Updated on 2026-07-03 GMT+08:00

FAQs

  1. Can a synthesis operator be placed in the middle of a workflow?

    No. In the current version, synthesis operators can only be placed at the end of a workflow. If you need to perform further processing on the synthesized results, you are advised to split the process into two separate tasks:

    1. Task 1: processing operator + synthesis operator
    2. Task 2: further processing of the synthesized results
  2. Is text-to-image cross-modal synthesis supported?

    The current version does not support cross-modal synthesis. It only supports same-modality data synthesis, for example:

    • Text → text (Q&A rewriting)
    • Text → image (not supported)
    • Image → text (not supported)