Model Serve resource quick reference

Get descriptions of all quick start models

To get a list of the names and descriptions of all of the quick start models, use the following URI:

/mlops/api/v1/modelhub/models

Monitor all quick start models

To get the statuses of and information about all of the quick start models, use the following URI:

/mlops/api/v1/modelhub/monitor

Get information about a quick start model

To get the status of and information about a single quick start model, use the following URI:

/mlops/api/v1/modelhub/preBuiltModel?name=<model name>

Use TEXT_TRANSLATION or IMAGE_CLASSIFICATION as the model name.

Get the code template file

To get the code template file that you need to define a machine learning model, use the following URI:

/mlops/api/v1/model/template/custom

Upload a model file

When you upload a model file, first generate a model file ID and then use the ID to upload the file.

To generate a model file ID, use the following URI:

/mlops/api/v1/model/upload/generateId

To upload or update a model file, use the following URI:

/mlops/api/v1/model/upload/<model file ID>

Register a machine learning model

Register a machine learning model after you upload the model file that defines the machine learning algorithm.

To register a machine learning model, use the following URI:

/frs/v1/Projects('<model ID>')/Documents

Edit a machine learning model

To edit a registered machine learning model, use the following URI:

/frs/v1/Documents('<ID>')

Download a model file

To download the model file from a registered machine learning model, use the following URI:

/mlops/api/v1/model/download/<ID>

Create a model deployment

To create a model deployment based on a registered machine learning model, use the following URI:

/frs/v1/Projects('<deployment ID>')/Documents

Edit a model deployment

To edit a model deployment, use the following URI:

/frs/v1/Documents('<deployment ID>')

Monitor deployments

To monitor the status of all model deployments in your organization, use the following URI:

/mlops/api/v1/deployment/monitor

To monitor the status of one model deployment, use the following URI:

/mlops/api/v1/deployment/monitor/<deployment ID>

Control a quick start model or model deployment

You can start a quick start model or model deployment to make it available for predictions, restart it while it's running, and stop it to release the resources.

To start a deployment, use the following URI:

/mlops/api/v1/deployment/control/<deployment ID>/start

To restart a deployment, use the following URI:

/mlops/api/v1/deployment/control/<deployment ID>/restart

To stop a deployment, use the following URI:

/mlops/api/v1/deployment/control/<deployment ID>/stop

Generate predictions

To generate predictions from an available quick start model or model deployment, use the following URI:

/mlops/api/v1/deployment/request/<deployment ID>

Delete assets

To delete a machine learning model or model deployment, use the following URI:

/frs/api/v1/Documents('<ID>')