This model is a fine-tuned model based on the Llama 2_7b architecture. It has been specifically trained on a dataset comprising USMLE (United States Medical Licensing Examination) questions and answers, as well as conversations between doctors and patients.