Putting Deep Learning into Production

Deep learning models are achieving state-of-the-art results in speech, image/video classification and numerous other areas, but deploying them to production often involves a unique set of challenges including prediction latency, significant training cost, device memory requirements, etc.

This conference will focus on some best practices for deploying deep learning models into production.  We'll explore some

  • What are some ways to speed up training time?
  • How can pre-trained models be used?
  • Can knowledge from a different domain be "transferred" to the task at hand?
  • Can the model size be reduced to increase prediction latency?


Jan 21, 2017, 9:30a - 5p



To be announced


Topics Covered

Increasing training speed

TensorFlow APIs

Model Zoo

Reusing pre-trained models

Reducing model size

Transfer learning across tasks