Tensorfow DNN quantization with mixed precision

AI & Data Science Deep Learning (Training & Inference) TensorRT
#1

Hello everyone,

I have a tensorflow model trained and converted to a a frozen graph (.pb).
I want to use tensorrt with python in order to apply mixed precision quantization on this model.

Can you please give me some examples?
Thank you

#2

Hi,

Please refer to below links for mixed precision python sample:
https://docs.nvidia.com/deeplearning/sdk/tensorrt-archived/tensorrt-601/tensorrt-developer-guide/index.html#unique_299310851
https://docs.nvidia.com/deeplearning/sdk/tensorrt-sample-support-guide/index.html#int8_caffe_mnist

Thanks

#3

Hello and thanks for the answer,

However, what I wanted to do is to use the Tensorflow TensorRT (TF-TRT) API to apply mixed precision quantization.

PS: Tensorflow 1.15 + TensorRT 5