Exercise

Building a training loop

Let's write a simple training loop from scratch!

The pipeline you've created in the previous exercise is available as the nlp object. It already contains the entity recognizer with the added label 'GADGET'.

The small set of labelled examples that you've created previously is available as the global variable TRAINING_DATA. To see the examples, you can print them in your script or in the IPython shell. spacy and random have already been imported for you.

Instructions 1/3

XP

1
2
3

Call nlp.begin_training, create a training loop for 10 iterations and shuffle the training data.

Take Hint (-15 XP)