ODSC-39392: resolve the comments

Author: Ziqun Ye

ads/model/deployment/model_deployment.py

@@ -17,7 +17,6 @@
 from ads.common import auth as authutil
 import pandas as pd
 from ads.model.serde.model_input import JsonModelInputSERDE
-from ads.common import auth, oci_client
 from ads.common.oci_logging import (
     LOG_INTERVAL,
     LOG_RECORDS_LIMIT,
@@ -63,6 +62,7 @@
 
 MODEL_DEPLOYMENT_KIND = "deployment"
 MODEL_DEPLOYMENT_TYPE = "modelDeployment"
+MODEL_DEPLOYMENT_INFERENCE_SERVER_TRITON = "TRITON"
 
 MODEL_DEPLOYMENT_INSTANCE_SHAPE = "VM.Standard.E4.Flex"
 MODEL_DEPLOYMENT_INSTANCE_OCPUS = 1
@@ -926,10 +926,7 @@ def predict(
         if model_name and model_version:
             header['model-name'] = model_name
             header['model-version'] = model_version
-        elif not model_version and not model_name:
-            
-            pass
-        else:
+        elif bool(model_version) ^ bool(model_name):
             raise ValueError("`model_name` and `model_version` have to be provided together.")
         prediction = send_request(
             data=data, endpoint=endpoint, is_json_payload=is_json_payload, header=header,
@@ -1404,9 +1401,9 @@ def _update_from_oci_model(self, oci_model_instance) -> "ModelDeployment":
             infrastructure.CONST_WEB_CONCURRENCY,
             runtime.env.get("WEB_CONCURRENCY", None),
         )
-        if runtime.env.get("CONTAINER_TYPE", None) == "TRITON":
+        if runtime.env.get("CONTAINER_TYPE", None) == MODEL_DEPLOYMENT_INFERENCE_SERVER_TRITON:
             runtime.set_spec(
-                runtime.CONST_INFERENCE_SERVER, "triton"
+                runtime.CONST_INFERENCE_SERVER, MODEL_DEPLOYMENT_INFERENCE_SERVER_TRITON.lower()
             )
 
         self.set_spec(self.CONST_INFRASTRUCTURE, infrastructure)

ads/model/deployment/model_deployment_runtime.py

@@ -562,6 +562,9 @@ def with_inference_server(self, inference_server: str = "triton") -> "ModelDeplo
             
         Example
         -------
+        >>> from ads.model.deployment import ModelDeployment, ModelDeploymentContainerRuntime, ModelDeploymentInfrastructure
+        >>> import ads
+        >>> ads.set_auth("resource_principal")
         >>> infrastructure = ModelDeploymentInfrastructure()\
         ...                 .with_project_id(<project_id>)\
         ...                 .with_compartment_id(<comparment_id>)\
@@ -578,7 +581,7 @@ def with_inference_server(self, inference_server: str = "triton") -> "ModelDeplo
         ...                 .with_model_uri(<model_id>)\
         ...                 .with_env({"key":"value", "key2":"value2"})\
         ...                 .with_inference_server("triton")
-        ...                 deployment = ModelDeployment()\
+        >>> deployment = ModelDeployment()\
         ...                 .with_display_name("Triton Example")\
         ...                 .with_infrastructure(infrastructure)\
         ...                 .with_runtime(runtime)