SqueezeAttention-PyTorch

SqueezeAttention-PyTorch copied to clipboard

Thanks for your code! I have a question, if using convolution + bn + relu, why convolution has bias? I think it is unnecssary.

HsLOL

model

2

能写个带有sablock的resnet网络吗

lewisxiaoxu

关于论文

2

个人感觉 你的实现 没有什么问题，但是我咋感觉这篇论文写得逻辑有问题呢，就是很多细节忽略了

doudou123456

Based on your understanding, where does the sa module fit into the model

Sunshine066