Last updated 10th January, 2020.
Objective
Before being able to deploy a machine learning model, you need to create a namespace on your public cloud project. This guide covers this first step.
Requirements
- A public cloud project
Instructions
Step 1 - Go to the ML Serving menu
All actions on the ML Serving are made on the Public Cloud section
of the OVHcloud control panel and are located under
Data & Analytics > ML Serving
Step 2 - Start the creation of a new namespace
Once you read general information and validated this product's contract
terms, you can start the creation of a namespace by clicking on the
Create a namespace button.
Step 3 - Select the region for your namespace
One or several regions are available for your namespace. Each region
corresponds to a physical OVHcloud datacenter. Just select the region in
which you want your namespace to be deployed and click next.
Step 4 - Choose a name for your namespace
You can create as many namespaces as you want in your public cloud
project. Just choose a name identifying your namespace and click next.
Step 5 - Attach your object storage container
Each namespace is linked to a data container inside public cloud object storage.
You can choose to link to an existing container or create a new one.
Step 6 - Check that everything is created
Your namespace is now created and you can check general information about it.
You can check that the data container was created (if it was not
already) by going on the menu under Storage > Object Storage.
Your container should appear in the list.
Going Further
- You can check the official documentation of the open source product ML Serving about Namespaces