Deployment details area

From the Deployments page, select a deployment name to access the model and deployment details.

The deployment area page includes access to everything you need to access the model details, deployment settings, performance graphs and more. The following tabs include specific information to help you set up and monitor your deployed models.
Models
Displays the champion and challenger models for each deployment and provides access to model information and data and promotion capabilities. This tab also includes the details about the response variable and all predictor variables. For more information on using champion and challenger models, go to Use champion/challenger models in a deployment.
Performance
Access the Drift and Stability reports.
Drift report
The Drift report shows drift in the champion model. Drift reports update when new data are received.
Stability report
The Stability report shows the performance of the champion and challenger models. Stability reports update daily.
Settings
Includes the observation ID settings, drift alert thresholds, and email alert settings. For more information, go to Deployment settings.
Audit Log
Records actions on the deployment. For more information, go to Deployment audit log.
Service Metrics
Service metrics include measures of the health of the endpoint that you use to access the model.
Endpoint Status Summary
The endpoint status summary shows the HTTP status code for each HTTP request. Codes in the range 200 to 299 represent successes. Codes in the range 400 to 599 represent errors. If the graph includes unexpected codes, investigate the cause.
Response Time
The response time shows the time to complete HTTP requests. Usually, the line is relatively flat. If the graph includes an increasing trend or unexpected spikes, investigate the cause.

Storage metrics show the storage utilization for your deployment for the current day. Minitab Model Ops has a limit of 1440 data rows per deployment, per day.

Integration
Includes details about the API keys, as well as code formats to use to generate prediction and stability requests.

View model details

Minitab Model Ops stores basic model information for each model in the deployment. Use the following steps to view the model information.
  1. Open the Deployments page.
  2. Select a deployment name to open the deployment details area.
  3. Open the Models tab.
  4. Open the settings of the model .
  5. Select Model information.
The model information includes the following details.
Model Name
The descriptive name for the model assigned upon import.
Model Type
The type of model.
Created By
The name of the user who imported the model.
Created
The date and time of the model's creation.

Rename an existing model

  1. Open the Deployments page.
  2. Select a deployment name to open the deployment details area.
  3. Open the Models tab.
  4. Open the settings of the model .
  5. Select Rename model.

Download baseline data

  1. Open the Deployments page.
  2. Select a deployment name to open the deployment details area.
  3. Open the Models tab.
  4. Open the settings of the model .
  5. Select Download baseline data.
  6. Select either CSV or MWX format.

Remove a model from a deployment

You can remove a challenger model from a deployment. When you remove a challenger model, the model is still available in the model repository.
  1. Open the Deployments page.
  2. Select a deployment name to open the deployment details area.
  3. Open the Models tab.
  4. Open the settings of the challenger model .
  5. Select Remove model.