Add helm inference

[shantanutrip]

What's changing and why?

• CLI and SDK change for mig profile support

Before/After UX

Before:

chadchc@80a997306578 hyperpod-jumpstart-inference-template % hyp create hyp-jumpstart-endpoint --help
Usage: hyp create hyp-jumpstart-endpoint [OPTIONS]

Options:
  --version TEXT                  Schema version to use
  --debug BOOLEAN                 Enable debug mode
  --namespace TEXT                Kubernetes namespace
  --accept-eula BOOLEAN           Whether model terms of use have been accepted  [default: False]
  --metadata-name TEXT            Name of the jumpstart endpoint object
  --model-id TEXT                 Unique identifier of the model within the hub  [required]
  --model-version TEXT            Semantic version of the model to deploy (e.g. 1.0.0)
  --instance-type TEXT            EC2 instance type for the inference server  [required]
  --endpoint-name TEXT            Name of SageMaker endpoint; empty string means no creation
  --tls-certificate-output-s3-uri TEXT
                                  S3 URI to write the TLS certificate
  --help                          Show this message and exit.

After:

chadchc@80a997306578 hyperpod-jumpstart-inference-template % hyp create hyp-jumpstart-endpoint --help
Usage: hyp create hyp-jumpstart-endpoint [OPTIONS]

Options:
  --version TEXT                  Schema version to use
  --debug BOOLEAN                 Enable debug mode
  --namespace TEXT                Kubernetes namespace
  --accept-eula BOOLEAN           Whether model terms of use have been accepted  [default: False]
  --metadata-name TEXT            Name of the jumpstart endpoint object
  --model-id TEXT                 Unique identifier of the model within the hub  [required]
  --model-version TEXT            Semantic version of the model to deploy (e.g. 1.0.0)
  --instance-type TEXT            EC2 instance type for the inference server  [required]
  --accelerator-partition-type TEXT
                                  MIG profile to use for GPU partitioning
  --accelerator-partition-validation TEXT
                                  Enable MIG validation for GPU partitioning. Default is true.  [default: True]
  --endpoint-name TEXT            Name of SageMaker endpoint; empty string means no creation
  --tls-certificate-output-s3-uri TEXT
                                  S3 URI to write the TLS certificate
  --help                          Show this message and exit.

How was this change tested?

1. Test with local setup inference operator
2. Using this command

chadchc@80a997306578 hyperpod-jumpstart-inference-template % hyp create hyp-jumpstart-endpoint \     
  --version 1.1 \
  --model-id deepseek-llm-r1-distill-qwen-1-5b \
  --instance-type ml.p4d.24xlarge \
  --endpoint-name js-test \
  --accelerator-partition-type "mig-4g.20gb" \
  --accelerator-partition-validation true --debug true --tls-certificate-output-s3-uri s3://sagemaker-cc-helm-test-9a02e56c-tls-ae472dc0

Are unit tests added?

Yes

Are integration tests added?

No

Reviewer Guidelines

‼️ Merge Requirements: PRs with failing integration tests cannot be merged without justification.

One of the following must be true:

• All automated PR checks pass
• Failed tests include local run results/screenshots proving they work
• Changes are documentation-only

[piyushdaftary]

Can you look into UT failures

FAILED test/unit_tests/cli/test_inference.py::test_js_create_with_mig_profile - NameError: name 'mock_load_schema' is not defined
FAILED test/unit_tests/inference/test_hp_jumpstart_endpoint.py::TestHPJumpStartEndpoint::test_create - AssertionError: Expected 'call_create_api' to be called once. Called 0 times.
FAILED test/unit_tests/inference/test_hp_jumpstart_endpoint.py::TestHPJumpStartEndpoint::test_create_missing_name_and_endpoint_name - AssertionError: Exception not raised
FAILED test/unit_tests/inference/test_hp_jumpstart_endpoint.py::TestHPJumpStartEndpoint::test_create_validation_logic_priority - AssertionError: Expected 'validate_mig_profile' to be called once. Called 0...
FAILED test/unit_tests/inference/test_hp_jumpstart_endpoint.py::TestHPJumpStartEndpoint::test_create_with_accelerator_partition_validation - AssertionError: Expected 'validate_mig_profile' to be called once. Called 0...
FAILED test/unit_tests/inference/test_hp_jumpstart_endpoint.py::TestHPJumpStartEndpoint::test_create_with_metadata - AssertionError: Expected 'call_create_api' to have been called once. Called...
FAILED test/unit_tests/inference/test_hp_jumpstart_endpoint.py::TestHPJumpStartEndpoint::test_create_without_accelerator_partition_validation - AssertionError: Expected 'validate_instance_type' to be called once. Called...