A slim tensorflow wrapper that provides syntactic sugar for tensor variables. This library will be helpful for practical deep learning researchers not beginners.
I have a question here, In sugartensor, a convolution can be done by adopt "tensor.sg_conv(size=3, dim=1), and a 3x3 conv kernel is randomly initialized. But if I want the conv kernel is initialized by a specific kernel, such as [1,2,3;3,2,1;2,1,3], how should I do it?
I have a question here, In sugartensor, a convolution can be done by adopt "tensor.sg_conv(size=3, dim=1), and a 3x3 conv kernel is randomly initialized. But if I want the conv kernel is initialized by a specific kernel, such as [1,2,3;3,2,1;2,1,3], how should I do it?