LiuQL2/MedicalChatbot ? reverse-engineered prompt

Build me a Python project that recreates a medical diagnosis chat system from a research paper. It should chat with a patient, ask about symptoms one by one, and then predict the most likely disease based on what the user says. The goal is to feel like a simple medical assistant that can keep track of the conversation, choose the next symptom to ask about, and make a final diagnosis at the end.

I already have the dataset, so wire the code to load the data from a folder path I can set easily. Make it easy to run from the command line, with a clear help screen showing the main options and how to train and test it. If you need to check anything unclear, look up the paper or current docs online. Keep the project clean and runnable, with sensible defaults and basic instructions so I can get it working quickly.