Prebuilts on Cerebrium

Introduction

This repo consists of the prebuilt models for Cerebrium.
Each folder in ./Prebuilts contains a cortex deployment that can be used on Cerebrium. This consists of:

• config.yaml: The configuration file for the cortex deployment. Contains information on the resources required for the deployment.
• main.py: The main file that is executed during the runtime
  • This has three parts
    • The Item which is the parameters for the API call
    • The init part which is only executed during the build time
    • The predict part which is executed during the runtime.
• requirements.txt: The dependencies for the model
• pkglist.txt: The list of apt packages that are installed during the build time

Additionally, it is adviced to place a small README.md to describe the model and its usage.

How to use

Deploy using the cerebrium dashboard.

The primary method to deploy prebuilt models is using the dashboard.
Simply navigate to the pre-built models tab and click deploy for the model you would like to deploy!

Deploy with the Cerebrium CLI

Alternatively, you can use the Cerebrium CLI to deploy the model since each of the models in this repo are, by default, valid Cortex projects.

• Clone this repo and open your terminal.
• cd into the folder of the model you would like to deploy.
• If you have not done so already, login to the Cerebrium CLI by running:

cerebrium login <YOUR PRIVATE API KEY>

• Deploy your model with the following command:

cerebrium deploy <YOUR NAME FOR YOUR MODEL> --config-file config.yaml 

How to contribute

To contribute to this repo, please fork this repo and create a pull request.
Additionally, please create an issue for any bugs or feature requests.

Don't hesitate to reach out to the team on slack or discord if there is anything you need help with or if you have any suggestions.