The purpose of this container is to create a Deep Learning Conda environment with Jupyter Lab.
Here are some Deep Learning packages installed.
- TensorFlow v2.0
- PyTorch v1.3
- MXNet
- spaCy
- AllenNLP
Note that this container requires nvidia-docker2. You will also need CUDA 10.0, cuDNN 7, and NCCL 2.3 installed on your host computer.
On Ubuntu 19+, nvidia-docker2
is not yet published. You need to do the following.
distribution=ubuntu18.04
curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | \
sudo tee /etc/apt/sources.list.d/nvidia-docker.list
sudo apt-get update && sudo apt-get install -y nvidia-container-toolkit
sudo systemctl restart docker
# check runtime hook is installed
which nvidia-container-runtime-hook
# make sure container detects gpu
docker run --gpus all --rm nvidia/cuda nvidia-smi
Build it.
./build.sh
Run it.
docker run -it \
--shm-size=5g \
--gpus all \
-p 8888:8888 \
-p 6006:6006 \
conda-deeplearning:local
Run it with a mounted host folder.
docker run -it \
-v $HOME/git/docker-containers/conda-deeplearning/ipynb:/root/ipynb \
-p 8888:8888 \
-p 6006:6006 \
--shm-size=5g \
--gpus all \
conda-deeplearning:local
Run it with password protection.
docker run -it \
-v $HOME/git/docker-containers/conda-deeplearning/ipynb:/root/ipynb \
-p 8888:8888 \
-p 6006:6006 \
-e NOTEBOOK_PASSWORD=sha1:6676da7235c8:9c7d402c01e330b9368fa9e1637233748be11cc5 \
--shm-size=5g \
--gpus all \
conda-deeplearning:local
Observe it.
- GitHub NVIDIA Docker
- NVIDIA CUDA docker image
- Anaconda Install Scripts
- Install Anaconda in silent mode
- Include .whl installation in requirements.txt
- Tensorflow Install GPU
Check out Alonzo Church.
@misc{oneoffcoder_conda_deeplearning_2019,
title={Docker container for Deep Learning with Jupyter Lab},
url={https://github.com/oneoffcoder/docker-containers/tree/master/conda-deeplearning},
journal={GitHub},
author={One-Off Coder},
year={2019},
month={Jul}}