I’m currently doing token classification on french biomedical data for my intership. I tried different model such as Bert, Camembert-base, Camembert-large and achieve 70% F1 score. I see a research paper where they do similar stuff and then they increase performance using a voting strategy between 5 of their model. Does someone know how i can do it ? I know in sklearn there is VotingClassifier() fonction for this kind of stuff, does hugging face have something similar ?