emiliopaolini

Support boost::multiprecision::cpp_int as BaseType

1

Hi everyone, as the title says, i was trying to use cpp_int defined in boost::multiprecision as base type. But when i declare a fixed point using cpp_int like this: using...

Power of 2 quantization

1

I was wondering if there is a way to perform power of 2 quantization with Larq. Maybe a specific quantizer is needed? Any suggestions?

is BatchNorm binarized?

5

Hi everyone, i have a simple question: if i look at the summary of every model in larq, it looks like that computation for batch normalization has been binarized. I...

Documentation

3

Hi all, i am trying to use the library with python bindings. However i am finding some issues and I cannot find a documentation in which the methods available are...

Hi everyone, i am asking if it is possible to define custom types for weights, gradients and activations. If this is not possible, do you have any suggestions on how...