Closed ricardobarroslourenco closed 4 years ago
Hi Richardo @ricardobarroslourenco
Sorry for the late reply. based on what I heard from the teammates using gpu, it should work with code, but quantized network is not optimized to run fast in gpus for inference.
Ok. No problems. I was more concerned on the training portion.
I would like to know if QKeras is architecture agnostic (following TF requirements) since it lists as import the standard
tensorflow
python library (I assume if that is correct,tensorflow-gpu
would be supported as well).