docker

All-in-one Jupyter Docker image for GPU Deep Learning using PyCUDA, PyTORCH, CUDA etc.

Build and Run the GPU image (see below for more info)

docker build -t quantscientist/pycuda -f Dockerfile.gpu3 .

nvidia-docker run -it -p 5555:5555 -p 7842:7842 -p 8787:8787 -p 8786:8786 -p 8788:8788 -v ~/db/Dropbox/dev2/:/root/sharedfolder quantscientist/pycuda bash

This repository includes utilities to build and run my NVIDIA Docker image for the Deep Learning School: https://www.meetup.com/TensorFlow-Tel-Aviv/events/241762893/

NOTE: Building this image may take several hours since CMAKE is being built from source. https://github.com/QuantScientist/deep-ml-meetups

Also available on docker hub (Build on docker hub usually failes because of the long build time): https://hub.docker.com/r/quantscientist/deep-learning-boot-camp/

docker pull quantscientist/deep-learning-boot-camp

Please be aware that this project is currently experimental.

CUDA requirements

Running a CUDA container requires a machine with at least one CUDA-capable GPU and a driver compatible with the CUDA toolkit version you are using.

NVIDIA drivers are backward-compatible with CUDA toolkits versions:

CUDA toolkit version	Minimum driver version
7.0	>= 346.46
7.5	>= 352.39

** We use CUDA 8.0. **

Get the toolkit:

sudo apt-get install nvidia-cuda-toolkit

Get nsight IDE:

sudo apt-get install nvidia-nsight

Install nvidia-docker and nvidia-docker-plugin

wget -P /tmp https://github.com/NVIDIA/nvidia-docker/releases/download/v1.0.1/nvidia-docker_1.0.1-1_amd64.deb sudo dpkg -i /tmp/nvidia-docker*.deb && rm /tmp/nvidia-docker*.deb

Get nvidia docker (requires docker-engine NOT docker.io):

nvidia-docker run --rm nvidia/cuda nvidia-smi

Image contents

On top of all the fancy deep learning libraries, this docker image contains:

Ubuntu 16.04 CUDA 8.0 (GPU version only) cuDNN v5 (GPU version only) Tensorflow Theano Keras iPython/Jupyter Notebook Numpy, SciPy, Pandas, Scikit Learn, Matplotlib A few common libraries used for deep learning

• ArrayFire
• PyCUDA
• Python
• LLVM
• LLDB
• Snappy
• Numba

Build the image

GPU version

docker build -t quantscientist/pycuda -f Dockerfile.gpu3 .

CPU version

docker build -t quantscientist/pycuda -f Dockerfile.cpu .

Run the GPU image

nvidia-docker run -it -p 5555:5555 -p 7842:7842 -p 8787:8787 -p 8786:8786 -p 8788:8788 -v ~/db/Dropbox/dev2/:/root/sharedfolder quantscientist/pycuda bash

Run the GPU image

docker run -it -p 5555:5555 -p 7842:7842 -p 8787:8787 -p 8786:8786 -p 8788:8788 -v /myhome/data-science/:/root/sharedfolder --env="DISPLAY" quantscientist/pycuda bash

Run Jupyter

chmod +x run_jupyter.sh ./run_jupyter.sh

OR

docker build -t quantscientist/gpu -f Dockerfile.gpu .

Issues and Contributing

• Please let us know by filing a new issue
• You can contribute by opening a pull request
  You will need to send a signed copy of the Contributor License Agreement to [email protected] before your change can be accepted.