gw0 / docker-keras

Docker image: Keras in Docker for reproducible deep learning on CPU or GPU

docker-keras

docker-keras is a minimal Docker image built from Debian 9 (amd64) for reproducible deep learning based on Keras. It features minimal images for Python 2 or 3, TensorFlow, Theano, or CNTK backends, processing on CPU or GPU, and uses only Debian and Python packages (no manual installations). Each tag is using the latest released versions at a specific date.

Open source project:

• home: http://gw.tnode.com/docker/keras/
• github: http://github.com/gw0/docker-keras/
• technology: debian, keras, tensorflow, theano, cntk, cuda toolkit, openblas, python, numpy, h5py
• docker hub: https://hub.docker.com/r/gw000/keras/

Available tags:

• 2.1.4-py2, 2.1.4-cpu, 2.1.4, latest points to 2.1.4-py2-tf-cpu
• 2.1.4-py3 points to 2.1.4-py3-tf-cpu
• 2.1.4-gpu points to 2.1.4-py2-tf-gpu
• 2.1.4-py2-tf-cpu/2.1.4-py2-tf-gpu/2.1.4-py3-tf-cpu/2.1.4-py3-tf-gpu [2018-02-15]: Python 2.7/3.5 + Keras (2.1.4) + TensorFlow (1.5.0) on CPU/GPU (Dockerfile.py2-tf-cpu/.py2-tf-gpu/.py3-tf-cpu/.py3-tf-gpu)
• 2.1.4-py2-th-cpu/2.1.4-py2-th-gpu/2.1.4-py3-th-cpu/2.1.4-py3-th-gpu [2018-02-15]: Python 2.7/3.5 + Keras (2.1.4) + Theano (1.0.1) on CPU/GPU (Dockerfile.py2-th-cpu*/.py2-th-gpu.py3-th-cpu/.py3-th-gpu)
• 2.1.4-py2-cntk-cpu/2.1.4-py2-cntk-gpu/2.1.4-py3-cntk-cpu/2.1.4-py3-cntk-gpu [2018-02-15]: Python 2.7/3.5 + Keras (2.1.4) + CNTK (2.4) on CPU/GPU (Dockerfile.py2-cntk-cpu*/.py2-cntk-gpu.py3-cntk-cpu/.py3-cntk-gpu)
• 2.1.3-py2-tf-cpu/2.1.3-py2-tf-gpu/2.1.3-py3-tf-cpu/2.1.3-py3-tf-gpu [2018-01-17]: Python 2.7/3.5 + Keras (2.1.3) + TensorFlow (1.4.1) on CPU/GPU (Dockerfile.py2-tf-cpu/.py2-tf-gpu/.py3-tf-cpu/.py3-tf-gpu)
• 2.1.3-py2-th-cpu/2.1.3-py2-th-gpu/2.1.3-py3-th-cpu/2.1.3-py3-th-gpu [2018-01-17]: Python 2.7/3.5 + Keras (2.1.3) + Theano (1.0.1) on CPU/GPU (Dockerfile.py2-th-cpu*/.py2-th-gpu.py3-th-cpu/.py3-th-gpu)
• 2.1.3-py2-cntk-cpu/2.1.3-py2-cntk-gpu/2.1.3-py3-cntk-cpu/2.1.3-py3-cntk-gpu [2018-01-17]: Python 2.7/3.5 + Keras (2.1.3) + CNTK (2.3) on CPU/GPU (Dockerfile.py2-cntk-cpu*/.py2-cntk-gpu.py3-cntk-cpu/.py3-cntk-gpu)
• 2.1.1-py2-tf-cpu/2.1.1-py2-tf-gpu/2.1.1-py3-tf-cpu/2.1.1-py3-tf-gpu [2017-12-01]: Python 2.7/3.5 + Keras (2.1.1) + TensorFlow (1.4.0) on CPU/GPU
• 2.1.1-py2-th-cpu/2.1.1-py2-th-gpu/2.1.1-py3-th-cpu/2.1.1-py3-th-gpu [2017-12-01]: Python 2.7/3.5 + Keras (2.1.1) + Theano (1.0.0) on CPU/GPU
