24  Downloads Updated  3 months ago

Latxa is based on Llama-3.1, which is trained on a Basque corpus comprising 4.3M documents and 4.2B tokens using language adaptation techniques.