DL workspace is an open source toolkit that allows you to setup a cluster that can run deep learning training job, interactive exploration job, and evaluation service. The cluster also support big data analytic toolkit such as Hadoop/Spark.

DL Workspace is still in pre-release alpha stage. If you encounter issues in either deployment and/or usage, please open an issue at Github, or contact the DL Workspace team.

Development environment.

Please setup the dev environment of DL workspace as this.

Detailed Step-by-step setup insturction for a selected set of clusters.

DL workspace cluster can be deployed to either public cloud (e.g., Azure), or to on-prem cluster. The deployment to public cloud is more straightforward, as the environment is more uniform. The deployment instruction are as follows:

Azure Container Service

Azure Cluster

Azure Deployment using Azure Resource Management (ARM) templates

We give instruction on the deployment of DL Workspace to an on-prem cluster as well. Please note that because each on-prem cluster is different in hardware (and maybe software) configuration, the deployment procedure is more tricky. The basic deployment step is as follows.

On-Prem, Ubuntu

On-Prem, CoreOS

On-Prem, Ubuntu, Single Node

Additional information on general deployment can be found at here.