Fix the bug of dropout and prenet. #205
Merged
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
We should put DecoderPrenetWrapper outside AttentionWrapper. Otherwise, the AttentionWrapper concatenates the input and previous attention vector before calling the internal cell, so dropout in the DecoderPrenetWrapper will drop 50% information of previous attention vector.
|
Hi I would like to ask you that if we cut off half of samples in |
|
LJSpeech is about 23 hours, and I think maybe 10+ hours of data is enough to learn the alignments. |
|
Thanks for the fixes! |
|
@npuichigo have you tested this on reduction factor of 5 (outputs_per_step=5)? With this change, my model is not wanting to align. Usually it'll align at about 10k steps. Currently at 30 and it's not wanting to align correctly.
Changed found here. |
|
@LearnedVector I tested on reduction factor 2 with LJSpeech and alignments can be learned after about 20k steps. |
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Outdated suggestions cannot be applied.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.
I have tested that with both of the bugs fixed, good alignments can be learned with reduction factor 2 easily.