_norm": 0.42524874427823406, "acc_norm_stderr": 0.035172374705053376, "mc1": 0.30599755201958384, "mc1_stderr": 0.016132229728155048, "mc2": 0.4511289179951579, "mc2_stderr": 0.015760909140053084 }, "harness|arc:challenge|25": { "acc": 0.42235494880546076, "acc_stderr": 0.014434138713379977, "acc_norm": 0.43686006825938567, "acc_norm_stderr": 0.014494421584256515 }, "harness|hellaswag|10": { "acc": 0.5289782911770564, "acc_stderr": 0.004981394110706144, "acc_norm": 0.692989444333798, "acc_norm_stderr": 0.004603111343213064 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.34, "acc_stderr": 0.04760952285695236, "acc_norm": 0.34, "acc_norm_stderr": 0.04760952285695236 }, # 其他任务的结果... }
配置详情
-
配置名称:
harness_arc_challenge_25- 数据文件:
- 分割:
2024_03_01T08_52_40.899023- 路径:
**/details_harness|arc:challenge|25_2024-03-01T08-52-40.899023.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|arc:challenge|25_2024-03-01T08-52-40.899023.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_gsm8k_5- 数据文件:
- 分割:
2024_03_01T08_52_40.899023- 路径:
**/details_harness|gsm8k|5_2024-03-01T08-52-40.899023.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|gsm8k|5_2024-03-01T08-52-40.899023.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_hellaswag_10- 数据文件:
- 分割:
2024_03_01T08_52_40.899023- 路径:
**/details_harness|hellaswag|10_2024-03-01T08-52-40.899023.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|hellaswag|10_2024-03-01T08-52-40.899023.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_hendrycksTest_5- 数据文件:
- 分割:
2024_03_01T08_52_40.899023- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-anatomy|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-astronomy|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-college_biology|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-college_physics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-computer_security|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-econometrics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-global_facts|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_geography|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_physics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-human_aging|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-human_sexuality|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-international_law|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-jurisprudence|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-machine_learning|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-management|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-marketing|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-medical_genetics|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-miscellaneous|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-moral_disputes|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-nutrition|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-philosophy|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-prehistory|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-professional_accounting|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-professional_law|5_2024-03-01T08-52-40.899023.parquet**/details_harness|hendrycksTest-professional_medicine|5_2024-03-01T08-52-40.899023.parquet- `**/
- 路径:
- 分割:
- 数据文件:



