medicure rag chatbot Download - medicure rag chatbot Source code download

medicure rag chatbot

Other source code

v0.0.1

Download

Medicure RAG Chatbot?

An AI chatbot implementing RAG technique with Meta-Llama2-7b Large Language Model, along with langchain and pinecone vector databases. Resource Used : The Gale Encyclopedia of Medicine

Screenshots

Technologies Used

Streamlit- WebApp UI
Pinecone - Vector Database
Langchain and sentence-transformers - RetrieveQAChain and Embedding Model
Meta Llama-2-7b-chat quantized Model - Large Language Model(LLM) from Hugging Face Hub

Solution Approach

Pinecone vector db stores the text_chunks embeddings generated from the Book Pdf. LangChain is used for building the LLMChain with promptTemplate to perform the similarity search from pinecone and then fine-grain the output with LLM.

Running Web App Locally

To run web app locally, follow these steps:

1.Clone the Repo:

git clone https://github.com/4darsh-Dev/medicure-rag-chatbot.git

Configure poetry:

  pip install poetry
  poetry init
  poetry shell

Install Requirements:
```
  poetry install
  
```
Run the Streamlit App:
```
poetry streamlit run app.py
```
Access Your App: After running the command, Streamlit will start a local web server and provide a URL where you can access your app. Typically, it will be something like http://localhost:8501. Open this URL in your web browser.
Stop the Streamlit Server: To stop the Streamlit server, go back to the terminal or command prompt where it's running and press Ctrl + C to terminate the server.