• 2.1.1-py2-cntk-cpu/2.1.1-py2-cntk-gpu/2.1.1-py3-cntk-cpu/2.1.1-py3-cntk-gpu [2017-12-01]: Python 2.7/3.5 + Keras (2.1.1) + CNTK (2.3) on CPU/GPU
• 2.0.8-py2-tf-cpu/2.0.8-py2-tf-gpu/2.0.8-py3-tf-cpu/2.0.8-py3-tf-gpu [2017-09-14]: Python 2.7/3.5 + Keras (2.0.8) + TensorFlow (1.3.0) on CPU/GPU
• 2.0.8-py2-th-cpu/2.0.8-py2-th-gpu/2.0.8-py3-th-cpu/2.0.8-py3-th-gpu [2017-09-14]: Python 2.7/3.5 + Keras (2.0.8) + Theano (0.9.0) on CPU/GPU
• 2.0.8-py2-cntk-cpu/2.0.8-py2-cntk-gpu/2.0.8-py3-cntk-cpu/2.0.8-py3-cntk-gpu [2017-09-14]: Python 2.7/3.5 + Keras (2.0.8) + CNTK (2.1) on CPU/GPU
• 2.0.6-py2-tf-cpu/2.0.6-py2-tf-gpu/2.0.6-py3-tf-cpu/2.0.6-py3-tf-gpu [2017-07-13]: Python 2.7/3.5 + Keras (2.0.6) + TensorFlow (1.2.1) on CPU/GPU
• 2.0.6-py2-th-cpu/2.0.6-py2-th-gpu/2.0.6-py3-th-cpu/2.0.6-py3-th-gpu [2017-07-13]: Python 2.7/3.5 + Keras (2.0.6) + Theano (0.9.0) on CPU/GPU
• 2.0.5-py2-tf-cpu/2.0.5-py2-tf-gpu/2.0.5-py3-tf-cpu/2.0.5-py3-tf-gpu [2017-06-13]: Python 2.7/3.5 + Keras (2.0.5) + TensorFlow (1.1.0) on CPU/GPU
• 2.0.5-py2-th-cpu/2.0.5-py2-th-gpu/2.0.5-py3-th-cpu/2.0.5-py3-th-gpu [2017-06-13]: Python 2.7/3.5 + Keras (2.0.5) + Theano (0.9.0) on CPU/GPU
• 2.0.4-py2-tf-cpu/2.0.4-py2-tf-gpu/2.0.4-py3-tf-cpu/2.0.4-py3-tf-gpu [2017-05-01]: Python 2.7/3.5 + Keras (2.0.4) + TensorFlow (1.1.0) on CPU/GPU
• 2.0.4-py2-th-cpu/2.0.4-py2-th-gpu/2.0.4-py3-th-cpu/2.0.4-py3-th-gpu [2017-05-01]: Python 2.7/3.5 + Keras (2.0.4) + Theano (0.9.0) on CPU/GPU
• 2.0.3-py2-tf-cpu/2.0.3-py2-tf-gpu/2.0.3-py3-tf-cpu/2.0.3-py3-tf-gpu [2017-04-19]: Python 2.7/3.5 + Keras (2.0.3) + TensorFlow (1.0.1) on CPU/GPU
• 2.0.3-py2-th-cpu/2.0.3-py2-th-gpu/2.0.3-py3-th-cpu/2.0.3-py3-th-gpu [2017-04-19]: Python 2.7/3.5 + Keras (2.0.3) + Theano (0.9.0) on CPU/GPU
• 2.0.2-py2-tf-cpu/2.0.2-py2-tf-gpu/2.0.2-py3-tf-cpu/2.0.2-py3-tf-gpu [2017-03-27]: Python 2.7/3.5 + Keras (2.0.2) + TensorFlow (1.0.1) on CPU/GPU
• 2.0.2-py2-th-cpu/2.0.2-py2-th-gpu/2.0.2-py3-th-cpu/2.0.2-py3-th-gpu [2017-03-27]: Python 2.7/3.5 + Keras (2.0.2) + Theano (0.9.0) on CPU/GPU
• 2.0.0-py2-tf-cpu/2.0.0-py2-tf-gpu/2.0.0-py3-tf-cpu/2.0.0-py3-tf-gpu [2017-03-15]: Python 2.7/3.5 + Keras (2.0.0) + TensorFlow (1.0.1) on CPU/GPU
• 2.0.0-py2-th-cpu/2.0.0-py2-th-gpu/2.0.0-py3-th-cpu/2.0.0-py3-th-gpu [2017-03-15]: Python 2.7/3.5 + Keras (2.0.0) + Theano (0.8.2) on CPU/GPU
• 1.2.2-py2-tf-cpu/1.2.2-py2-tf-gpu/1.2.2-py3-tf-cpu/1.2.2-py3-tf-gpu [2017-02-17]: Python 2.7/3.5 + Keras (1.2.2) + TensorFlow (1.0.0) on CPU/GPU
