Lambda Stack:
One Line Install for TensorFlow GPU on Ubuntu 18.04

One line Lambda Stack Install:

LAMBDA_REPO=$(mktemp) && \
wget -O${LAMBDA_REPO} https://lambdal.com/static/files/lambda-stack-repo.deb && \
sudo dpkg -i ${LAMBDA_REPO} && rm -f ${LAMBDA_REPO} && \
sudo apt-get update && sudo apt-get install -y lambda-stack-cuda

Congrats! You've now installed Lambda Stack! You now have TensorFlow, PyTorch, and Caffe with full GPU support. Your machine learning frameworks, CUDA libraries, NVIDIA GPU drivers, and other system libraries are all taken care of. If you ever need to upgrade your system, just run:

sudo apt-get update && sudo apt-get upgrade

Installing on Ubuntu 16.04 or 18.04 Server Edition? Use this:

This should be used when installing on a server as it doesn't include any graphical user interface packages.

LAMBDA_REPO=$(mktemp) && \
wget -O${LAMBDA_REPO} https://lambdal.com/static/files/lambda-stack-repo.deb && \
sudo dpkg -i ${LAMBDA_REPO} && rm -f ${LAMBDA_REPO} && \
sudo apt-get update && \
sudo apt-get --yes upgrade && \
sudo apt-get install --yes --no-install-recommends lambda-server && \
sudo apt-get install --yes --no-install-recommends nvidia-390 libcuda1-390 nvidia-opencl-icd-390 && \
sudo apt-get install --yes --no-install-recommends lambda-stack-cuda

Want an un-attended installation for Docker, Ansible, or shell scripts?

This will let you install the entire lambda stack within a script without being interrupted by the license prompt. Installation this way assumes that you have read and agree to the CUDNN license.

LAMBDA_REPO=$(mktemp) && \
wget -O${LAMBDA_REPO} https://lambdal.com/static/files/lambda-stack-repo.deb && \
sudo dpkg -i ${LAMBDA_REPO} && rm -f ${LAMBDA_REPO} && \
sudo apt-get update && \
sudo apt-get --yes upgrade && \
echo "cudnn cudnn/license_preseed select ACCEPT" | sudo debconf-set-selections && \
sudo apt-get install --yes --no-install-recommends lambda-server && \
sudo apt-get install --yes --no-install-recommends nvidia-390 libcuda1-390 nvidia-opencl-icd-390 && \
sudo apt-get install --yes --no-install-recommends lambda-stack-cuda

Lambda Stack is:

  • Frameworks: TensorFlow, PyTorch, Caffe, Caffe2, Keras
  • No longer worrying about your LD_LIBRARY_PATH or your PATH
  • GPU Software: CUDA, CuDNN, NVIDIA Drivers
  • Dev Tools: git, tmux, screen, vim, emacs, htop, valgrind, build-essential
  • Installed system-wide (works on every user!)
  • Designed for Ubuntu 16.04 LTS and Ubuntu 18.04 LTS

What does Lambda Stack Include?

Lambda Stack installs the following libraries for your system. Compiled with GPU support.

Benefits of Lambda Stack

Save weeks of setup time

You’ll be ready to start training just moments after the installation.

Make money

Calculate how many hours you'll save after you stop futzing with your Linux install and multiply that by your hourly rate!

Simple and scalable

Lambda Stack works on notebooks, workstations, servers, cloud instances, and more.

Once again, how to install Lambda Stack:

LAMBDA_REPO=$(mktemp) && \
wget -O${LAMBDA_REPO} https://lambdal.com/static/files/lambda-stack-repo.deb && \
sudo dpkg -i ${LAMBDA_REPO} && rm -f ${LAMBDA_REPO} && \
sudo apt-get update && sudo apt-get install -y lambda-stack-cuda

Give us feedback.

Have a question or suggestion on how we can improve Lambda Stack? Send us an email:

software@lambdalabs.com