Medical-ChatBot-using-llama-2

Project - Medical ChatBot

Architecture :

Backend :

1. Data Ingestion : Medical book (Pdf file)
2. Extract data
3. Create different text chunks (break down big content into small chunks(parts) for to feed into model)
4. Embedding - vector
5. Build semantic index - (Connecting vector)
6. Knowledge base - (pinecone, vector store)

Frontend :

1. User questions -> Query embedding -> Knowledge base
2. Knowledge base -> Ranked result -> LLM model (llam2) -> user answer

Tech stacks:

1. Language - Python
2. Framework - Langchain
3. Frontend/webapp - Flask, HTML, CSS
4. LLM - meta llama 2
5. Vector DB - pinecone.

Description

This project is a Medical ChatBot built using the Llama-2 model. The primary goal of this chatbot is to provide accurate and relevant medical information to users based on their queries.

The architecture of the project is divided into two main parts: the backend and the frontend.

Backend:

1. Data Ingestion: The backend starts with ingesting data from a medical book in PDF format.
2. Data Extraction: The data from the PDF is then extracted and processed.
3. Text Chunking: The extracted data is broken down into smaller chunks or parts. These chunks are then fed into the model.
4. Embedding: Each chunk of text is converted into a vector representation, also known as embedding.
5. Semantic Index Building: A semantic index is built to connect these vectors, which aids in understanding the context and meaning of the text chunks.
6. Knowledge Base: The embeddings are stored in a knowledge base using Pinecone, a vector database.

Frontend:

1. User Query Processing: When a user asks a question, it is converted into a query embedding.
2. Knowledge Base Lookup: This query embedding is used to search the knowledge base for the most relevant information.
3. LLM Model Processing: The ranked results from the knowledge base are then passed through the LLM (Llama-2) model.
4. User Answer Generation: Finally, the model generates an answer to the user's question based on the information it has processed.

The tech stack used for this project includes:

1. Python: The primary programming language used for developing the chatbot.
2. Langchain: A framework used for building the chatbot.
3. Flask, HTML, CSS: These technologies are used for building the frontend or web application of the chatbot.
4. Meta Llama-2: The language model used for processing and generating responses to user queries.
5. Pinecone: A vector database used for storing the text embeddings.

In summary, this Medical ChatBot project leverages advanced NLP techniques and a robust tech stack to provide users with accurate medical information based on their queries. It represents a significant step forward in the field of AI-driven healthcare solutions.

Output