What Oracle Cloud Infrastructure feature allows catalogued models to serve as HTTP endpoints?

The feature that allows catalogued models to serve as HTTP endpoints is Model Deployments. In Oracle Cloud Infrastructure, Model Deployments enable you to take a trained machine learning model and make it accessible via an API. This allows applications and services to interact with the model, making predictions or performing inference in real-time through HTTP requests.

By deploying a model, you create a serving infrastructure that can handle incoming requests and return responses based on the model's predictions. This process is essential for integrating machine learning capabilities into applications, thereby providing a seamless user experience and enhancing functionality.

Other options, while related to the machine learning lifecycle, do not serve the same purpose. Model Frameworks refers to the specific technologies or libraries used to build models, Model Catalogs relate to the organization and management of models but do not directly involve serving them, and Model Serving is a broader term that encompasses the process but does not specifically designate the deployment step that transforms a model into an HTTP endpoint.