open-llm-leaderboard-old/details_NotAiLOL__Knight-Miqu-27B-MoE
收藏数据集概述
该数据集是在对模型 NotAiLOL/Knight-Miqu-27B-MoE 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。
数据集组成
- 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" 分割始终指向最新的结果。
- 额外的 "results" 配置存储所有运行结果的聚合,用于计算和显示 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_NotAiLOL__Knight-Miqu-27B-MoE", "harness_winogrande_5", split="train")
最新结果
以下是 2024-04-21T01:28:01.048272 运行的最新结果:
python { "all": { "acc": 0.2690590125333566, "acc_stderr": 0.031219067685780297, "acc_norm": 0.26882939322699473, "acc_norm_stderr": 0.03206876676863873, "mc1": 0.2778457772337821, "mc1_stderr": 0.01568092936402465, "mc2": 0.526959336790919, "mc2_stderr": 0.016725653049287824 }, "harness|arc:challenge|25": { "acc": 0.3199658703071672, "acc_stderr": 0.013631345807016195, "acc_norm": 0.3771331058020478, "acc_norm_stderr": 0.014163366896192589 }, "harness|hellaswag|10": { "acc": 0.30103565026887075, "acc_stderr": 0.004577707025031371, "acc_norm": 0.3741286596295559, "acc_norm_stderr": 0.00482908153282651 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.29, "acc_stderr": 0.04560480215720683, "acc_norm": 0.29, "acc_norm_stderr": 0.04560480215720683 }, ... }
配置详情
-
config_name: harness_arc_challenge_25
- data_files:
- split: 2024_04_21T01_28_01.048272
- **/details_harness|arc:challenge|25_2024-04-21T01-28-01.048272.parquet
- split: latest
- **/details_harness|arc:challenge|25_2024-04-21T01-28-01.048272.parquet
- split: 2024_04_21T01_28_01.048272
- data_files:
-
config_name: harness_gsm8k_5
- data_files:
- split: 2024_04_21T01_28_01.048272
- **/details_harness|gsm8k|5_2024-04-21T01-28-01.048272.parquet
- split: latest
- **/details_harness|gsm8k|5_2024-04-21T01-28-01.048272.parquet
- split: 2024_04_21T01_28_01.048272
- data_files:
-
config_name: harness_hellaswag_10
- data_files:
- split: 2024_04_21T01_28_01.048272
- **/details_harness|hellaswag|10_2024-04-21T01-28-01.048272.parquet
- split: latest
- **/details_harness|hellaswag|10_2024-04-21T01-28-01.048272.parquet
- split: 2024_04_21T01_28_01.048272
- data_files:
-
config_name: harness_hendrycksTest_5
- data_files:
- split: 2024_04_21T01_28_01.048272
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-04-21T01-28-01.048272.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2024-04-21T01-28-01.048272.parquet ... }
- split: 2024_04_21T01_28_01.048272
- data_files:



