[News] Double the Speed, Keep the Quality: Omni-Turbo Model

In the NEXT update, a Turbo version of Omni-Small will be available.

This Turbo version aims to preserve the quality of the Omni-Small model while boosting speed.

Speed:

The turbo model nearly doubled the speed of Omni-Small, while still faster than the quality mode of RTX-SuperRes

Quality:

The de-noise capability of Turbo is significantly better than Mini. But when compared to Omni-Small, Turbo's result is less sharp when zoomed in.

🖼️Compare results on Imgsli:

🔗VS Omni-Mini

🔗VS Omni-Small

🔗VS RTX-SuperRes

This is only the first version of the Omni-Turbo model. I'm still working on future updates for this model. Thank you for supporting this project! 💕