• 1.2.2-py2-th-cpu/1.2.2-py2-th-gpu/1.2.2-py3-th-cpu/1.2.2-py3-th-gpu [2017-02-17]: Python 2.7/3.5 + Keras (1.2.2) + Theano (0.8.2) on CPU/GPU
• 1.2.1-py2-tf-cpu/1.2.1-py2-tf-gpu/1.2.1-py3-tf-cpu/1.2.1-py3-tf-gpu [2017-01-20]: Python 2.7/3.5 + Keras (1.2.1) + TensorFlow (0.12.1) on CPU/GPU
• 1.2.1-py2-th-cpu/1.2.1-py2-th-gpu/1.2.1-py3-th-cpu/1.2.1-py3-th-gpu [2017-01-20]: Python 2.7/3.5 + Keras (1.2.1) + Theano (0.8.2) on CPU/GPU
• 1.2.0-py2-tf-cpu/1.2.0-py2-tf-gpu/1.2.0-py3-tf-cpu/1.2.0-py3-tf-gpu [2016-12-20]: Python 2.7/3.5 + Keras (1.2.0) + TensorFlow (0.12.0) on CPU/GPU
• 1.2.0-py2-th-cpu/1.2.0-py2-th-gpu/1.2.0-py3-th-cpu/1.2.0-py3-th-gpu [2016-12-20]: Python 2.7/3.5 + Keras (1.2.0) + Theano (0.8.2) on CPU/GPU
• 1.1.1-py2-tf-cpu/1.1.1-py2-tf-gpu/1.1.1-py3-tf-cpu/1.1.1-py3-tf-gpu [2016-10-31]: Python 2.7/3.5 + Keras (1.1.1) + TensorFlow (0.10.0) on CPU/GPU
• 1.1.1-py2-th-cpu/1.1.1-py2-th-gpu/1.1.1-py3-th-cpu/1.1.1-py3-th-gpu [2016-10-31]: Python 2.7/3.5 + Keras (1.1.1) + Theano (0.8.2) on CPU/GPU
• 1.1.0-py2-tf-cpu/1.1.0-py2-tf-gpu/1.1.0-py3-tf-cpu/1.1.0-py3-tf-gpu [2016-09-20]: Python 2.7/3.5 + Keras (1.1.0) + TensorFlow (0.10.0) on CPU/GPU
• 1.1.0-py2-th-cpu/1.1.0-py2-th-gpu/1.1.0-py3-th-cpu/1.1.0-py3-th-gpu [2016-09-20]: Python 2.7/3.5 + Keras (1.1.0) + Theano (0.8.2) on CPU/GPU
• 1.0.8-py2-tf-cpu/1.0.8-py2-tf-gpu/1.0.8-py3-tf-cpu/1.0.8-py3-tf-gpu [2016-08-28]: Python 2.7/3.5 + Keras (1.0.8) + TensorFlow (0.9.0) on CPU/GPU
• 1.0.8-py2-th-cpu/1.0.8-py2-th-gpu/1.0.8-py3-th-cpu/1.0.8-py3-th-gpu [2016-08-28]: Python 2.7/3.5 + Keras (1.0.8) + Theano (0.8.2) on CPU/GPU
• 1.0.6-py2-tf-cpu/1.0.6-py2-tf-gpu/1.0.6-py3-tf-cpu/1.0.6-py3-tf-gpu [2016-07-20]: Python 2.7/3.5 + Keras (1.0.6) + TensorFlow (0.9.0) on CPU/GPU
• 1.0.6-py2-th-cpu/1.0.6-py2-th-gpu/1.0.6-py3-th-cpu/1.0.6-py3-th-gpu [2016-07-20]: Python 2.7/3.5 + Keras (1.0.6) + Theano (0.8.2) on CPU/GPU
• 1.0.4-py2-tf-cpu/1.0.4-py2-tf-gpu/1.0.4-py3-tf-cpu/1.0.4-py3-tf-gpu [2016-06-16]: Python 2.7/3.5 + Keras (1.0.4) + TensorFlow (0.8.0) on CPU/GPU
• 1.0.4-py2-th-cpu/1.0.4-py2-th-gpu/1.0.4-py3-th-cpu/1.0.4-py3-th-gpu [2016-06-16]: Python 2.7/3.5 + Keras (1.0.4) + Theano (0.8.2) on CPU/GPU
• 1.0.1-py2-th-cpu/1.0.1-py2-th-gpu [2016-04-16]: Python 2.7 + Keras (1.0.1) + Theano (0.8.1) on CPU/GPU
• 0.3.3-py2-th-cpu/0.3.3-py2-th-gpu [2016-03-31]: Python 2.7 + Keras (0.3.3) + Theano (0.8.1) on CPU/GPU

