For this activity follow Neural machine translation with a Transformer and Keras.
Submit completed Colab notebook showing generated output.
The data used in this tutorial is a Portuguese-to-English translation dataset that contains about 52,000 training examples. The pre-processing for this data begins by tokenizing the text so that common words and parts of words are saved as tokens to use when analyzing text later.