Tensorflow2.0: What is the replacement of `tf.contrib.layers.fully_connected`?

Question

I want to follow this beginner guide , but I found it quite lot of obsoleted and deprecated package/class/function.

So, as question title ask, who know how to replace the ...layers.fully_connected to what function in TF2.0?

Or you think the guide I follow is too old to practice, would you please guide me some blog/guide for TF2.0 or the up-to-date one.

Answer 1

yeah that's too old to practice with i recommend o'reilly 'hands on machine learning with sckitlearn, keras and tensorflow' it's all good and new but for your problem, when you use dense or any other type it will connect it fully and completely, for example:

model.add(keras.layers.Dense(300,activation="relu"))