MLFlow Integration #1542
Open
MLFlow Integration #1542
+620
−25
Conversation
This file contains hidden or 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
Signed-off-by: Naveenraj Kamalakannan <[email protected]>
Signed-off-by: Naveenraj Kamalakannan <[email protected]>
Phlip79
requested changes
Dec 2, 2025
Co-authored-by: Philip Petrakian <[email protected]> Signed-off-by: Naveenraj Kamalakannan <[email protected]>
Signed-off-by: Naveenraj Kamalakannan <[email protected]>
Author
|
Hi @Phlip79 - implemented all the changes. |
yaoyu-33
reviewed
Dec 2, 2025
| mlflow.set_experiment(logger_cfg.mlflow_experiment) | ||
|
|
||
| # Prepare tags and params | ||
| def safe_serialize(obj: Any) -> str: |
Contributor
There was a problem hiding this comment.
can we move this to utils? it's duplicated with
Contributor
|
thanks for the contribution, all look good. Had one comment. |
maanug-nv
requested changes
Dec 3, 2025
Contributor
maanug-nv
left a comment
maanug-nv
left a comment
There was a problem hiding this comment.
Thanks for adding the support. Left 1 suggestion. Can I also ask you to add unit tests similar to what we do for wandb?
| 1) Install MLFlow (if not already available): | ||
|
|
||
| ```bash | ||
| pip install mlflow |
Contributor
There was a problem hiding this comment.
I think we should add this as an optional dependency, and maybe in validate, or LoggerConfig.finalize() assert that mlflow is importable. Maybe even a hard dep, since wandb is hard dep.
Signed-off-by: Naveenraj Kamalakannan <[email protected]>
Signed-off-by: Naveenraj Kamalakannan <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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Outdated suggestions cannot be applied.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.
What does this PR do ?
Add MLFlow integration for experiment tracking and artifact logging.
This PR adds comprehensive MLFlow support to Megatron Bridge, enabling users to log training metrics, configuration parameters, and checkpoint artifacts to MLFlow tracking servers. The integration includes automatic checkpoint artifact logging with iteration-based artifact paths.
MLFlow logging follows the same pattern as the existing W&B integration and can be enabled via LoggerConfig with mlflow_experiment and mlflow_run_name parameters.
Before your PR is "Ready for review"
Pre checks:
If you haven't finished some of the above items you can still open "Draft" PR.
Additional Information
cc @Phlip79