关于Pytorch中view函数的使用

If you are viewing collapsing non-neighboring dimensions, you are interleaving the pixels. You could try it on an image tensor and your result should be interleaved. 注意,在使用view函数的时候一定要配合permute函数使用,类似 patches = patches.contiguous().view(B, C, -1, kernel_size*kernel_size) print(patches.shape) # [B,