Quantized (QAT) EfficientNet Classification Model TensorRT engine

[panahikhas]

Quantized (QAT) EfficientNet Classification Model TensorRT engine

I saw you have a QAT performed version of Efficientnet and a container to use it for workaround. I suggest to compile this model with TensorRT engine and put the engine public to people can use it in their researches. I think this is such helpful.

This process of working with quantized models and TensorRT is very hard because weak support of PyTorch.