SDXL is able to be used alongside a Refiner model, specifically designed to take a mostly denoised image from an SDXL base model and complete the generation process.
When used in conjunction with a base model, this is known as an ensemble of expert denoisers, and can require fewer denoising steps to achieve high quality generations.
Refiner Settings
Refiner Model
Model used during the refiner portion of the generation process. Similar to the base model used for generations.
Scheduler
Scheduler used during the refiner portion of the generation process. Similar to the Scheduler used with the base model.
Steps
Number of steps that will be performed during the refiner portion of the generation process. Similar to the Steps set for the base model.
CFG Scale
Controls how much the prompt influences the generation process. Similar to the Generation CFG Scale.
Refiner Start
Where in the generation process the refiner will start to be used. 0 means the refiner will be used for the entire generation process, 0.8 means the refiner will be used for the last 20% of the generation process.
Positive Aesthetic Score
SDXL was trained on images with associated aesthetic scores. The Positive Aesthetic Score setting guides generations to be more similar to images with a high aesthetic score.
Negative Aesthetic Score
SDXL was trained on images with associated aesthetic scores. The Negative Aesthetic Score setting guides generations to be more similar to images with a low aesthetic score.
Was this article helpful?
That’s Great!
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
Feedback sent
We appreciate your effort and will try to fix the article