search

aws / amazon-sagemaker-examples

Example 📓 Jupyter notebooks that demonstrate how to build, train, and deploy machine learning models using 🧠 Amazon SageMaker.
https://sagemaker-examples.readthedocs.io
Apache License 2.0
9.96k stars 6.73k forks source link

Registering a PyTorchModel to SageMaker console without deploying #760

Open austinmw opened 5 years ago

austinmw commented 5 years ago

I have a PyTorchModel object defined like this:

model = PyTorchModel(name='model-name',
                     model_data=model_uri,
                     role=role,
                     framework_version='1.0.0',
                     entry_point='serve.py',
                     source_dir='src',
                     sagemaker_session=sagemaker_session)

Through the Python SDK, how can I just register this model with the SageMaker console without creating a new endpoint_configuration and deploying a new endpoint?

I've looked through the SDK documentation and can't find any methods for accomplishing this.

laurenyu commented 5 years ago

hi @austinmw, you use the undocumented method _create_sagemaker_model() (code) to create the SageMaker model without creating an endpoint configuration or deploying to an endpoint. We do have an item in our backlog to consider making this a public method, and are always reprioritizing our backlog based on customer feedback. Thanks for using SageMaker!