[Feature Request] Add reinforcement learning (RL) training support for Qwen-Image-Edit-2509 in DiffSynth.

[zk1009]

Is there a plan to add reinforcement learning (RL) training support for Qwen-Image-Edit-2509 in DiffSynth?

[LaFeuilleMorte]

Same Request, there's a new RL methods to use: https://github.com/PKU-YuanGroup/Edit-R1

[Deng-Xian-Sheng]

The Z-Image model seems to use similar reinforcement learning techniques to achieve excellent instruction compliance.

However, this might be limited by the visual understanding capabilities of the VLM?

If we assume something that the VLM cannot understand, perhaps the effect would be zero?