Freezing model but not sure it's working

[LeeDoYup (Doyup Lee)]

same question in

Parameters with requires_grad = False are updated during training

Hello. I’am trying to freeze front layers during training. Before starting optimization, the optimizer is constructed by optimizer = torch.optim.SGD(net.parameters(), lr, ...) Then, during training, i changed the front layers’ requires_grad=False. Specifically, for epoch in range(total_epoch): if epoch == freeze_epoch: net.conv1.weight.requires_grad = False train_one_epoch() #update network in an epoch However, I found that the weights of the front layers are still updated. I als…

momentum, weight decay affect updating params with requires_grad=False