SciLifeLab Serve has added MLFlow to its existing suite of tools related to building and using machine learning models for life science research. MLFlow is the most popular open source solution for tracking machine learning experiments.

Those affiliated with a Swedish research institution can login to SciLifeLab Serve and use MLFlow to track machine learning experiments regardless of where they take place. This means that you can track experiments done on your computer, your research group machine, or national compute resources. Users need only connect to servers at SciLifeLab Serve, and begin sending metrics and artifacts from the experiments. These metrics and artifacts, as well as other features, become available in the graphical user interface of the user’s MLFlow instance on SciLifeLab Serve. For more information on how to do this, see the documentation page on SciLifeLab Serve.