open-llm-leaderboard-old/details_NeuralNovel__Ignis-7B-DPO
收藏数据集概述
数据集简介
该数据集是在评估模型NeuralNovel/Ignis-7B-DPO在Open LLM Leaderboard上的运行过程中自动创建的。
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,每个运行可以在每个配置中作为一个特定的分割找到,分割名称使用运行的时间戳。"train"分割始终指向最新的结果。
额外配置
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_NeuralNovel__Ignis-7B-DPO", "harness_winogrande_5", split="train")
最新结果
这些是最新结果的示例: python { "all": { "acc": 0.591254836807939, "acc_stderr": 0.0334070366622428, "acc_norm": 0.5955459854243185, "acc_norm_stderr": 0.03410660590947557, "mc1": 0.4785801713586291, "mc1_stderr": 0.01748743214471181, "mc2": 0.6545579637974486, "mc2_stderr": 0.01605098507055363 }, "harness|arc:challenge|25": { "acc": 0.6390784982935154, "acc_stderr": 0.014034761386175456, "acc_norm": 0.6629692832764505, "acc_norm_stderr": 0.013813476652902276 }, "harness|hellaswag|10": { "acc": 0.6716789484166501, "acc_stderr": 0.0046864258512532815, "acc_norm": 0.8485361481776539, "acc_norm_stderr": 0.0035776774950640805 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.3, "acc_stderr": 0.046056618647183814, "acc_norm": 0.3, "acc_norm_stderr": 0.046056618647183814 }, # 其他任务的结果... }
配置详情
-
config_name: harness_arc_challenge_25
- 数据文件:
- 分割: 2024_02_29T12_08_32.687568
- 路径: **/details_harness|arc:challenge|25_2024-02-29T12-08-32.687568.parquet
- 分割: latest
- 路径: **/details_harness|arc:challenge|25_2024-02-29T12-08-32.687568.parquet
- 分割: 2024_02_29T12_08_32.687568
- 数据文件:
-
config_name: harness_gsm8k_5
- 数据文件:
- 分割: 2024_02_29T12_08_32.687568
- 路径: **/details_harness|gsm8k|5_2024-02-29T12-08-32.687568.parquet
- 分割: latest
- 路径: **/details_harness|gsm8k|5_2024-02-29T12-08-32.687568.parquet
- 分割: 2024_02_29T12_08_32.687568
- 数据文件:
-
config_name: harness_hellaswag_10
- 数据文件:
- 分割: 2024_02_29T12_08_32.687568
- 路径: **/details_harness|hellaswag|10_2024-02-29T12-08-32.687568.parquet
- 分割: latest
- 路径: **/details_harness|hellaswag|10_2024-02-29T12-08-32.687568.parquet
- 分割: 2024_02_29T12_08_32.687568
- 数据文件:
-
config_name: harness_hendrycksTest_5
- 数据文件:
- 分割: 2024_02_29T12_08_32.687568
- 路径:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_computer_science|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_european_history|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_geography|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_mathematics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_microeconomics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_physics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_psychology|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_statistics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_us_history|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-high_school_world_history|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-human_aging|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-human_sexuality|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-international_law|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-jurisprudence|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-logical_fallacies|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-machine_learning|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-management|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-marketing|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-medical_genetics|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-miscellaneous|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-moral_disputes|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-moral_scenarios|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-nutrition|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-philosophy|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-prehistory|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-professional_accounting|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-professional_law|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-professional_medicine|5_2024-02-29T12-08-32.687568.parquet
- **/details_harness|hendrycksTest-professional_psychology|5_
- 路径:
- 分割: 2024_02_29T12_08_32.687568
- 数据文件:



