open-llm-leaderboard-old/details_vicgalle__NeuralBeagle-11B
收藏数据集概述
数据集摘要
该数据集是在对模型 vicgalle/NeuralBeagle-11B 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。"train" 分割始终指向最新结果。
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。以下是部分配置的详细信息:
配置示例
-
config_name: harness_arc_challenge_25
- data_files:
- split: 2024_01_21T05_36_08.681056
- path: **/details_harness|arc:challenge|25_2024-01-21T05-36-08.681056.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2024-01-21T05-36-08.681056.parquet
- split: 2024_01_21T05_36_08.681056
- data_files:
-
config_name: harness_gsm8k_5
- data_files:
- split: 2024_01_21T05_36_08.681056
- path: **/details_harness|gsm8k|5_2024-01-21T05-36-08.681056.parquet
- split: latest
- path: **/details_harness|gsm8k|5_2024-01-21T05-36-08.681056.parquet
- split: 2024_01_21T05_36_08.681056
- data_files:
-
config_name: harness_hellaswag_10
- data_files:
- split: 2024_01_21T05_36_08.681056
- path: **/details_harness|hellaswag|10_2024-01-21T05-36-08.681056.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2024-01-21T05-36-08.681056.parquet
- split: 2024_01_21T05_36_08.681056
- data_files:
-
config_name: harness_hendrycksTest_5
- data_files:
- split: 2024_01_21T05_36_08.681056
- path:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_computer_science|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_european_history|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_geography|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_mathematics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_microeconomics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_physics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_psychology|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_statistics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_us_history|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-high_school_world_history|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-human_aging|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-human_sexuality|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-international_law|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-jurisprudence|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-logical_fallacies|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-machine_learning|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-management|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-marketing|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-medical_genetics|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-miscellaneous|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-moral_disputes|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-moral_scenarios|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-nutrition|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-philosophy|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-prehistory|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-professional_accounting|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-professional_law|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-professional_medicine|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-professional_psychology|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-public_relations|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-security_studies|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-sociology|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-us_foreign_policy|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-virology|5_2024-01-21T05-36-08.681056.parquet
- **/details_harness|hendrycksTest-world_religions|5_2024-01-21T05-36-08.681056.parquet
- path:
- split: 2024_01_21T05_36_08.681056
- data_files:
最新结果
以下是 2024-01-21T05:36:08.681056 运行 的最新结果:
python { "all": { "acc": 0.6427615979936613, "acc_stderr": 0.032364857864304215, "acc_norm": 0.6436873631017901, "acc_norm_stderr": 0.03302566377855622, "mc1": 0.587515299877601, "mc1_stderr": 0.017233299399571207, "mc2": 0.7135848334628236, "mc2_stderr": 0.015017949998169619 }, "harness|arc:challenge|25": { "acc": 0.710750853242



