Now that the model is properly trained, it may be used to generate text output. By examining the prompt a user presents, the language model can realize what is currently being stated or questioned. From there, the model works by using the information gathered, preprocessed, and examined in instruction to locate and best craft a meaningful reaction.