Add continuous-time consistency models #234
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
Implements continuous-time consistency models as described in [1]
[1] Lu, C., & Song, Y. (2024).
Simplifying, Stabilizing and Scaling Continuous-Time Consistency Models
arXiv preprint arXiv:2410.11081
|
First observation: Loading a model trained in Tensorflow in Jax works with equal quality, so the difference is somewhere in the training process, not in the sampling. |
|
The difference was caused by a coincidence, which led me to discover an error in the loss calculation. It is now fixed, and all backends lead to similar results. |
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Outdated suggestions cannot be applied.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.
Implements continuous-time consistency models as described in [1].
I observe differences in the quality of the samples depending on the backend (all else equal) that I cannot explain yet. Tensorflow performs consistenly better than Jax and PyTorch. If you find any indications for the reasons of this difference, please let me know.
[1] Lu, C., & Song, Y. (2024).
Simplifying, Stabilizing and Scaling Continuous-Time Consistency Models
arXiv preprint arXiv:2410.11081