Zipformer MVQ #1190
Open
Zipformer MVQ #1190
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 results: 100 hours:
960 hours:
|
danpovey
reviewed
Aug 3, 2023
| sp=sp, | ||
| params=params, | ||
| ) | ||
| # if not params.print_diagnostics: |
There was a problem hiding this comment.
is there a particular reason this was commented?
There was a problem hiding this comment.
It was just for speed during my exp. Will make sure to un-comment them when I merge the PR.
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Outdated suggestions cannot be applied.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.
This PR makes training with knowledge distillation an option in the Zipformer recipe. The knowledge distillation method is MVQ-KD.
The teacher targets can be downloaded via the following command:
To turn on knowledge distillation, you will need to set
--enable-distillation True. It is applicable to both streaming and non-streaming Zipformers.Detailed results will follow.