Deep Learning Docker Image
MIT License
Don't waste time on setting up a deep learning environment while you can get a deep learning environment with everything pre-installed.
Variant | Tag | Conda | PyTorch | TensorFlow | Image size |
---|---|---|---|---|---|
Conda | conda |
✔️ | ❌ | ❌ | |
Tensorflow | tf |
❌ | ❌ | ✔️ | |
PyTorch | torch |
❌ | ✔️ | ❌ | |
PyTorch + Tensorflow |
tf-torch , latest
|
❌ | ✔️ | ✔️ | |
PyTorch + Tensorflow + Conda | tf-torch-conda |
✔️ | ✔️ | ✔️ |
You can see the full list of tags https://hub.docker.com/r/matifali/dockerdl/tags.
docker run --gpus all --rm -it -h dockerdl matifali/dockerdl bash
docker run --gpus all --rm -it -h dockerdl -p 8888:8888 matifali/dockerdl jupyter lab --no-browser --port 8888 --ServerApp.token='' --ip='*'
Connect by opening http://localhost:8888 in your browser.
git clone https://github.com/matifali/dockerdl.git
Modify the corresponding [Dockerfile]
to add or delete packages.
[!NOTE] You may have to rebuild the
dockerdl-base
if you are building a custom image and then use it as a base image. See Build section.
The following --build-arg
are available for the dockerdl-base
image.
Argument | Description | Default | Possible Values |
---|---|---|---|
USERNAME |
User name | coder |
Any string or $USER
|
USERID |
User ID | 1000 |
$(id -u $USER) |
GROUPID |
Group ID | 1000 |
$(id -g $USER) |
CUDA_VER |
CUDA version | 12.4.1 |
|
UBUNTU_VER |
Ubuntu version | 22.04 |
22.04 , 20.04 , 18.04
|
[!WARNING] Not all combinations of
--build-arg
are tested.
Build the base image
docker build -t dockerdl-base:latest --build-arg USERNAME=coder --build-arg CUDA_VER=12.4.1 --build-arg UBUNTU_VER=22.04 -f base.Dockerfile .
Build the image you want with the base image as the base image.
docker build -t dockerdl:tf --build-arg TF_VERSION=2.12.0 -f tf.Dockerfile .
or
docker build -t dockerdl:torch --build-arg -f torch.Dockerfile .
Follow the instructions here.
If you find any issue please feel free to create an issue and submit a PR.
[^1]: This image is based on nvidia/cuda and uses nvidia-container-toolkit to access the GPU.