能只用lora吗

[zwukong]

我看到你们有个Lora，wan2.2的低噪+ lora不知道支持吗。要是能支持comfyui原生工作流就完美了

[EasonXiao-888]

Thank you for your interest. To isolate the influence of the model itself, we train a general editing model from scratch. As a result, the released LoRA weights may not generalize well to other base models. However, you can directly fine-tune Wan2.2 using our dataset — in our experience, it is relatively easy to obtain strong results.
In addition, our other project, JoyAI-Image-Edit, has provided a ComfyUI workflow you could follow.