Change backbone finetuning strategy to allow for DDP #205
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ksikka
requested review from
danbider and
themattinthehatt
and removed request for
danbider
ksikka
force-pushed
the
main
branch
2 times, most recently
from
913b9f2 to
de97315
Compare
|
Run with and without code changes: |
ksikka
commented
Oct 15, 2024
danbider
reviewed
Oct 15, 2024
lightning_pose/callbacks.py
Outdated
| pl_module.current_epoch, backbone_lr, upsampling_lr | ||
| ) | ||
|
|
||
| def _get_next_backbone_lr(self, current_epoch, backbone_lr, upsampling_lr): |
There was a problem hiding this comment.
good to comment on the if conditions, what is this the LR schedule that is implemented here
See Lightning-AI/pytorch-lightning#20340 for the original issue. The implemented workaround no longer freezes/unfreezes layers via require_grad, instead we only modulate the learning rate of the backbone layer.
themattinthehatt
approved these changes
Oct 16, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
See Lightning-AI/pytorch-lightning#20340 for the original issue. The implemented workaround no longer freezes/unfreezes layers via require_grad, instead we only modulate the learning rate of the backbone layer.
Unblocks #138