everdockers / tensorflow_cpu

Dockerfile and files for baikangwang/tensorflow_cpu, python + tensorflow

Tensorflow (cpu)

• repository
  • baikangwang/tensorflow_cpu
    
• tags
  • tfonly,latest
  • tfonly_py2
  • jupyter,jupyter_py3
  • jupyter_py2
  • emnist

tfonly, latest

Description

provides fundamental developing environment of machine learning, python 3.5+tensorflow

Docker
Docker Pull	docker pull baikangwang/tensorflow_cpu[:tfonly]
Dockerfile	https://github.com/EverDockers/tensorflow_cpu/blob/tfonly/Dockerfile

Usage

docker run -it --name <container name> -v <local working dir>:/projects -p <host port>:6006 baikangwang/tensorflow_cpu:tfonly [/bin/bash]

Components

• nvidia-docker: 1.0
• cuda: 8.0.61
• cudnn: 6.0.21
• tensorflow: 1.3.0
• python: 3.5

tfonly_py2

Description

provides fundamental developing environment of machine learning with python 2.7, python 2.7+tensorflow

Docker
Docker Pull	docker pull baikangwang/tensorflow_cpu:tfonly_py2
Dockerfile	https://github.com/EverDockers/tensorflow_cpu/blob/tfonly2/Dockerfile

Usage

docker run -it --name <container name> -v <local working dir>:/projects -p <host port>:6006 baikangwang/tensorflow_cpu:tfonly_py2 [/bin/bash] 

Components

• nvidia-docker: 1.0
• cuda: 8.0.61
• cudnn: 6.0.21
• tensorflow: 1.3.0
• python: 2.7

jupyter,jupyter_py3

Description

based on baikangwang/tensorflow_cpu:tfonly and provides python development environment, python 3.5+tensorflow+jupyter

Docker
Docker Pull	docker pull baikangwang/tensorflow_cpu:jupyter
Dockerfile	https://github.com/EverDockers/tensorflow_cpu/blob/jupyter2/Dockerfile

Usage

docker run -it --name <container name> -v <local working dir>:/projects -p <host port>:8888 baikangwang/tensorflow_cpu:jupyter [/bin/bash]

NOTE: the default command is /run_jupyter.sh --allow-root, which will automatically launch jupyter notebook. When create the container with a initial command as /bin/bash, should manually launch the jupyter notebook within the container, /run_jupyter.sh --allow-root

Components

• nvidia-docker: 1.0
• cuda: 8.0.61
• cudnn: 6.0.21
• tensorflow: 1.3.0
• python: 3.5
• jupyter

jupyter_py2

Description

based on baikangwang/tensorflow_cpu:tfonly_py2 and provides python 2.7 development environment, python+tensorflow+jupyter

Docker
Docker Pull	docker pull baikangwang/tensorflow_cpu:jupyter_py2
Dockerfile	https://github.com/EverDockers/tensorflow_cpu/blob/jupyter2/Dockerfile

Usage

nvidia-docker run -it --name <container name> -v <local working dir>:/projects -p <host port>:8888 baikangwang/tensorflow_cpu:jupyter_py2 [/bin/bash]

NOTE: the default command is /run_jupyter.sh --allow-root, which will automatically launch jupyter notebook. When create the container with a initial command as /bin/bash, should manually launch the jupyter notebook within the container, /run_jupyter.sh --allow-root

Components

• nvidia-docker: 1.0
• cuda: 8.0.61
• cudnn: 6.0.21
• tensorflow: 1.3.0
• python: 2.7
• jupyter

emnist

Description

based on baikangwang/tensorflow_cpu:tfonly and provides python development environment, python 3.5+tensorflow+jupyter+keras

Docker
Docker Pull	docker pull baikangwang/tensorflow_cpu:emnist
Dockerfile	https://github.com/EverDockers/tensorflow_cpu/blob/emnist/Dockerfile

Usage

docker run -it --name <container name> -v <local working dir>:/projects -p <host port>:8888 -p <host port>:5000 baikangwang/tensorflow_cpu:emnist [/bin/bash]

Components

• nvidia-docker: 1.0
• cuda: 8.0.61
• cudnn: 6.0.21
• tensorflow: 1.3.0
• python: 3.5
• jupyter
• keras