Table of contents

jupyterlab-gpu — bare JupyterLab on GPU

A clean JupyterLab environment on top of CUDA 12 — no opinionated ML libs preinstalled, just Jupyter and Python. Bring your own pip install.

Image tag

docker.io/manvarharsh/jupyterlab-gpu:cuda12

What’s in this image

  • Base: nvidia/cuda:12.4.1-cudnn-devel-ubuntu22.04
  • Python 3.10 (conda)
  • JupyterLab + JupyterHub
  • Podstack authenticator
  • OpenSSH server
  • Build essentials and common CUDA dev headers

Default ports

PortService
22SSH
8000JupyterHub

Use cases

  • A blank-slate notebook environment when none of the preset images fits
  • Teaching / coursework where students install their own stack
  • Quick sandbox for trying a new library on GPU

Environment variables

VariableDescription
ENABLE_SSHEnable SSH server
ENABLE_JUPYTERHUBEnable JupyterHub on port 8000
PODSTACK_API_URLBackend URL for JupyterHub token validation
SSH_PUBLIC_KEYPublic key for SSH

Persistence

Mount at /data. Notebooks and any installed packages (pip install --target=/data/site-packages) live there.

See also