I try to run a FC4 model on Huawei P40 PRO using Tensorflow Lite benchmark tool. To get better performances, i have quantized my model in INT8 when converting it. It appears tha
