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