Our inferernce may take a long time to complete. Given using Ray serve as the backend, is there any good practice on the client side? Async HTTP ? Polling ? or send a callback URL with the http request?
Thanks a lot
Our inferernce may take a long time to complete. Given using Ray serve as the backend, is there any good practice on the client side? Async HTTP ? Polling ? or send a callback URL with the http request?
Thanks a lot