Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Bias in catconv2d #63

Open
joihn opened this issue Jan 18, 2024 · 0 comments
Open

Bias in catconv2d #63

joihn opened this issue Jan 18, 2024 · 0 comments

Comments

@joihn
Copy link

joihn commented Jan 18, 2024

hardnet v1 does not have conv2d bias.
ptsemseg/models/hardnet.py:11

        self.add_module('conv', nn.Conv2d(in_channels, out_channels, kernel_size=kernel,
                                          stride=stride, padding=kernel//2, bias = False))

However, hardnetv2, which use catconv2d does have conv2d bias
CatConv2d/catconv2d.py:39

        self.bias = nn.Parameter(torch.Tensor(out_channels))

Why does catconv2d has bias ? isn't it useless ?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant