My usual settings

DPM++ SDE Karras with 30 steps
CFG Scale 7
Hires. fix with Lanczos upscaler
20 Hires steps and Denoising strength 0.4-0.5
Upscale by 1.25
txt2img ugly face? just inpaint it 👍
Some example images has been face inpaint for better results.

If you like my LoRAs please consider to support me by buying me a coffee here so I can make a new LoRA faster.

Trigger word: goyoonjung

Checkpoints used in example: RealisticVision
Maybe works good with other models too but I haven't tried it yet so let me know.
Use it wisely. I don't responsible for anything you generated with this LoRa.

• The training materials provided by me are only for learning and communication, sharing AI painting experience.

• Do not use for commercial purposes, do not use for pornographic production, do not mass disseminate.

• If some people insist on using the portraits of public figures for pornographic materials, please do not name the source of this LORA, thank you very much.