/dlcontainer

A docker image for deep learning in fMRI. GPU supported. Pytorch, Keras, Tensorflow, Nilearn, Pymc3

Primary LanguageDockerfile

Dlcontainer

This docker image is designed for GPU supported(only) deep learning for common usage and fMRI and EEG analysis. The former settings of jupyter notebook was guided from my friend @lizequn. Now it is more convenient. If you want to apply it as a remote python interpreter in pycharm, try to modify the docker daemon.json file add the following lines:

{
"default-runtime" = nvidia,
"runtimes": {
        "nvidia": {
            "path": "/usr/bin/nvidia-container-runtime",
            "runtimeArgs": []
        }
    }
}

Then set the python interpreter in pycharm to the specific docker image. More fMRI tools would be added to this image in the future.

Such settings are not recommended.

How to build

Just clone this project and run the cli:

docker build -t <image-name:tags> .

How to run

docker run --rm -d --gpus all -p 8888:8888 -p 6006:6006 -e PASSWORD=yourPasswd -v yourpath:/notebook <image-name:tags>

Notice

Expample cli commend is:

docker run --rm -d --gpus all -p 8888:8888 -p 6006:6006 -e PASSWORD=yourPasswd -v yourpath:/notebook xin0214/dlcontianer:latest