You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
请问当我应用iaff模块时候存在以下情况:1、F的batch_size为1;2、全局平均池化将特征图的面积变为1*1。这导致了经过BN层的时候会报错ValueError: Expected more than 1 value per channel when training, got input size torch.Size([1, 256, 1, 1]),请问这个有处理方法吗
The text was updated successfully, but these errors were encountered:
请问当我应用iaff模块时候存在以下情况:1、F的batch_size为1;2、全局平均池化将特征图的面积变为1*1。这导致了经过BN层的时候会报错ValueError: Expected more than 1 value per channel when training, got input size torch.Size([1, 256, 1, 1]),请问这个有处理方法吗
The text was updated successfully, but these errors were encountered: