It is a PyTorch-based framework developed by Facebook to give ease to the experience in the implementation of the end-to-end workflow of Natural Language Processing. It is designed to reduce the gap between practicing research and serving the models of an NLP app.
Different Features of PyText
- It is based on torch text.
- It provides supports of distributed training.
- It also includes support of multi-tasking.
PyText Different Uses
It can be used to implement different tasks of NLP -
- For implementing text classifiers.
- For implementing sequence taggers.
- For implementing the Joint intent -slot model.
- For implementing contextual intent -slot models.