There are good instructions on how train a model using torchtext. But how do I preper it to production? How to create a PREDICT pipeline?
