Use MLflow as logging system

microsoft / dstoolkit-mlops-base

Support ML teams to accelerate their model deployment to production leveraging Azure

MIT License

89 stars 39 forks source link

Use MLflow as logging system #24

Open mariamedp opened 2 years ago

mariamedp commented 2 years ago

MLflow is becoming the most common model management library. Enabling it by default in the template is a requirement as the new AML SDK version will rely on it more heavily.

Changes needed:

[ ] Change AML run.log() methods by mlflow equivalent.
[ ] Use mlflow to log models as well.
[ ] Add examples on how to log images in experiment.

Instructions in AML docs: https://docs.microsoft.com/azure/machine-learning/how-to-use-mlflow

abhi-rf commented 2 years ago

Hello! I am new to contributing and I would like to give this issue a try! @mariamedp

mariamedp commented 2 years ago

Hi @abhi-rf, that'd be great, thank you! I'm currently working on issue #27 that will introduce some modifications to the training scripts, can you give me a few hours to commit the changes before you start working on this? Meanwhile you can start by setting up your pipelines for testing (instructions), if you haven't done it already.

mariamedp commented 2 years ago

Hi again @abhi-rf, this is the PR: https://github.com/microsoft/dstoolkit-mlops-base/pull/57. It's pending review & approval by my colleagues, but you can have a sneak peak of the changes by looking at that branch. It will be merged next week.