address reviewer feedback

Author: siddvenk

doc/frameworks/djl/using_djl.rst

@@ -49,7 +49,11 @@ Alternatively, you can provide full specifications to the DJLModel to have full
         },
         image_uri=<djl lmi image uri>,
     )
+    # Deploy the model to an Amazon SageMaker Endpoint and get a Predictor
+    predictor = djl_model.deploy("ml.g5.12xlarge",
+                                 initial_instance_count=1)
 
+Regardless of how you create your model, a ``Predictor`` object is returned.
 Each ``Predictor`` provides a ``predict`` method, which can do inference with json data, numpy arrays, or Python lists.
 Inference data are serialized and sent to the DJL Serving model server by an ``InvokeEndpoint`` SageMaker operation. The
 ``predict`` method returns the result of inference against your model.

src/sagemaker/djl_inference/model.py

@@ -147,9 +147,8 @@ def _infer_engine(self) -> Optional[str]:
             logger.info("Using provided engine %s", self.engine)
             return self.engine
 
-        if self.task is not None:
-            if self.task == "text-embedding":
-                return "OnnxRuntime"
+        if self.task == "text-embedding":
+            return "OnnxRuntime"
         return "Python"
 
     def _infer_image_uri(self):

src/sagemaker/serve/model_server/djl_serving/utils.py

@@ -98,7 +98,7 @@ def _get_default_max_tokens(sample_input, sample_output) -> tuple:
 
 def _get_default_djl_configurations(
     model_id: str, hf_model_config: dict, schema_builder: SchemaBuilder
-) -> tuple:
+) -> tuple[dict, int]:
     """Placeholder docstring"""
     default_tensor_parallel_degree = _get_default_tensor_parallel_degree(hf_model_config)
     if default_tensor_parallel_degree is None: