meetup

Everyone who has tried to do ML development knows that it is complex.
Beyond the usual concerns in the software development, ML development comes with new challenges:

  • There are a myriad tools
  • It’s hard to track experiments
  • It’s hard to reproduce results
  • It’s hard to deploy ML
And because of these challenges, it is clear that ML development has to evolve a lot to become as robust,
predictable and wide-spread as traditional software development In this talk we have had Abdulrahman Alfozan as our presenter he took us through the machine learning lifecycle and its development challenges and will feature a demo of model training, tuning and serving using the MLflow framework.
MLflow which is an open source machine learning platform aims to simplify the management of machine learning pipelines

Slides : Notebook: