Ray is a unified framework for scaling AI and Python applications. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
Model composition APIs are incomplete as part of ray.serve.* API docs. While we have examples for all those deployment patterns, there are no API docs or code snippets for any of the primitives: InputNode, Model, DAGDriver, etc.
We have spoken at length about model composition at meetups and conferences, yet no complete & comprehensive API docs. Let's make all this part of our efforts to improve Ray Serve developer UX experience.
Description
Model composition APIs are incomplete as part of ray.serve.* API docs. While we have examples for all those deployment patterns, there are no API docs or code snippets for any of the primitives: InputNode, Model, DAGDriver, etc.
We have spoken at length about model composition at meetups and conferences, yet no complete & comprehensive API docs. Let's make all this part of our efforts to improve Ray Serve developer UX experience.
Link
This API page is incomplete This Serve API page should be comprehensive and should include all the APIs we use in your deployment graph examples