Getting started

My Dockerhub.

Clone the repository and setup the correct folder structure.

git clone https://github.com/HannahKniesel/HowToDockern.git

cd HowToDockern
mkdir workspace/data
mkdir workspace/output

Add you wandb API key to the .env file in the workspace directory.

Build the docker image

docker build -t <image-name> .

Run the docker image (locally - for development)

docker run --gpus all -it --rm --ipc=host -v /mnt/e/Documents/UbuntuCode/1_DockerTest/HowToDockern/workspace/:/workspace/ --name <container-name> <image-name>

Push the docker to dockerhub

docker tag <image-name> <user-name>/<image-name>:<version>
docker push <user-name>/<image-name>:<version>

Use on cluster

Your startscript.sh should look like this:

python classifier.py

Then you can submit it as follows:

submit ./startscript.sh --name <name> --custom <user-name>/<image-name>:<version>