balena.yaml

name: balena-openvino
type: sw.application
description: >-
   Get started with OpenVINO™. Develop, tweak and test models direclty on edge devices. 
post-provisioning: >-
   OpenVINO™ is an open-source toolkit for optimizing and deploying AI inference. This project is built around this toolkit, and is aimed to reduce friction in developing AI models on the 
   edge. The aim of this project is to get you started with OpenVINO™ , and allow you to develop, tweak and test models on edge devices. To this end, it includes Jupyter Notebook, a popular 
   environment for data science and machine learning applications, and some tools that allow you to interact with the Intel OpenVino Model Zoo.

   This is meant to be both a demo of OpenVino running on balena, containing two models, one for object recognition, and one for face detection, but also a playground where you can download, 
   convert and run your models directly on an edge device. Upon deploying this fleet, in Jupyter, you'll see three notebooks, the two examples, and a template file you can use to work on your 
   model.
version: 0.0.3
assets:
  repository:
    type: blob.asset
    data:
      url: 'https://github.com/cristidragomir97/balena-openvino'
  logo:
    type: blob.asset
    data:
      url: 'https://raw.githubusercontent.com/balena-labs-projects/balena-openvino/master/logo.png'
data:
  defaultDeviceType: intel-nuc
  supportedDeviceTypes:
    - intel-nuc
    - generic-amd64
    - genericx86-64-ext