TensorFlow implementation for paper "AdderNet: Do We Really Need Multiplications in Deep Learning?"
This code repository has not been tested well following the original paper.
Is set to automatic (signSGD) rather than by full-precision gradient.
[1] AdderNet: Do We Really Need Multiplications in Deep Learning?, https://arxiv.org/abs/1912.13200
[2] Code, https://github.com/huawei-noah/AdderNet