Skip to content

Predictor#

You can get a ModelServing instance using Project.get_model_serving. Once you have it, you can create a predictor using ModelServing.create_predictor. Predictors can also be accessed from the Deployment metadata objects:

deployment.predictor

Predictor #

Bases: DeployableComponent

Metadata object representing a predictor in Model Serving.

api_protocol property writable #

api_protocol

API protocol enabled in the predictor (e.g., HTTP or GRPC).

artifact_files_path property #

artifact_files_path

Path of the artifact files deployed by the predictor.

artifact_path property #

artifact_path

Path of the model artifact deployed by the predictor. Resolves to /Projects/{project_name}/Models/{name}/{version}/Artifacts/{artifact_version}/{name}{version}.zip.

artifact_version property writable #

artifact_version

Artifact version deployed by the predictor.

Deprecated

Artifact versions are deprecated in favor of deployment versions.

config_file property writable #

config_file

Model server configuration file passed to the model deployment.

It can be accessed via CONFIG_FILE_PATH environment variable from a predictor or transformer script. For LLM deployments without a predictor script, this file is used to configure the vLLM engine.

created_at property #

created_at

Created at date of the predictor.

creator property #

creator

Creator of the predictor.

description property writable #

description

Description of the predictor.

environment property writable #

environment

Name of the inference environment.

id property #

id

Id of the predictor.

inference_logger property writable #

inference_logger

Configuration of the inference logger attached to this predictor.

model_framework property writable #

model_framework

Model framework of the model to be deployed by the predictor.

model_name property writable #

model_name

Name of the model deployed by the predictor.

model_path property writable #

model_path

Model path deployed by the predictor.

model_server property #

model_server

Model server used by the predictor.

model_version property writable #

model_version

Model version deployed by the predictor.

name property writable #

name

Name of the predictor.

project_name property writable #

project_name

Name of the project the deployment belongs to.

project_namespace property writable #

project_namespace

Kubernetes project namespace.

requested_instances property #

requested_instances

Total number of requested instances in the predictor.

script_file property writable #

script_file

Script file used to load and run the model.

serving_tool property writable #

serving_tool

Serving tool used to run the model server.

transformer property writable #

transformer

Transformer configuration attached to the predictor.

version property #

version

Version of the predictor.

deploy #

deploy()

Create a deployment for this predictor and persists it in the Model Serving.

Example 
import hopsworks

project = hopsworks.login()

# get Hopsworks Model Registry handle
mr = project.get_model_registry()

# retrieve the trained model you want to deploy
my_model = mr.get_model("my_model", version=1)

# get Hopsworks Model Serving handle
ms = project.get_model_serving()

my_predictor = ms.create_predictor(my_model)
my_deployment = my_predictor.deploy()

print(my_deployment.get_state())
RETURNS DESCRIPTION

Deployment. The deployment metadata object of a new or existing deployment.

describe #

describe()

Print a JSON description of the predictor.