Template for creating a Docker container that provides access to a GPU for Pytorch
Serves a simple Docker setup for local, containerized, Pytorch + GPU development with Jupyter.
The main thing to remember to change is the absolute path to the notebooks folder to your local
path.
Otherwise, this can be cloned and used as is.
First, you'll need to make sure that you have the appropriate NVIDIA drivers already installed. If you're running this in WSL, you'll also need to toggle the relevant virtualization item in your BIOS - don't worry, this is usually pretty easy!
Once you have the prerequisites installed, you can clone this repo locally, then specify the absolute path to your notebooks directory in the docker-compose.yaml file.
After that, use docker-compose from your command line to start the repo:
docker-compose up -d --buildI prefer running the container in the background with the -d flag, but you can do whatever you want!
If you want to run the notebook or prototype code using jupyter, all you need to do is open a new tab in your browser and navigate to localhost:8888 and you will have access to a Jupyter Lab interface.