promptdojo_

Batch vs realtime serving — step 6 of 7

Write choose_mode(user_waiting) that returns "realtime" when a user is waiting on the prediction, and "batch" when nothing is waiting.

full-screen editor opens — close anytime to keep reading.