open-llm-leaderboard-old/details_Technoculture__MT7Bi-alpha-dpo-v0.2
收藏数据集概述
该数据集是在对模型Technoculture/MT7Bi-alpha-dpo-v0.2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_Technoculture__MT7Bi-alpha-dpo-v0.2", "harness_winogrande_5", split="train")
最新结果
以下是2024-02-09T12:50:13.790724运行的最新结果:
python { "all": { "acc": 0.5274798937736077, "acc_stderr": 0.034244329313021585, "acc_norm": 0.5324573781856667, "acc_norm_stderr": 0.034973478659411146, "mc1": 0.29008567931456547, "mc1_stderr": 0.01588623687420952, "mc2": 0.45484028768936574, "mc2_stderr": 0.015178684073869702 }, "harness|arc:challenge|25": { "acc": 0.5204778156996587, "acc_stderr": 0.01459913135303501, "acc_norm": 0.5469283276450512, "acc_norm_stderr": 0.014546892052005628 }, "harness|hellaswag|10": { "acc": 0.5714997012547302, "acc_stderr": 0.0049385003039902845, "acc_norm": 0.7589125672176857, "acc_norm_stderr": 0.004268690572638815 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.26, "acc_stderr": 0.0440844002276808, "acc_norm": 0.26, "acc_norm_stderr": 0.0440844002276808 }, ... }
配置详情
-
config_name: harness_arc_challenge_25
- split: 2024_02_09T12_50_13.790724
- path: **/details_harness|arc:challenge|25_2024-02-09T12-50-13.790724.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2024-02-09T12-50-13.790724.parquet
- split: 2024_02_09T12_50_13.790724
-
config_name: harness_gsm8k_5
- split: 2024_02_09T12_50_13.790724
- path: **/details_harness|gsm8k|5_2024-02-09T12-50-13.790724.parquet
- split: latest
- path: **/details_harness|gsm8k|5_2024-02-09T12-50-13.790724.parquet
- split: 2024_02_09T12_50_13.790724
-
config_name: harness_hellaswag_10
- split: 2024_02_09T12_50_13.790724
- path: **/details_harness|hellaswag|10_2024-02-09T12-50-13.790724.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2024-02-09T12-50-13.790724.parquet
- split: 2024_02_09T12_50_13.790724
-
config_name: harness_hendrycksTest_5
- split: 2024_02_09T12_50_13.790724
- path:
- **/details_harness|hendrycksTest-abstract_algebra|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-anatomy|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-astronomy|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-college_biology|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-college_physics|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-computer_security|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-econometrics|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2024-02-09T12-50-13.790724.parquet
- **/details_harness|hendrycksTest-global_facts|5_2024-02-09T12-50-13.790724.parquet
- path:
- split: 2024_02_09T12_50_13.790724



