Guangzhi Li

@jagadeesh09 Model parameters are not the only one that occupies the GPU memory, reduce batch_size to 16 or smaller would help.

@jagadeesh09 The type of parameter "bias" in torch.nn.Conv2d should be bool(Check API at http://pytorch.org/docs/master/nn.html), but conv.bias is FloatTensor, change to bias = True if conv.bias is not None, otherwise False.

@fenollp Thanks for your reply. What would you suggest to do if I would like to achieve the binary optimization? modify PyTorch core?

Thank you @cow8 That's very helpful!

@jiecaoyu Sounds great. Good luck with the paper.