Usage

Quick experiment with latest Keras (with TensorFlow backend on CPU) and your Python 2 code in current directory (will be mapped to /srv):

$ docker run -it --rm -v $(pwd):/srv gw000/keras /srv/run.py

Or using TensorFlow backend on GPUs in Python 2 (see docker-debian-cuda):

$ docker run -it --rm $(ls /dev/nvidia* | xargs -I{} echo '--device={}') $(ls /usr/lib/*-linux-gnu/{libcuda,libnvidia}* | xargs -I{} echo '-v {}:{}:ro') -v $(pwd):/srv gw000/keras:2.1.4-py2-tf-gpu /srv/run.py

Or using Theano backend on GPUs in Python 3 (see docker-debian-cuda):

$ docker run -it --rm $(ls /dev/nvidia* | xargs -I{} echo '--device={}') $(ls /usr/lib/*-linux-gnu/{libcuda,libnvidia}* | xargs -I{} echo '-v {}:{}:ro') -v $(pwd):/srv gw000/keras:2.1.4-py3-th-gpu /srv/run.py

Additional parameters in above commands explicitly expose your GPU devices and CUDA Driver library from the host system into the container. The vendor specific nvidia-docker tool performs the same thing in a less transparent way and is incompatible with other Docker tools. For more instructions see docker-debian-cuda.

In practice you are supposed to extend this image by writing your own Dockerfile that installs all your application dependencies (either using apt-get or pip). Eg. if you need Matplotlib, PIL/pillow, Pandas, Scikit-learn, and Statsmodels:

FROM gw000/keras:2.1.4-py3-tf-cpu

# install dependencies from debian packages
RUN apt-get update -qq \
 && apt-get install --no-install-recommends -y \
    python-matplotlib \
    python-pillow

# install dependencies from python packages
RUN pip --no-cache-dir install \
    pandas \
    scikit-learn \
    statsmodels

# install your app
ADD ai/ /srv/ai/
RUN chmod +x /srv/ai/run.py

# default command
CMD ["/srv/ai/run.py"]

If you are looking for a full deep learning research environment based on Keras and Jupyter, check out docker-keras-full.

Feedback

If you encounter any bugs or have feature requests, please file them in the issue tracker or even develop it yourself and submit a pull request over GitHub.

License

Copyright © 2016-2018 gw0 [http://gw.tnode.com/] <[email protected]>

All code is licensed under the GNU Affero General Public License 3.0+ (AGPL-3.0+). Note that it is mandatory to make all modifications and complete source code publicly available to any user.

docker-keras's Issues

.

Index files failed to download on keras:2.1.4_tf_gpu

Hi, I am trying to add some extra libraries on top of your keras:2.1.4_tf_gpu image. However, when I try to start building your dockerfile I get the following error:

W: The repository 'http://deb.debian.org/debian stretch Release' does not have a Release file.
W: The repository 'http://deb.debian.org/debian stretch-updates Release' does not have a Release file.
E: Failed to fetch http://deb.debian.org/debian/dists/stretch/main/binary-amd64/Packages  404  Not Found
E: Failed to fetch http://deb.debian.org/debian/dists/stretch-updates/main/binary-amd64/Packages  404  Not Found
E: Some index files failed to download. They have been ignored, or old ones used instead.
The command '/bin/sh -c apt-get update -qq  && apt-get install --no-install-recommends -y     build-essential     vim     g++     git     openssh-client     python3     python3-dev     python3-pip     python3-setuptools     python3-virtualenv     python3-wheel     pkg-config     libopenblas-base     python3-numpy     python3-scipy     python3-h5py     python3-yaml     python3-pydot  && apt-get clean  && rm -rf /var/lib/apt/lists/*' returned a non-zero code: 100

How do I fix this?

Not working

This docker image not supporting Volta architecture. Can you add CUDA 9?
I have Tesla V100-SXM2-16GB GPU and got the following error trying to use it inside the container

2018-01-18 15:59:34.504171: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:152] no NVIDIA GPU device is present: /dev/nvidia0 does not exist

Missing SSH package

Could you add the open ssh package within this containers so we can connect with git over SSH ?

apt-get update && apt-get install openssh-client

Not able to access the Juypter notebook from Host

Hello,
I am not able to access the Juypter note book from host, I am trying "localhost:8888", getting below error on container dos prompt:
No web browser found: could not locate runnable browser.

something wrong with the latest version keras:2.1.4-py3-tf-gpu

Hi,

When I do "singularity pull docker://gw000/keras:2.1.4-py3-tf-gpu", many packages like:
tar: var/lib/dpkg/info/xz-utils.prerm: Cannot write: No space left on device
cannot write, but I still have enough space

Can't find `nvidia-smi` in 2.1.4-py3-tf-gpu

$ docker run --runtime=nvidia -it --rm gw000/keras:2.1.4-py3-tf-gpu nvidia-smi 
docker: Error response from daemon: OCI runtime create failed: container_linux.go:296: starting container process caused "exec: \"nvidia-smi\": executable file not found in $PATH": unknown.

1.1.1-py2-th-gpu not compiled

Dockerhub is reporting an error on the compilation of this tag

There is no PIL library