0.4.2 update

included bigger and more capable model. Responses are much better than the 1B model. But generation time is bit lengthy. Especially the first prompt. Requires a high end device with more than 8GB ram.