PhotoReal DiT

On this model we work together (Ai Art Vision). This model is trained on a 26k photo realistic dataset ad mainly focus on realistic image style.

Recommended Parameters:

• works best with the normal "Euler" "and "dpmpp_2m sgm_uniform" sampler with a CFG scale of around 5-6 for the best results.

  • Don't use the upscaler in WebUI mode, as it will generate noisy images (a common problem with all SD3-M models). If you are comfortable with ComfyUI, use Ultimate SD Upscaler instead.

🌐 Any inquiries about my model are more than welcome. Reach out on Discord ➡️ shopon_skp. Let's keep the conversation going.