MediBot

This repository hosts a Context-Aware Medical Chatbot developed with the Llama 3.2 LLM fine-tuned on oncology-specific patient-doctor dialogues. The chatbot integrates with the LangChain Retrieval-Augmented Generation (RAG) framework, leveraging cancer-specific PDFs embedded in a Chroma Vector Database to provide accurate, contextual responses. Using Low-Rank Adaptation (LoRA) and Parameter-Efficient Fine-Tuning (PEFT), the model is optimized for memory and performance, with 8-bit quantization enhancing efficiency. The chatbot is designed as a multi-turn, memory-based conversational AI, supporting patients and healthcare providers with reliable cancer-related insights. Skills utilized include text generation, model quantization, RAG for information retrieval, and Hugging Face Transformers for model deployment.