Session Ready
Exercise

Train the encoder decoder network

In the last exercise, you have created the Encoder and the Decoder networks separately. In this exercise, you'll combine them using Keras functional API to create the Encoder-Decoder architecture. You'll also compile and train the network.

You have the input vectors to the Encoder and Decoder saved in variables eng_input_data and fra_input_data respectively. You also have the target data available in variable target_data. The encoder and decoder input layers are saved in encoder_input and decoder_input. The decoder output layer is saved in decoder_out.

Instructions 1/2
undefined XP
  • 1
  • 2
  • Create the encoder-decoder model using Model().
  • Compile the model using 'rmsprop' optimizer and 'categorical_crossentropy' loss.
  • Print model summary.