Skip to Content

Deploy the Image Classification Retrained Model

test
0 %
Deploy the Image Classification Retrained Model
Details
// Explore More Tutorials

Deploy the Image Classification Retrained Model

11/06/2018

Discover how to deploy the retraining model for the SAP Leonardo Machine Learning foundation Image Classification Retraining scenario

You will learn

  • How to check the content of the model repository
  • How to request a new model deployment to be deployed
  • How to check the model deployment status

Step 1: Check the Model Repository Model

Once a job is completed, the model will be automatically stored in the model repository.

Only model store in the repository can then be deployed.

To check the list of model in the repository, you can execute the following command:

cf sapml retraining models -m image

Your model should be listed in the output.

  NAME     VERSION  UPDATED AT                CREATED AT                CHECKSUM
  flowers  1        2018-11-03T15:53:09.418Z  2018-11-03T15:53:02.624Z

Note: You may have more than one version of the same model available in the repository

Log on to answer question
Step 2: Submit your Model for Deployment

Once the model is available in the repository, you can request the model to be deployed using the following command:

cf sapml retraining model_deploy flowers 1 -m image

Once the deployment request completed, you should get the following output:

Deployment request submitted with id "xx-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx".

Copy in your notepad the submission id from the console (***xx-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx***)

Log on to answer question
Step 3: Check your Deployed Model

You can check that your model was properly deployed using the following command:

cf sapml retraining model_deployments -m image

Your model should be listed in the output with a status that will change during the deployment process:

  ID         MODEL NAME MODEL VERSION  HOST    PORT  STATUS     DESCRIPTION
  xxxxxxxxx  flowers    1              zzzzzz  443   PENDING    Loading model...

Then:

  ID         MODEL NAME MODEL VERSION  HOST    PORT  STATUS     DESCRIPTION
  xxxxxxxxx  flowers    1              zzzzzz  443   PENDING    Finalizing the deployment resources creation

And finally:

  ID         MODEL NAME MODEL VERSION  HOST    PORT  STATUS     DESCRIPTION
  xxxxxxxxx  flowers    1              zzzzzz  443   SUCCEEDED  Model is deployed and ready to serve.

Copy in your notepad the host id from the console (***HOST***)

Provide an answer to the question below then click on Validate.

Paste the host value for your deployed model.
×

Next Steps

Back to top