Python environment setup
python3 -m venv cuda
source cuda/bin/activate
pip3 install -r requirements.txt
Jupyter notebooks:
python3 -m ipykernel install --user --name=cuda --display-name "cuda-gpt"
python3 -m notebook --ip=0.0.0.0 --port=8000
Connect web browser to
http://localhost:8000, open notebook, set kernel to cuda-gpt for gpu access.
