The nlp-with-transformers/notebooks repository is a collection of Jupyter notebooks that accompany a published book on natural language processing with transformers.
The repository provides executable code examples aligned with chapters covering transformer-based NLP tasks. The notebooks are designed to run on cloud platforms like Google Colab or on local machines, though most chapters require GPU acceleration for practical execution times. The examples focus on PyTorch implementations, with the codebase actively maintained for that framework.
This resource suits developers learning transformer architectures and NLP techniques through hands-on practice, or those seeking reference implementations for common NLP workflows. The notebooks integrate with cloud platforms for accessibility without local setup overhead. Developers should be aware that TensorFlow code paths are no longer tested and may not run against current Keras versions, and that Chapter 7 on question answering relies on deprecated dependencies and is unmaintained; the README directs users to current Haystack documentation for that topic instead.
The project maintains PyTorch examples as its primary focus while acknowledging that TensorFlow sections have fallen out of sync with current library versions. One chapter has been explicitly marked as unmaintained due to upstream deprecations in its dependencies, with guidance provided to external documentation for current approaches.
