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.

Sign up for GitHub

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

Jump to bottom

Check ResNet compat TF Torch #9

Open
zaccharieramzi opened this issue Apr 8, 2022 · 1 comment
Open

Check ResNet compat TF Torch #9

zaccharieramzi opened this issue Apr 8, 2022 · 1 comment

Comments

@zaccharieramzi
Copy link
Collaborator

One last point omitted was this one: https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py#L107-L111

Need to make sure that the same is used in both frameworks
@zaccharieramzi
Copy link
Collaborator Author

In TF, the stride is at the same place as in Torch, i.e. in conv2, see here.

In Keras, however, the stride is in conv1, see here.
In block2, the stride is in conv2, but the architecture also features preactivation, unlike PyTorch, which is why they called it 1.5 in PyTorch comments...

Maybe we can accept this slight change... Wdyt @tomMoral ?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@zaccharieramzi