qubvel/train_imagenet

Template

Requierments

1. CUDA 9.0
2. nvidia-docker 18.06.0-ce

Instructions

Complete following steps to run your jupyter server inside docker container.
Some preinstalled software:

1. Anaconda 5.1 package with Python 3.6
2. Keras 2.1.5 (tensorflow 1.7.0)
3. Pytorch 0.4.0
4. OpenCV 3.4.2
5. imgaug, albumentations

0. Modify configuration file.

config.yaml file:

project_name: template

docker_image_name: template_image
docker_container_name: template_container

jupyter_port: 8888
tensorboard_port: 6006

1. Build docker image.

$ ./docker_build.sh

2. Launch container with notebook server.

$ ./docker_run.sh