Yes, it's somewhat similar to the 2B model as it uses the same vocabulary size.