A ML Model can only begin to add value in an organization when that model’s insights routinely become available to the users for which it was built. The process of taking a trained ML model and making its predictions available to users or other systems is known as Deployment. Deployment is entirely distinct from routine machine learning tasks like feature engineering, model selection, or model evaluation.
1. To access ML Deployment, click MLOps located on left navigation bar.
2. Click ML Deployments.
3. To add deployment, click Add Deployment.
4. In the appeared form, select specific AI Usecase for which you are planning to work, select Data Source for the related ML Deployment, select specific ML Model. In item type select the type of deployment you want to create. Select item name. Then you need to select the assignee, who will approve your ML Deployment.
After creating ML Deployment, it will go to Task Pending of the assigned approver. The assigned approver has to go to his list of Task Pending for Approving or Rejecting the request. Both the Assignor and the Approver can track the status of the same task from their respective Task Pending page.
Below are the details of the process step by step -
- Approver firstly needs to click on the specific Task Name. Here approver will get the details related to task in the Task Brief, he can also see the relevant documents attached for that usecase and can also comment on the task.
- After that, Approver can click the Approve button. Then approver will need to fill the form appeared. He can also assign the ML Deployment to other team member.
- Then Approver can click on Approve button if he wants to approve. To reject, he can click on Reject button. He can also click on Request to Probyto, if he wants Probyto to work on that task.
Then Select the Due date, it is a kind of expiry date of your data source. Then click on Create button.
5. After adding ML Deployment, you will see all the deployments you have added on the home screen of ML Deployment with its Usecase ID, status(approved or not), Last updated by, type, activation status etc. as shown in the below picture.
6. To edit or configure ML Deployment, Click on the name of any ML Deployment.
7. On the second Level page of ML Deployment, You will find Configure, Click on Configure.
8. Fill all the details such as-
- Enter the Key- In this column, you need to provide configurations for successful integrations.
- Enter the Value-In this option, you need to give the detail or the link related to your Key.
Then click Submit.
After adding all the configurations, it will be shown below the description of ML Deployment.
9. You can also edit your ML Deployment details, To edit ML Deployment, Click Edit ML Deployments.
Fill all the relevant details, click Update.