byw·11 miesięcy temu·discussI wonder if it has more to do with larger training data than the languages themselves.