As we said TensorFlow had brought the Keras API into the TensorFlow project and by doing so you get the following:
- Keras compatibility tf.keras module is introduced in TensorFlow as an implementation of the Keras specifications that's built from the ground up for TensorFlow
- Also, they are introducing into core TensorFlow new data structures like layers
- Also, models which are containers for the graph of layers used to be Keras data structures and now they will be shared between core TensorFlow and this tf.keras module
- Finally, it's fully compatible with all the TensorFlow advanced features like the experiment API.
So if it happens that you are a Keras a TensorFlow user, how will you get affected by such ...