Berlin LLM Hackathon

Code for the Berlin LLM Hackathon organized by Weaviate. Evaluating multiple retreival strategies for a Q&A bot using MLOps slack messages

Make sure to create an .env file with your environement variable

OPENAI_API_KEY=sk-xxxx
COHERE_API_KEY=xxx
WEVIATE_URL=https://xxxx.weaviate.network
OPENAI_ORGANIZATION=org-xxxx

Local setup

Create environment

Create a new virtual environment and install the requirements

python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Download the data

python download_data.py

Run the ingestion pipeline

python ingest.py

Running locally

python app.py "What is the best way to get started with MLOps?"

Running in agenta

Installing Agenta

You need to install the agenta platform locally. You can find the instructions in the github repository:

https://github.com/agenta-ai/agenta

and in the documenation

https://docs.agenta.ai

Running the app in the agenta web-ui

agenta variant serve

This deploys the app locally. You can access the api here: http://localhost/llm_hackathon/original/ with its documentation in http://localhost/llm_hackathon/original/docs

You can experiment with your app in the playground. 🎮 Go to: http://localhost:3000/apps/llm_hackathon/playground