These implementations are for demonstration purposes. They are less efficient than the implementations in the Python standard library.
In what follows we explain how you can download and use our pretrained XLM (English-only) BERT model. Then we explain how you can train your own monolingual model, and how you can fine-tune it on the ...