Integer-Arithmetic-Only

Quantization aware training 量化背后的技术——Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference

1，概述　　模型量化属于模型压缩的范畴，模型压缩的目的旨在降低模型的内存大小，加速模型的推断速度（除了压缩之 […]