open-llm-leaderboard/details_BEE-spoke-data__TinyLlama-1.1bee
收藏数据集概述
数据集简介
该数据集是在评估模型BEE-spoke-data/TinyLlama-1.1bee在Open LLM Leaderboard上的自动创建的。
数据集结构
- 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分片找到,分片名称使用运行的时间戳。
- "train"分片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_BEE-spoke-data__TinyLlama-1.1bee", "harness_winogrande_5", split="train")
最新结果
以下是2023-10-29T06:23:03.017532的最新结果: python { "all": { "em": 0.0018875838926174498, "em_stderr": 0.00044451099905590025, "f1": 0.03736262583892626, "f1_stderr": 0.0011684343114881778, "acc": 0.27343398918005496, "acc_stderr": 0.007654321426298602 }, "harness|drop|3": { "em": 0.0018875838926174498, "em_stderr": 0.00044451099905590025, "f1": 0.03736262583892626, "f1_stderr": 0.0011684343114881778 }, "harness|gsm8k|5": { "acc": 0.002274450341167551, "acc_stderr": 0.0013121578148674216 }, "harness|winogrande|5": { "acc": 0.5445935280189423, "acc_stderr": 0.013996485037729782 } }
配置详情
-
harness_arc_challenge_25
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|arc:challenge|25_2023-09-22T04-13-14.200799.parquet
-
harness_drop_3
- 分片: 2023_10_29T06_23_03.017532, latest
- 路径: **/details_harness|drop|3_2023-10-29T06-23-03.017532.parquet
-
harness_gsm8k_5
- 分片: 2023_10_29T06_23_03.017532, latest
- 路径: **/details_harness|gsm8k|5_2023-10-29T06-23-03.017532.parquet
-
harness_hellaswag_10
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hellaswag|10_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: 多个路径,例如**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_abstract_algebra_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-abstract_algebra|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_anatomy_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-anatomy|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_astronomy_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-astronomy|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_business_ethics_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-business_ethics|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_clinical_knowledge_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_college_biology_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-college_biology|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_college_chemistry_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-college_chemistry|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_college_computer_science_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-college_computer_science|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_college_mathematics_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-college_mathematics|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_college_medicine_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-college_medicine|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_college_physics_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-college_physics|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_computer_security_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-computer_security|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_conceptual_physics_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-conceptual_physics|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_econometrics_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-econometrics|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_electrical_engineering_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-electrical_engineering|5_2023-09-22T04-13-14.200799.parquet
-
harness_hendrycksTest_elementary_mathematics_5
- 分片: 2023_09_22T04_13_14.200799, latest
- 路径: **/details_harness|hendrycksTest-elementary_mathematics|5_2023-09-22T04-13-14.200799.parquet



