open-llm-leaderboard/details_gaodrew__gaodrew-gorgonzola-13b
收藏数据集概述
该数据集是在评估模型 gaodrew/gaodrew-gorgonzola-13b 在 Open LLM Leaderboard 上的自动创建的。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。
结果配置
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_gaodrew__gaodrew-gorgonzola-13b", "harness_winogrande_5", split="train")
最新结果
以下是 2023-09-23T16:30:20.571069 运行的最新结果:
python { "all": { "em": 0.04121224832214765, "em_stderr": 0.0020357012531483946, "f1": 0.13222525167785196, "f1_stderr": 0.002559817666324549, "acc": 0.4265177812231289, "acc_stderr": 0.010193838735770604 }, "harness|drop|3": { "em": 0.04121224832214765, "em_stderr": 0.0020357012531483946, "f1": 0.13222525167785196, "f1_stderr": 0.002559817666324549 }, "harness|gsm8k|5": { "acc": 0.10007581501137225, "acc_stderr": 0.008266274528685634 }, "harness|winogrande|5": { "acc": 0.7529597474348856, "acc_stderr": 0.012121402942855573 } }
配置详情
-
config_name: harness_arc_challenge_25
- split: 2023_08_23T21_07_18.787653
- path: **/details_harness|arc:challenge|25_2023-08-23T21:07:18.787653.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2023-08-23T21:07:18.787653.parquet
- split: 2023_08_23T21_07_18.787653
-
config_name: harness_drop_3
- split: 2023_09_21T04_07_38.110729
- path: **/details_harness|drop|3_2023-09-21T04-07-38.110729.parquet
- split: 2023_09_23T16_30_20.571069
- path: **/details_harness|drop|3_2023-09-23T16-30-20.571069.parquet
- split: latest
- path: **/details_harness|drop|3_2023-09-23T16-30-20.571069.parquet
- split: 2023_09_21T04_07_38.110729
-
config_name: harness_gsm8k_5
- split: 2023_09_21T04_07_38.110729
- path: **/details_harness|gsm8k|5_2023-09-21T04-07-38.110729.parquet
- split: 2023_09_23T16_30_20.571069
- path: **/details_harness|gsm8k|5_2023-09-23T16-30-20.571069.parquet
- split: latest
- path: **/details_harness|gsm8k|5_2023-09-23T16-30-20.571069.parquet
- split: 2023_09_21T04_07_38.110729
-
config_name: harness_hellaswag_10
- split: 2023_08_23T21_07_18.787653
- path: **/details_harness|hellaswag|10_2023-08-23T21:07:18.787653.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2023-08-23T21:07:18.787653.parquet
- split: 2023_08_23T21_07_18.787653
-
config_name: harness_hendrycksTest_5
- split: 2023_08_23T21_07_18.787653
- path:
- **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-anatomy|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-astronomy|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-college_biology|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-college_physics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-computer_security|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-econometrics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-global_facts|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_geography|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_physics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-human_aging|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-human_sexuality|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-international_law|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-jurisprudence|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-machine_learning|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-management|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-marketing|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-medical_genetics|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-miscellaneous|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-moral_disputes|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-nutrition|5_2023-08-23T21:07:18.787653.parquet
- **/details_harness|hendrycksTest-philosophy|5_2023-08-23T21:07:18.787653.parquet
- path:
- split: 2023_08_23T21_07_18.787653



