Skip to content
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

Use bfloat16 for eval #66

Open
tbaker2 opened this issue Jan 26, 2024 · 1 comment
Open

Use bfloat16 for eval #66

tbaker2 opened this issue Jan 26, 2024 · 1 comment

Comments

@tbaker2
Copy link

tbaker2 commented Jan 26, 2024

I'm running paxml on an Intel Xeon CPU server using the paxml/main.py program. I'm trying to create a model that creates weights in bfloat16, and uses that datatype during eval. I modified the LmCloudSpmd2B configuration with the following lines:

MODEL_DTYPE = jnp.bfloat16
ICI_MESH_SHAPE = [1, 1, 1]

The training status output includes the following output.

model.dtype : type/jax.numpy/float32
model.fprop_dtype : dtype[bfloat16]

All of the other operator datatypes are float32. When I run that model with the --eval switch all of the computation is in float32. How can I direct paxml to use bfloat16?

Tom

@tbaker2
Copy link
Author

tbaker2 commented Feb 12, 2024

Any comments on this?

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

No branches or pull requests

1 participant