Skip to content

Conversation

@romitjain
Copy link
Collaborator

ODM will try to import cuml first. If it is available, it will use GPU accelerated KMeans.

The reason to not add it as a dependency in pyproject.toml is because managing cuml dependencies via pip is extremely challenging. For our use cases, since we use the NVCR image, it will already be available in the image and can leverage it.

Signed-off-by: romit <romit@ibm.com>
Signed-off-by: romit <romit@ibm.com>
@ashokponkumar ashokponkumar merged commit 3a9d3fd into foundation-model-stack:main Dec 12, 2025
9 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.

2 participants