open-llm-leaderboard-old/details_HanNayeoniee__LHK_DPO_v1
收藏数据集概述
该数据集是在对模型 HanNayeoniee/LHK_DPO_v1 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 2 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。
- 一个额外的配置 "results" 存储所有运行的聚合结果,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_HanNayeoniee__LHK_DPO_v1", "harness_winogrande_5", split="train")
最新结果
以下是 最新结果 的摘要:
python { "all": { "acc": 0.6558324843188587, "acc_stderr": 0.032073971899271705, "acc_norm": 0.6547598388796718, "acc_norm_stderr": 0.03276116533502233, "mc1": 0.6548347613219094, "mc1_stderr": 0.016643103319274943, "mc2": 0.7989231486574115, "mc2_stderr": 0.013454899328675057 }, "harness|arc:challenge|25": { "acc": 0.7201365187713311, "acc_stderr": 0.013119040897725923, "acc_norm": 0.7474402730375427, "acc_norm_stderr": 0.012696728980207702 }, "harness|hellaswag|10": { "acc": 0.7267476598287194, "acc_stderr": 0.004447185883327433, "acc_norm": 0.8930491933877713, "acc_norm_stderr": 0.0030841908180933085 }, # 其他任务的结果... }
配置详情
-
harness_arc_challenge_25
- 分割:2024_02_09T15_00_21.741552, 2024_02_12T16_42_16.638684, latest
- 路径:
**/details_harness|arc:challenge|25_2024-02-09T15-00-21.741552.parquet,**/details_harness|arc:challenge|25_2024-02-12T16-42-16.638684.parquet
-
harness_gsm8k_5
- 分割:2024_02_09T15_00_21.741552, 2024_02_12T16_42_16.638684, latest
- 路径:
**/details_harness|gsm8k|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|gsm8k|5_2024-02-12T16-42-16.638684.parquet
-
harness_hellaswag_10
- 分割:2024_02_09T15_00_21.741552, 2024_02_12T16_42_16.638684, latest
- 路径:
**/details_harness|hellaswag|10_2024-02-09T15-00-21.741552.parquet,**/details_harness|hellaswag|10_2024-02-12T16-42-16.638684.parquet
-
harness_hendrycksTest_5
- 分割:2024_02_09T15_00_21.741552, 2024_02_12T16_42_16.638684, latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-anatomy|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-astronomy|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-business_ethics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-college_biology|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-college_chemistry|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-college_computer_science|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-college_mathematics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-college_medicine|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-college_physics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-computer_security|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-conceptual_physics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-econometrics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-electrical_engineering|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-formal_logic|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-global_facts|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_biology|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_chemistry|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_computer_science|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_european_history|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_geography|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_mathematics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_physics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_psychology|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_statistics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_us_history|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-high_school_world_history|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-human_aging|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-human_sexuality|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-international_law|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-jurisprudence|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-logical_fallacies|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-machine_learning|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-management|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-marketing|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-medical_genetics|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-miscellaneous|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-moral_disputes|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-moral_scenarios|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-nutrition|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-philosophy|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-prehistory|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-professional_accounting|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-professional_law|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-professional_medicine|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-professional_psychology|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-public_relations|5_2024-02-09T15-00-21.741552.parquet,**/details_harness|hendrycksTest-security_studies|5_2024-02-09T15-00-21.741552.parquet, `**/



