While there are many great tutorials and talks showing how to train machine learning models, there is little information on what happens after we have trained our model. How can we store, utilize, and update it?

Open Source frameworks such as Spark, TensorFlow, MXNet, or PyTorch enable anyone to model and train Machine Learning Models. In this webinar, Mesosphere technical lead Joerg Schad looks at the complete deep learning pipeline, including a live demo of Mesosphere DC/OS. Joerg will addresses commonly asked questions such as:

  • How can we easily deploy distributed deep learning frameworks on any public or private infrastructure?
  • How can we manage different deep learning frameworks on a single cluster, especially considering heterogeneous resources such as GPUs?
  • What is the best UI for a data scientist to work with the cluster?
  • How can we store & serve models at scale?
  • How can we update models that are currently in use without causing downtime for the service using them?
  • How can we monitor the entire pipeline and track performance of the deployed models?

View the slides from the webinar on Slideshare.