open-llm-leaderboard/details_openbmb__UltraLM-65b
收藏数据集概述
数据集简介
该数据集是在评估模型openbmb/UltraLM-65b在Open LLM Leaderboard上的运行过程中自动创建的。数据集包含64个配置,每个配置对应一个评估任务。
数据集结构
数据集由3次运行结果组成,每次运行的结果可以在每个配置中找到,并以运行的时间戳命名。"train"分割总是指向最新的结果。此外,一个额外的配置"results"存储了所有运行的聚合结果,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_openbmb__UltraLM-65b", "harness_winogrande_5", split="train")
最新结果
以下是2023-09-23T05:14:21.286059运行的最新结果: python { "all": { "em": 0.23804530201342283, "em_stderr": 0.004361481495925771, "f1": 0.2999853187919465, "f1_stderr": 0.004304795126990332, "acc": 0.5694431396390439, "acc_stderr": 0.011961137264223144 }, "harness|drop|3": { "em": 0.23804530201342283, "em_stderr": 0.004361481495925771, "f1": 0.2999853187919465, "f1_stderr": 0.004304795126990332 }, "harness|gsm8k|5": { "acc": 0.32752084912812734, "acc_stderr": 0.012927102210426474 }, "harness|winogrande|5": { "acc": 0.8113654301499605, "acc_stderr": 0.010995172318019811 } }
配置详情
-
config_name: harness_arc_challenge_25
- split: 2023_08_04T22_09_07.792369
- path: **/details_harness|arc:challenge|25_2023-08-04T22:09:07.792369.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2023-08-04T22:09:07.792369.parquet
- split: 2023_08_04T22_09_07.792369
-
config_name: harness_drop_3
- split: 2023_09_18T23_27_44.207127
- path: **/details_harness|drop|3_2023-09-18T23-27-44.207127.parquet
- split: 2023_09_23T05_14_21.286059
- path: **/details_harness|drop|3_2023-09-23T05-14-21.286059.parquet
- split: latest
- path: **/details_harness|drop|3_2023-09-23T05-14-21.286059.parquet
- split: 2023_09_18T23_27_44.207127
-
config_name: harness_gsm8k_5
- split: 2023_09_18T23_27_44.207127
- path: **/details_harness|gsm8k|5_2023-09-18T23-27-44.207127.parquet
- split: 2023_09_23T05_14_21.286059
- path: **/details_harness|gsm8k|5_2023-09-23T05-14-21.286059.parquet
- split: latest
- path: **/details_harness|gsm8k|5_2023-09-23T05-14-21.286059.parquet
- split: 2023_09_18T23_27_44.207127
-
config_name: harness_hellaswag_10
- split: 2023_08_04T22_09_07.792369
- path: **/details_harness|hellaswag|10_2023-08-04T22:09:07.792369.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2023-08-04T22:09:07.792369.parquet
- split: 2023_08_04T22_09_07.792369
-
config_name: harness_hendrycksTest_5
- split: 2023_08_04T22_09_07.792369
- path:
- **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-anatomy|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-astronomy|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-college_biology|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-college_physics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-computer_security|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-econometrics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-global_facts|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_geography|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_physics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-human_aging|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-human_sexuality|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-international_law|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-jurisprudence|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-machine_learning|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-management|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-marketing|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-medical_genetics|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-miscellaneous|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-moral_disputes|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-nutrition|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-philosophy|5_2023-08-04T22:09:07.792369.parquet
- **/details_harness|hendrycksTest-prehistory|5_2023-08-04T22:09:07.
- path:
- split: 2023_08_04T22_09_07.792369



