Can you add the ability to select the text encoder models manually? Stable Diffusion 3 can use alternate text encoders, and when using perturbed models, they don't include one. They work fine in comfy ui, but can't be used in Stability Matrix inference.
Can you add the ability to select the text encoder models manually? Stable Diffusion 3 can use alternate text encoders, and when using perturbed models, they don't include one. They work fine in comfy ui, but can't be used in Stability Matrix inference.