def forward(self, X):
out1 = self.net1(X)
out2 = self.net2(X)
out = grouped_conv(out1, out2)
return self.fc(out)
def forward(self, X):
out1 = self.features_net(X)
out2 = self.features_net(X)
out = grouped_conv(out1, out2)
return self.fc(out)
def forward(self, X):
out1 = self.features_net(X)
out2 = self.filters_net(X)
out = grouped_conv(out1, out2)
out = self.bn(out.view(out2.shape[0], 64, 6, 6)).view(out2.shape[0], -1)
out = self.fc(out)
return out
def forward(self, X):
out1 = self.features_net(X)
out2 = self.filters_net(X)
out = grouped_conv(out1, out2)
out = self.bn(out.view(out2.shape[0], -1, out.shape[2],
out.shape[3])) # Batchnorm and relu
return self.fc(out.view(out2.shape[0], -1))
def forward(self, X):
"""
Forward Prop
:param X: Input dataset with batch dimension
:return: Output of model and parameters
"""
out1 = self.net1(X)
out2 = self.net2(X)
out = grouped_conv(out1, out2)
out = self.bn(out.view(out2.shape[0], -1, out.shape[2], out.shape[3])) # Batchnorm and relu
out = self.fc(out.view(out2.shape[0], -1))
return out
def forward(self, X):
"""
Forward Prop
:param X: Input dataset with batch dimension
:return: Output of model and parameters
"""
out1 = self.net1(X)
out2 = self.net2(X)
out3 = grouped_conv(out1, out2)
out3 = self.fc(out3)
return out3