open-llm-leaderboard/details_TFLai__Orca-Nova-13B
收藏数据集概述
数据集名称
Evaluation run of TFLai/Orca-Nova-13B
数据集描述
该数据集是在模型 TFLai/Orca-Nova-13B 在 Open LLM Leaderboard 上的评估运行期间自动创建的。
数据集结构
- 该数据集包含 64 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作为一个特定的拆分找到,拆分名称使用运行的时间戳。"train" 拆分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_TFLai__Orca-Nova-13B", "harness_winogrande_5", split="train")
最新结果
这些是最新结果(来自 2023-10-28T09:46:46.108882 运行): python { "all": { "em": 0.003984899328859061, "em_stderr": 0.0006451805848102392, "f1": 0.07519924496644283, "f1_stderr": 0.0016030527256702374, "acc": 0.46032756632616734, "acc_stderr": 0.010706817769913408 }, "harness|drop|3": { "em": 0.003984899328859061, "em_stderr": 0.0006451805848102392, "f1": 0.07519924496644283, "f1_stderr": 0.0016030527256702374 }, "harness|gsm8k|5": { "acc": 0.14480667172100076, "acc_stderr": 0.009693234799052708 }, "harness|winogrande|5": { "acc": 0.7758484609313339, "acc_stderr": 0.011720400740774108 } }
配置详情
-
harness_arc_challenge_25
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|arc:challenge|25_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|arc:challenge|25_2023-09-12T12-05-50.844177.parquet
-
harness_drop_3
- 拆分:2023_10_28T09_46_46.108882
- 路径:
**/details_harness|drop|3_2023-10-28T09-46-46.108882.parquet - 拆分:latest
- 路径:
**/details_harness|drop|3_2023-10-28T09-46-46.108882.parquet
-
harness_gsm8k_5
- 拆分:2023_10_28T09_46_46.108882
- 路径:
**/details_harness|gsm8k|5_2023-10-28T09-46-46.108882.parquet - 拆分:latest
- 路径:
**/details_harness|gsm8k|5_2023-10-28T09-46-46.108882.parquet
-
harness_hellaswag_10
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hellaswag|10_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hellaswag|10_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-12T12-05-50.844177.parquet等 - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-12T12-05-50.844177.parquet等
-
harness_hendrycksTest_abstract_algebra_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_anatomy_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_astronomy_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_business_ethics_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_clinical_knowledge_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_college_biology_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_college_chemistry_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_college_computer_science_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_college_mathematics_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_college_medicine_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_college_physics_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_computer_security_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_conceptual_physics_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_econometrics_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-09-12T12-05-50.844177.parquet
-
harness_hendrycksTest_electrical_engineering_5
- 拆分:2023_09_12T12_05_50.844177
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-12T12-05-50.844177.parquet - 拆分:latest
- 路径:`**/details_harness|hendrycksTest-electrical



