Brian Tom

I don't think that converting layers in bnn.torch to a cpu layer is going to work without some effort. If you look at the implementation of the layers in bnn.torch,...

Hmmm. I don't think that works because the lua `convert` function relies on the destination module `nn` in this case to have the corresponding classes `SpatialConvolution` and `Binary`. nn does...

That's correct, but you only need to write the forward pass of it. 1. decode the weights from int32 to binary 2. perform regular spatial convolution from the binary weights...