Recently working with tlt for deploying optimized models into trt. I couldn’t find the code used to train the models I’m using (currently retinanet). Please coordinate me to the according github or other source in order to train the nets myself before applying tlt.
I saw a feature of pruning. It is a lot of benefit for me to take control over the source code of the models that are pretrained in order to anneal weights to 0 before using tlt. Thanks