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

Disabling GPU-related warnings #20

Open
epaillas opened this issue May 21, 2023 · 1 comment
Open

Disabling GPU-related warnings #20

epaillas opened this issue May 21, 2023 · 1 comment

Comments

@epaillas
Copy link

Hello - I was wondering if there is a way to disable the following warnings, which have started appearing when running pypower on interactive nodes in Perlmutter (and I guess they will also appear when submitting to the queue). I'm assuming these are related to the GPU capabilities of the code.

If not possible to disable them, is there a way so that they will only be printed by the root process? Each rank is printing its own copy, so the terminal is quickly filled up with these warnings.

[000002.72]  05-21 10:20  absl                         INFO     Unable to initialize backend 'tpu_driver': NOT_FOUND: Unable to find driver in registry given worker:
[000002.72]  05-21 10:20  absl                         INFO     Unable to initialize backend 'cuda': module 'jaxlib.xla_extension' has no attribute 'GpuAllocatorConfig'
[000002.72]  05-21 10:20  absl                         INFO     Unable to initialize backend 'rocm': module 'jaxlib.xla_extension' has no attribute 'GpuAllocatorConfig'
[000002.72]  05-21 10:20  absl                         INFO     Unable to initialize backend 'tpu': INVALID_ARGUMENT: TpuPlatform is not available.
[000002.72]  05-21 10:20  absl                         WARNING  No GPU/TPU found, falling back to CPU. (Set TF_CPP_MIN_LOG_LEVEL=0 and rerun for more info.)

Cheers,

@echaussidon
Copy link

If you are working with cpu, you can try :

export JAX_PLATFORM_NAME=cpu

or you can set it in the python code:

import jax; jax.config.update('jax_platform_name', 'cpu')

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

2 participants