open-llm-leaderboard-old/details_budecosystem__genz-70b
收藏数据集概述
数据集简介
该数据集是在对模型 budecosystem/genz-70b 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。
数据集结构
- 配置数量:64个配置,每个配置对应一个评估任务。
- 运行次数:数据集从2次运行中创建。每个运行在每个配置中作为一个特定的分割存在,分割名称使用运行的时间戳。
- 训练分割:"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,用于计算和显示聚合指标在 Open LLM Leaderboard 上。
数据加载示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_budecosystem__genz-70b", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-23T19:01:32.642131 运行 的最新结果: python { "all": { "em": 0.421875, "em_stderr": 0.005057576044482799, "f1": 0.5428481543624201, "f1_stderr": 0.004562270615925701, "acc": 0.5862101051177826, "acc_stderr": 0.011727291302229777 }, "harness|drop|3": { "em": 0.421875, "em_stderr": 0.005057576044482799, "f1": 0.5428481543624201, "f1_stderr": 0.004562270615925701 }, "harness|gsm8k|5": { "acc": 0.3373768006065201, "acc_stderr": 0.013023665136222105 }, "harness|winogrande|5": { "acc": 0.835043409629045, "acc_stderr": 0.010430917468237448 } }
配置详情
-
harness_arc_challenge_25
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|arc:challenge|25_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-09-13T09-54-04.852738.parquet
-
harness_drop_3
- 分割:2023_10_23T19_01_32.642131
- 路径:
**/details_harness|drop|3_2023-10-23T19-01-32.642131.parquet - 分割:latest
- 路径:
**/details_harness|drop|3_2023-10-23T19-01-32.642131.parquet
-
harness_gsm8k_5
- 分割:2023_10_23T19_01_32.642131
- 路径:
**/details_harness|gsm8k|5_2023-10-23T19-01-32.642131.parquet - 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-10-23T19-01-32.642131.parquet
-
harness_hellaswag_10
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hellaswag|10_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_abstract_algebra_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_anatomy_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_astronomy_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_business_ethics_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_clinical_knowledge_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_college_biology_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_college_chemistry_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_college_computer_science_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_college_mathematics_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_college_medicine_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_college_physics_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_computer_security_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_conceptual_physics_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_econometrics_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-09-13T09-54-04.852738.parquet
-
harness_hendrycksTest_electrical_engineering_5
- 分割:2023_09_13T09_54_04.852738
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-13T09-54-04.852738.parquet - 分割:latest
- 路径:`**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-13T09-54-04.852738



