Refactor TFP call to just sigmoid() #29641
Merged
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Outdated suggestions cannot be applied.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.
We're seeing issues caused by Tensorflow Probability + Keras 3 incompatibilities. Annoyingly, these are being thrown even when people are trying to use unrelated Torch models, as long as they have TF / TFP installed!
The underlying cause is that we import TFP in one of our pipeline files if it's present, and this file is run when
pipeline
is imported, even if the user is using an entirely different pipeline/framework. I removed the import entirely - it was only used to get the probabilities of a Bernoulli distribution given input logits, but this is just the sigmoid function, so I replaced the whole thing with a single call tosigmoid()
instead.