Update nn.py to use jnp.add rather than + operand. #19323
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
fchollet
reviewed
Mar 18, 2024
keras/backend/jax/nn.py
Outdated
| @@ -546,7 +546,7 @@ def batch_normalization( | |||
| offset = jnp.reshape(offset, shape) | |||
| res = res + offset | |||
|
|
|||
| return x * inv + res | |||
| return jnp.add(x + inv, res) | |||
There was a problem hiding this comment.
Thanks for catching, updated accordingly.
|
Was this change motivated by a specific error you encountered? |
fchollet
approved these changes
Apr 15, 2024
google-ml-butler
bot
added
kokoro:force-run
ready to pull
google-ml-butler
bot
removed
ready to pull
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Outdated suggestions cannot be applied.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.
Ref: https://jax.readthedocs.io/en/latest/notebooks/thinking_in_jax.html#numpy-lax-xla-jax-api-layering
jax.numpyimplicitly promotes arguments to allow operations of mixed data types, whilejax.laxdoes not. If usingjax.laxdirectly, it is advised to do explicit type promotion. Because the core functions in nn.py are the backbone for various model edge cases, it is better to use a safer operand.