Miaite

The network bug, conv layers were not supposed to share weights

1

In networks.py, when init class RED_CNN, only 4 nn.Conv2d were instantiated. ``` def __init__(self, out_ch=96): super(RED_CNN, self).__init__() self.conv_first = nn.Conv2d(1, out_ch, kernel_size=5, stride=1, padding=0) self.conv = nn.Conv2d(out_ch, out_ch, kernel_size=5, stride=1,...

https://app.mokahr.com/campus_apply/igg/28211#/jobs?zhineng=50086&location=%E6%88%90%E9%83%BD%E5%B8%82&page=1