Hi, thanks for your brilliant work! I'm currently working on some classification tasks and have encountered a challenge. The feature output from the mid block of UNet is of shape (1, 1024, 32, 32), which has a large number of channels. Could you please provide guidance on how to set up the downstream model for this classification task?
Hi, thanks for your brilliant work! I'm currently working on some classification tasks and have encountered a challenge. The feature output from the mid block of UNet is of shape (1, 1024, 32, 32), which has a large number of channels. Could you please provide guidance on how to set up the downstream model for this classification task?