Skip to content

Model Batching V4 #231

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Conversation

adenzler-nvidia
Copy link
Collaborator

Previous discussions in #109 , another attempt in #195

@btaba could you help test-driving this from the JAX side once that's ready?

@adenzler-nvidia adenzler-nvidia requested review from erikfrey and btaba May 9, 2025 08:16
Copy link
Collaborator

@btaba btaba left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi @adenzler-nvidia I'm not quite ready with the JAX shim to test, but I expect maybe by EOW. Nevertheless, I think this approach will work a lot better with JAX than prior versions, I can't foresee any big issues.

@thowell
Copy link
Collaborator

thowell commented May 12, 2025

@adenzler-nvidia thanks for making the updates to enable model batching for these fields!

similar to @btaba i think this pr has a good selection of batched Model fields and additional fields can easily be updated with a batch dimension if needed.

added a few minor comments for docstring formatting.

thanks!

Copy link
Collaborator

@erikfrey erikfrey left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

OK, looks great and conforms exactly to our plan and discussions. Let's go!

@adenzler-nvidia adenzler-nvidia merged commit f01468c into google-deepmind:main May 14, 2025
11 of 13 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants