Nn.init.xavier_uniform_

Hello!
Does anyone know what these two lines of code do Please?
nn.init.xavier_uniform_(self.conv1.weight)
nn.init.constant_(self.conv1.bias, 0.0)

The first line fills self.conv1.weight with samples from a uniform distribution, while the second line fills self.conv1.bias with zeros.

See here for more information.

1 Like

Am gonna see, Thank you so much

Do you know when i could use the uniform distribution please

That depends on your application. See here for a good tutorial on this.

2 Likes

Thank you so much :slightly_smiling_face: :slightly_smiling_face: