open-llm-leaderboard-old/details_vishesht27__22-Neuro_Model
收藏数据集概述
该数据集是在评估模型 vishesht27/22-Neuro_Model 的过程中自动创建的,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:来自1次运行,每个运行在每个配置中作为一个特定的分割存在,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。
- 汇总结果: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 上的聚合指标。
数据加载示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_vishesht27__22-Neuro_Model", "harness_winogrande_5", split="train")
最新结果
以下是 2024-01-10T20:10:57.394152 运行的最新结果:
python { "all": { "acc": 0.605571197032111, "acc_stderr": 0.03282075920315952, "acc_norm": 0.6179788321266033, "acc_norm_stderr": 0.033657408374297766, "mc1": 0.37821297429620565, "mc1_stderr": 0.01697633590754687, "mc2": 0.6022520577190992, "mc2_stderr": 0.016271569580854295 }, "harness|arc:challenge|25": { "acc": 0.46501706484641636, "acc_stderr": 0.01457558392201966, "acc_norm": 0.49146757679180886, "acc_norm_stderr": 0.014609263165632179 }, "harness|hellaswag|10": { "acc": 0.4519020115514838, "acc_stderr": 0.004966640868083856, "acc_norm": 0.6230830511850229, "acc_norm_stderr": 0.0048362341436554305 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.28, "acc_stderr": 0.04512608598542128, "acc_norm": 0.28, "acc_norm_stderr": 0.04512608598542128 }, # 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_01_10T20_10_57.394152- 路径:
**/details_harness|arc:challenge|25_2024-01-10T20-10-57.394152.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|arc:challenge|25_2024-01-10T20-10-57.394152.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_gsm8k_5- 数据文件:
- 分割:
2024_01_10T20_10_57.394152- 路径:
**/details_harness|gsm8k|5_2024-01-10T20-10-57.394152.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|gsm8k|5_2024-01-10T20-10-57.394152.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_hellaswag_10- 数据文件:
- 分割:
2024_01_10T20_10_57.394152- 路径:
**/details_harness|hellaswag|10_2024-01-10T20-10-57.394152.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|hellaswag|10_2024-01-10T20-10-57.394152.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_hendrycksTest_5- 数据文件:
- 分割:
2024_01_10T20_10_57.394152- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-anatomy|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-astronomy|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-college_biology|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-college_physics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-computer_security|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-econometrics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-global_facts|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_geography|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_physics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-human_aging|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-human_sexuality|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-international_law|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-jurisprudence|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-machine_learning|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-management|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-marketing|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-medical_genetics|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-miscellaneous|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-moral_disputes|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-nutrition|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-philosophy|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-prehistory|5_2024-01-10T20-10-57.394152.parquet**/details_harness|hendrycksTest-professional_accounting|5_2024-01-10T20-10-57.394152.parquet
- 路径:
- 分割:
- 数据文件:



