Example how to dockerize deep neural network created with Keras library
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data
.dockerignore
.gitignore
Dockerfile
Makefile
README.md
learn.py
model.h5
predict.py

README.md

Dockerized deep learning

This repository contains code to run already trained, dockerized Deep Convolutional Network. Article describing whole process can be found here.

Installing Docker

General installation instructions are on the Docker site, here are some quick links:

Running the container

We are using Makefile to simplify docker commands within make commands.

Firstly, insert images compatible with CIFAR10 dataset into data directory.

Next, build the container

$ make build

and run classifier

$ make predict

It will find all images under /data directory and classify them. Optionally, you can pass list of images to classify (glob patterns supported):

$ make classify IMAGES="airplane*.png deer4.png"

Raw commands can be found inside Makefile. Example without make:

$ docker run -it --rm -v $PWD/data:/srv/data cifar python predict.py

Licence

MIT