We evaluate many models used for biomedical and clinical nlp tasks, and train new models that perform much better. - View it on GitHub