Hello Everyone, we have a built a machine learnin...
# questions
a
Hello Everyone, we have a built a machine learning model using kedro pipelines, and we would like to integrate MLflow to keep track of different metrics, version etc. What is the best way to start? Are there any examples on how to use the 2 frameworks together? thank you very much!
a
I find it a bit hard to navigate on this page, but I guess all details can be found there?
n
If you are looking for something dead simple - https://kedro--3856.org.readthedocs.build/en/3856/integrations/mlflow.html, here is the documentation that is coming on the way (we haven't merged this yet).
If you plan to integrate it and utilize most of the mlflow feature,
kedro-mlflow
offers a nicer integration. I do recommend read through the docs to understand why things are structured in certain way. (artifact, deployment). These concepts are important. If you only need the tracking metrics and don't intend to use mlflow for deployment etc, the Kedro docs is a good start
a
thank you very much thats very helpful, I will go through the docs
n
and feel free to make any suggestion to the docs, @Yolan Honoré-Rougé (the author of
kedro-mlflow
is active in this Slack group).
a
will do, I am super new with MLflow so I think this is a good starting point
👍 2
👍🏼 1