DEEP Open Catalog

Welcome to the DEEP Open Catalog!

DEEP-Hybrid-DataCloud is delivering a comprehensive platform to easily develop, build, share and deploy Artificial Intelligence, Machine Learning and Deep Learning modules on top of distributed e-Infrastructures.

In the DEEP Open Catalog you can find ready to use modules in a variety of domains. These modules can be executed on your local laptop, on a production server or on top of computing e-Infrastructures supporting the DEEP-Hybrid-DataCloud stack.

Explore our marketplace!

TF Benchmarks

docker, tensorflow, cnn, trainable, api-v1

tf_cnn_benchmarks accessed via DEEPaaS API
Know more »»
semseg_vaihingen

tensorflow, semantic segmentation, cnn, trainable, inference, pre-trained, api-v1

2D semantic segmentation on the Vaihingen dataset
Know more »»
Train an audio classifier

tensorflow, docker, deep learning, trainable, inference, pre-trained, api-v1

Train your own audio classifier with your custom dataset. It comes also pretrained on the 527 AudioSet classes.
Know more »»
Body pose detection

tensorflow, docker, deep learning, inference, pre-trained, api-v1

Detect body poses in images.
Know more »»
Speech keywords classifier

tensorflow, docker, deep learning, trainable, inference, api-v1

Train a speech classifier to classify audio files between different keywords.
Know more »»
Development Docker image

development, tensorflow, docker, jupyterlab

This is a Docker image for developing new modules
Know more »»

Page 1 / 3

Quickstart

1. Explore the marketplace

Browse the list of available modules and pick the one of your choice.

The DEEP Generic Container is a good starting point to see how the API looks like (but it does not provide any functionality at all).

2. Try locally

With Docker:

$ docker search deephdc
$ docker run -ti -p 5000:5000 deephdc/deep-oc-generic-container

No Docker possible? Try udocker instead!

$ udocker search deephdc
$ udocker run -p 5000:5000  deephdc/deep-oc-generic-container

Need GPU access?

Use nvidia-docker together with Docker:

$ nvidia-docker run -ti -p 5000:5000 deephdc/deep-oc-generic-container

Or udocker way:

$ udocker pull deephdc/deep-oc-generic-container
$ udocker create --name=deep-oc-generic-container deephdc/deep-oc-generic-container
$ udocker setup --nvidia deep-oc-generic-container
$ udocker run -p 5000:5000  deep-oc-generic-container

3. Browse the API

Open your browser and visit http://127.0.0.1:5000/.
Read the Swagger documentation and test the module directly.