DynamicReLU

DynamicReLU copied to clipboard

in fact,the dynamic relu is well?who try to use it ?

3

daixiangzi

conv3d

1

Hello, can I ask how to operate with conv_type='3d'? about the code 'theta = torch.mean(theta, axis=-1)' when conv_type = '2d'? Do somebody know?Thanks a lot.

walynlee

Excuse me,could you please share the code of DynamicReluC,which is mentioned in the paper?

jianyulei1

Excuse me,could you please share the code of DynamicReluC,which is mentioned in the paper?

zhyyfaan

There may be a bug here.

1

def get_relu_coefs(self, x): print(x.shape) # axis? theta = torch.mean(x, dim=-1) if self.conv_type == '2d': # axis? theta = torch.mean(theta, dim=-1) theta = self.fc1(theta) theta = self.relu(theta) theta = self.fc2(theta) theta...

L-Icarus

About reduction ratios R

4

Thanks for your working! I have some questions: 1. That reduction ratios R=8 is the best trade-off in the paper , but the code defaults to 4, so why? 2....

bobo0810

About the results in the DynamicReLU paper

1

Hi, have you ever reimplemented the results as the paper reported, such as MobileNet or ResNet?

larenzhang