I noticed that in the implementation of pipeline_qwenimage_edit_plus.py, the following image size parameters are hardcoded:
CONDITION_IMAGE_SIZE = 384 * 384
VAE_IMAGE_SIZE = 1024 * 1024
(Reference: pipeline_qwenimage_edit_plus.py)
I would like to clarify whether the qwen-image-edit-2509 and qwen-image-edit-2511 models were trained using these exact same settings for CONDITION_IMAGE_SIZE and VAE_IMAGE_SIZE?