Say I have the convolution
conv = torch.nn.Conv2d(self.ch,4*self.ch,kernel_size=(1,3),padding=(0,1))
This convolutional network should be adjusting its kernel/weights over training. Is there a way for me to track the values of the kernel/weights of the convolution? Something like conv.parameters()?
I guess you can watch the weights of your conv via
print(conv.weight)
2 Likes
this works, thank you 