Freeze the weights / kernel

[Ashwin_Raju]

I can see many tutorial about freezing layers or freezing all weights in a layer but I would like to freeze only a subset of weights. For example, if the kernel size is 7x7 I would like to train only 5x5 starting from top left and freezing right bottom 2x2. I would like to declare it as 7x7 rather than 5x5. Is it possible to freeze these 2x2 ?

[Ranahanocka]

See this post:

Freezing part of the Layer Weights

Hello Everyone, How could I freeze some parts of the layer weights to zero and not the entire layer. I tried below code, but it doesn’t freeze the specific parts(1:10 array in 2nd dimension) of the layer weights. I am new to ML & started with Pytorch. Appreciate any help. Thanks. for child in model_ft.children(): print(“Freezing Parameters(1->10) on the Convolution Layer”,child) for param in child.parameters(): param.data[:,1:10,:,:].zero_() param.data[:,1:10,:,:].requires_grad = False o…

Essentially you just don’t those specefic parameters grad.