French language models. These model were trained in early 2021 following the then scaling laws and using the exact same training data as the CamemBERT