Deploying the model

1. Device Registration

You need to enter the machine ID details and select the hardware type. Machine ID can be retrieved using command cat/etc/machine-id. Click on submit which gives you the pop up of device registered.

Click on View Details in the pop up screen or go to device management and you will get a screen as shown below:

At this stage, the status of the device will be Inactive. 

You will need to run the two docker commands displayed on screen on your device to complete the registration of the device and set it to active status.

2. New Deployment  

a) Artifacts – Select the artifact you want to deploy. If you want to deploy the model from Model Zoo please select the Model Zoo option and if you want to deploy a model which you have trained and optimized using ENAP Studio please select the Custom Optimized option.

b) Hardware Type – Select the hardware type you want to deploy your model on.

c) Device – Select the device you have registered. Your registered device will be displayed in the dropdown list.

d) Select either the Model Zoo Artifact or Custom Optimized Artifact which you want to deploy. The corresponding artifacts will be displayed in the dropdown list.

e) Advanced Settings: You can replace the URL with a link to a YouTube video or an RTSP stream

Once you click on submit it will start a docker on the selected edge device. Inferencing will begin and you can view the logs in the Reports section.

3. Reports

You can view the deployments carried out by you. Clicking on info icon will provide you with detailed reports and logs of deployment.

4. Device Management

The user can see the summary of devices he has registered and the status for the same.

 

January 13, 2022