TensorFlow 2.0 Beta is available Learn more

Research and experimentation

Eager execution provides an imperative, define-by-run interface for advanced operations. Write custom layers, forward passes, and training loops with auto differentiation. Start with these notebooks, then read the eager execution guide.

  1. Eager execution
  2. Automatic differentiation and gradient tape
  3. Custom training: basics
  4. Custom layers
  5. Custom training: walkthrough