QuantiphiAssessment

Colab Notebook for quick testing Colab: QA-RAG Pipeline

API & Gradio setup for different environments

1. Local

For development in local using VS code. Run below commands. This will create virtual environment, activate it, install packages

Prerequisites - Python 3.12.2, pip

Windows	Mac/Linux
python -m venv env .\env\Scripts\activate pip install -r .\requirements.txt	python -m venv env source env/bin/activate pip install -r requirements.txt

Now run

1. python main.py file to start FastAPI server and check http://127.0.0.1:8181 OR
2. python gradio_app.py for Gradio app and check http://127.0.0.1:8181

2. Using docker on WINDOWS

From root directory run below command for fastapi server

docker-compose -f docker-compose-local.yml --env-file env_var_files/.env.local up -d

NOTE: From Dockerfile - Commnent & Uncomment below code to switch FastAPI with Gradio App

EXPOSE ${PORT}
CMD  uvicorn main:app --host ${HOST} --port ${PORT}

# EXPOSE ${GRADIOPORT}
# CMD python gradio_app.py

Note

1. Have not finished building the fully parameterised app for creating pipelines due to insufficient system RAM. You can adjust all necessary settings in env_var_files/.env.local to run app as desired. Going forward we can predfined the embedding & llm models, enabling us to utilise different input files to create a new vector database collection for conducting Q&A.
2. Hugging Face embedding model all-mpnet-base-v2
3. Hugging Face LLM model for QnA google/gemma-1.1-2b-it
4. Check postman collection to make request