Many bug fixes for v1.1.0 #335
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
facebook-github-bot
added
the
CLA Signed
Dinglet
pushed a commit
to Dinglet/audiocraft
that referenced
this pull request
Jul 30, 2024
* wip * fix bug in sample storage * fix bug in fsdp with pytorch 2.0.1 * fix kld * revert back passt change * fix n_quantizers * changes * adding warning * extra warnings and tests * missing changelog * missing changes between audiogen and musicgen
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Outdated suggestions cannot be applied.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 introduces quite a few bug fixes and some non backward compatible changes. From the changelog:
Fixed DAC support with non default number of codebooks.
Fixed bug when
two_step_cfgwas overriden when callinggenerate().Fixed samples being always prompted with audio, rather than having both prompted and unprompted.
Backward incompatible change: A
torch.no_gradaround the computation of the conditioning made its way in the public release. The released models were trained without this. Those impact linear layers applied to the output of the T5 or melody conditioners. We removed it, so you might need to retrain models.Backward incompatible change: Fixing wrong sample rate in CLAP (WARNING if you trained model with CLAP before).
Backward incompatible change: Renamed VALLEPattern to CoarseFirstPattern, as it was wrongly named. Probably no one retrained a model with this pattern, so hopefully this won't impact you!