A practical guide demonstrates fine-tuning DoctoBERT, a French medical language encoder pretrained on the FineMed corpus, across three clinical natural language processing tasks. It provides working code for named entity recognition on the QUAERO dataset, multi-label text classification on the MORFITT dataset, and semantic similarity scoring using sentence transformers on CLISTER pairs. The guide argues that compact encoder models remain the right tool for document-reading clinical tasks that do not require text generation, citing advantages in cost, inference speed, and the ability to deploy fully on-premises compared with large generative language models.
