possible to train a lora for it to do a certain task and multiple inputs for editing and referneces?
#4
by
krigeta
- opened
is it possible to train this one and can it accept multiple inputs? like we can able to do with qwen edit 2509?
Thanks for your interest. Regarding the architecture, it indeed supports multiple inputs. However, the currently released version was trained using only single-image editing data. We plan to support multi-image inputs in future version iterations. Please stay tuned!