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

Discrepancy between the SimSiam paper and lightning implementation? #572

Open
shashank2000 opened this issue Mar 3, 2021 · 0 comments
Open
Labels
enhancement New feature or request model

Comments

@shashank2000
Copy link
Contributor

https://github.com/PyTorchLightning/pytorch-lightning-bolts/blob/61d3a26f45b8f01232e76e18d22d55eecb4e6c77/pl_bolts/models/self_supervised/simsiam/models.py#L45

Possibly a nit - but it seems like we use the same hidden size for both the projection and prediction MLP, when in fact the projection MLP's hidden fc is supposed to be of dimension 2048 and the prediction MLP of hidden size 512. This is based on the Baseline settings sub-section of the paper's Methods section.

@akihironitta akihironitta added enhancement New feature or request model labels Mar 20, 2021
@akihironitta akihironitta added this to To do in Self Supervised Learning via automation Mar 20, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request model
Development

No branches or pull requests

2 participants