The notebook examples are available in the notebooks
folder. To run the examples, you need to have the following dependencies installed:
- Python 3.10 or higher
- Docker
- GPU with CUDA support (optional). If you don't have a GPU, you can still run the examples, but it will be slower. Adjust scripts to run on CPU.
To run the examples, follow these steps:
- Clone the repository
- Open a terminal and navigate to the repository folder
- Run the following command to start the Jupyter notebook server:
docker compose up -d jupyter-notebook
- Pull the model and be set up to run:
sh startup.sh
or
docker compose exec -d ollama-server ollama pull llama3:instruct
To-do:
- Add request to backend API
- Address issues with the model
- Integrate solutions with Chatbot API