Skip to main content

TensorFlow and Keras

Deep learning model development with TensorFlow and Keras, from the first layer to a served model.

Course duration: 9h​

What you will learn​

  • Build a model with the sequential API, then the functional one
  • Feed training with efficient tf.data pipelines
  • Write a custom training step when the standard loop is no longer enough
  • Monitor training with TensorBoard and callbacks
  • Export a model as a SavedModel and serve it

Prerequisites​

  • Python and NumPy
  • Deep learning fundamentals

Course modules​

  1. Tensors, variables and computation graphs
  2. Sequential API: a first model in a few lines
  3. Functional API: multiple inputs and branches
  4. Custom layers and Model subclassing
  5. tf.data: reading, transforming and prefetching
  6. Callbacks: checkpointing, early stopping, learning rate scheduling
  7. TensorBoard: tracking losses, weights and histograms
  8. Transfer learning with pretrained models
  9. Distributed training across several accelerators
  10. SavedModel and serving with TensorFlow Serving

Assessment and certificate​

The course ends with a 40-question exam covering every module. On success, a certificate of completion is issued; its number can be verified by anyone on the platform.

Free courses, by contrast, end with a 5-question quiz and a certificate preview, without certification.