BatteryPass-12K
收藏DataCite Commons2026-05-01 更新2026-05-03 收录
下载链接：
https://dataverse.harvard.edu/citation?persistentId=doi:10.7910/DVN/IPICAF
下载链接
链接失效反馈官方服务：
资源简介：
This is the first public benchmark for the digital battery passport conformance task: BatteryPass-12K. We used ChatGPT-5.1 Thinking (Standard) to automatically generate 2,000 synthetic samples and their metadata from each of the 6 pilot samples from the global battery alliance (GBA) (where 1,000 are conformant like the original while the other 1,000 are nonconformant). This means there are 12,000 data points in the dataset, which is split in the ratio 80:10:10 for the training, validation and test sets, respectively.
提供机构：
Harvard Dataverse
创建时间：
2026-04-30
