geezlab/gpu-deep-learning-notebook repository overview

⁠Content

• Nvidia CUDA 9.0
• TensorFlow
• PyTorch
• Keras
• NLP (spaCy, nltk, torchtext)
• Jupyter Notebook, Lab, Hub
• Machine learning libraries

⁠Pull Container Image

This can take a while depending on your internet connection as it is a fairly large container.

docker pull geezlab/gpu-deep-learning-notebook

⁠Run Container

Create an ephemeral container named dl-app, which will mount your current working directory to ~/dev within the container. Note that this container will be destroyed once Jupyter is stopped if you don't like this approach replace --rm with -d in the following command.

docker run --rm -p 10000:8888 -e JUPYTER_ENABLE_LAB=yes --runtime="nvidia" -v "$PWD":/home/geez/dev --name dl-app geezlab/gpu-deep-learning-notebook:latest

⁠Jupyter in Browser

The previous command will print a full URL with a token that is use to access Jupyter in browser as follows:

`http://<host-ip>:10000/?token=<token>`

That's it, happy coding!