Skip to main content

Deployments

Deployments are models hosted by serving engines that return predictions via REST or gRPC kserve v2 protocol.

Attributes

name Name of the deployment.

deployment_version   The version of the deployment.

engine_name The name of the serving engine hosting the deployment.

model_name   The name of the model.

model_version   The version of the model.

deployment_url     The fully qualified URL for prediction.

comment Optional comment associated with deployment.

error_text    Optional error set when a deployment fails.

Methods

Deployment.predict