Will Sagemaker endpoints update when the model on hub updates?

I deployed a huggingface model onto Sagemaker directly from the huggingface model hub like below.

I’m now wondering what happens if I update the model on the huggingface hub? Will my sagemaker endpoint automatically update to start using the new model or do i have to go through some sort of redeployment?

huggingface_model = HuggingFaceModel(
   env=hub,   #includes Huggingface model ID
   role=role, 
   transformers_version="4.6",
   pytorch_version="1.7", 
    # framework_version="2.2.0",
    py_version="py36", 
)
 predictor = huggingface_model.deploy(
    initial_instance_count=1,
    instance_type="ml.m5.xlarge"
 )


The Endpoint start using the new model when it is restarted or scaled. Meaning that on start up SageMaker caches the model inside the endpoint and to retrigger the cache you either need to stop and restart the endpoint, update it or scale-in/out.
You can also provide the revision of the git repo through the HF_REVISION_ID to always use the model associated with the revision.

Meaning that an updated flow could be:

  1. Upload new model up Hub
  2. Update endpoint with new revision
  3. Sagemaker will nicely update the endpoint and use the new revision

Thanks. How do you carry out the Step 2 ?