Keras DataGenerator slower by design than Pytorch DataLoader?

前端 未结 0 634
面向向阳花
面向向阳花 2021-01-18 06:38

An observation: DataGenerator\'s getitem() must return a batch. This means that if you don\'t make getitem() parallel under the hood then i

相关标签:
回答
  • 消灭零回复
提交回复
热议问题