mobilenetv3.pytorch icon indicating copy to clipboard operation
mobilenetv3.pytorch copied to clipboard

Published 20 hours ago • verified publisher icon d-li14

mobilenetv3 代码实现似乎和论文有出入

Open nlp-tokio opened this issue 4 years ago • 1 comments
trafficstars

Mobilenetv3模块实现部分好像少了一个pointwise convolution, 见 https://github.com/d-li14/mobilenetv3.pytorch/blob/master/mobilenetv3.py#:~:text=and%20inp%20%3D%3D%20oup-,if%20inp%20%3D%3D%20hidden_dim%3A,),-else%3A 在line94-104 这部分只适用于第一个模块，代码可以继续简化

nlp-tokio avatar Jun 30 '21 08:06 nlp-tokio

应该不缺吧，你可以对比下TensorFlow官方实现和PyTorch的官方实现

AlphaNext avatar Feb 06 '23 02:02 AlphaNext