Different upstream and downstream learning rates #525
Conversation
This file contains hidden or 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
|
Hi @xt0r3-Cambridge - just wanted to say thanks for implementing this. I recently used it when fine-tuning, since using such a high LR generally wipes all the pretraining knowledge. Hopefully the s3prl maintainers will take a look at this eventually. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
Implemented support for different upstream learning rates as raised in a feature request.
Use instructions:
Simply add another
upstream_lrparameter to the optimiser settings, like this:The upstream training only happens if the code is run with the
--upstream-trainable(or-f) flag.