add WSD-S scheduler #35
Open
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
|
Some num_decays, decay_step_perc configurations with their respective LR schedules. The standard linear warmup and linear decay: num_decays=1, decay_step_perc=1.0 (default) 
WSD: num_decays=1, decay_step_perc=0.1 WSD-S: num_decays=4, decay_step_perc=0.1 |
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Outdated suggestions cannot be applied.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.
Add WSD-S scheduler
The two main parameters added are the num_decays and decay_step_perc
num_decays represents the total number of times to decay throughout the training (eg LR will be decayed each time after completing training_steps / num_decays steps)
decay_step_perc represents the percentage of the total steps (at which LR will be decayed) to use for the decay phase
This choice of hparams, requires minimal effort and is compatible with the current train resuming functionality.