open-llm-leaderboard-old/details_HuggingFaceH4__zephyr-7b-alpha_private
收藏数据集概述
数据集摘要
该数据集是在对模型 HuggingFaceH4/zephyr-7b-alpha 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 61 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" 分割始终指向最新的结果。
- 额外的 "results" 配置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。
最新结果
以下是 2023-10-10T14:11:13.991325 运行的最新结果:
python { "all": { "acc": 0.6137978230566867, "acc_stderr": 0.03380754595328641, "acc_norm": 0.6176702382672306, "acc_norm_stderr": 0.03378555360789072, "mc1": 0.42717258261933905, "mc1_stderr": 0.017316834410963926, "mc2": 0.5790339154881958, "mc2_stderr": 0.015362629183533977 }, "harness|arc:challenge|25": { "acc": 0.5810580204778157, "acc_stderr": 0.014418106953639011, "acc_norm": 0.6100682593856656, "acc_norm_stderr": 0.01425295984889289 }, "harness|hellaswag|10": { "acc": 0.6409081856203943, "acc_stderr": 0.004787537385153006, "acc_norm": 0.8403704441346346, "acc_norm_stderr": 0.0036551361115537096 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.32, "acc_stderr": 0.046882617226215034, "acc_norm": 0.32, "acc_norm_stderr": 0.046882617226215034 }, # 其他任务的结果省略... "harness|truthfulqa:mc|0": { "mc1": 0.42717258261933905, "mc1_stderr": 0.017316834410963926, "mc2": 0.5790339154881958, "mc2_stderr": 0.015362629183533977 } }
配置详情
-
config_name: harness_arc_challenge_25
- split: 2023_10_10T14_11_13.991325
- path:
**/details_harness|arc:challenge|25_2023-10-10T14-11-13.991325.parquet
- path:
- split: latest
- path:
**/details_harness|arc:challenge|25_2023-10-10T14-11-13.991325.parquet
- path:
- split: 2023_10_10T14_11_13.991325
-
config_name: harness_hellaswag_10
- split: 2023_10_10T14_11_13.991325
- path:
**/details_harness|hellaswag|10_2023-10-10T14-11-13.991325.parquet
- path:
- split: latest
- path:
**/details_harness|hellaswag|10_2023-10-10T14-11-13.991325.parquet
- path:
- split: 2023_10_10T14_11_13.991325
-
config_name: harness_hendrycksTest_5
- split: 2023_10_10T14_11_13.991325
- path:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-anatomy|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-astronomy|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-business_ethics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-college_biology|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-college_chemistry|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-college_computer_science|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-college_mathematics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-college_medicine|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-college_physics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-computer_security|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-econometrics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-formal_logic|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-global_facts|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_biology|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_geography|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_physics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-human_aging|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-human_sexuality|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-international_law|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-jurisprudence|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-machine_learning|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-management|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-marketing|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-medical_genetics|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-miscellaneous|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-moral_disputes|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-nutrition|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-philosophy|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-prehistory|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-professional_accounting|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-professional_law|5_2023-10-10T14-11-13.991325.parquet**/details_harness|hendrycksTest-professional_medicine|5_2023-10-10T14-11-13.991325.parquet- `**/details_harness|hendrycksTest-professional_psychology|5_2023-10
- path:
- split: 2023_10_10T14_11_13.991325



