I'm experimenting with @fermat_app
’s upscaler model on Replicate. I'm getting some good results with these settings: https://
replicate.com/p/fl7jm5dbkjjv
zc5izhusmeo4ce
… – no prompt
– guess mode
– 0.45 prompt strength at 60 steps
– guidance scale 4 Original and upscaled.
Fermat App Upscaler Model Settings and Results on Replicate
By
–