I'd like to introduce sequence-classifier, a PyTorch library for sequence classification tasks such as named-entity recognition and part-of-speech tagging. It's designed for ease of use with Transformer-based models (BERT, RoBERTa, DeBERTa). It has a special focus on Conditional random fields (CRFs).

Key features include:

- Support for basic CRF functionality - Support for computing marginal log-likelihood and marginal probability - Support for an efficient computational algorithm on GPU based on Rush, 2020 (https://aclanthology.org/2020.acl-demos.38/)

You are invited to check out the GitHub repo, provide feedback, or ask any questions. If you find it useful, please give it a star!

GitHub link: https://github.com/yasufumy/sequence-classifier Link to example code: https://github.com/yasufumy/sequence-classifier/blob/main/ex...