Test NLP models in the wild!¶
Performance of NLP models should be measured not only in terms of well established metrics if we want to use them in the real life.
This module contains implementation of multiple functions designed to corrupt text in a way resembling naturally occuring mistakes (aspects).
- Quick Start Guide
- How to contribute