open-llm-leaderboard-old/details_rwitz__experiment26-truthy-iter-1
收藏数据集概述
该数据集是在对模型 rwitz/experiment26-truthy-iter-1 进行评估运行期间自动创建的。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,每次运行的详细结果可以在每个配置中找到,使用运行的时间戳作为分割名称。"train" 分割始终指向最新的结果。
数据集结构
配置详情
-
配置名称:
harness_arc_challenge_25- 数据文件:
- 分割:
2024_03_12T05_53_31.833751- 路径:
**/details_harness|arc:challenge|25_2024-03-12T05-53-31.833751.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|arc:challenge|25_2024-03-12T05-53-31.833751.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_gsm8k_5- 数据文件:
- 分割:
2024_03_12T05_53_31.833751- 路径:
**/details_harness|gsm8k|5_2024-03-12T05-53-31.833751.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|gsm8k|5_2024-03-12T05-53-31.833751.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_hellaswag_10- 数据文件:
- 分割:
2024_03_12T05_53_31.833751- 路径:
**/details_harness|hellaswag|10_2024-03-12T05-53-31.833751.parquet
- 路径:
- 分割:
latest- 路径:
**/details_harness|hellaswag|10_2024-03-12T05-53-31.833751.parquet
- 路径:
- 分割:
- 数据文件:
-
配置名称:
harness_hendrycksTest_5- 数据文件:
- 分割:
2024_03_12T05_53_31.833751- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-anatomy|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-astronomy|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-college_biology|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-college_physics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-computer_security|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-econometrics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-global_facts|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_geography|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_physics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-human_aging|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-human_sexuality|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-international_law|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-jurisprudence|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-machine_learning|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-management|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-marketing|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-medical_genetics|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-miscellaneous|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-moral_disputes|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-nutrition|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-philosophy|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-prehistory|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-professional_accounting|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-professional_law|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-professional_medicine|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-professional_psychology|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-public_relations|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-security_studies|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-sociology|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-us_foreign_policy|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-virology|5_2024-03-12T05-53-31.833751.parquet**/details_harness|hendrycksTest-world_religions|5_2024-03-12T05-53-31.833751.parquet
- 路径:
- 分割:
- 数据文件:
最新结果
以下是2024年3月12日 05:53:31.833751运行的最新结果:
python { "all": { "acc": 0.650096854091544, "acc_stderr": 0.032077230975316916, "acc_norm": 0.6491265947954293, "acc_norm_stderr": 0.03275212683346799, "mc1": 0.627906976744186, "mc1_stderr": 0.01692109011881403, "mc2": 0.776556754585214, "mc2_stderr": 0.013758880328500731 }, "harness|arc:challenge|25": { "acc": 0.71160409556314, "acc_stderr": 0.013238394422428173, "acc_norm": 0.7320819112627986, "acc_norm_stderr": 0.012942030195136438 }, "harness|hellaswag|10": { "acc": 0.716391157140012, "acc_stderr": 0.00449



