Hi everyone,

This is a sucession to my previous LoRA for SDXL models. The previous LoRA was for XL models models however this works with 1.5 models.

Use terms like Korean, detailed, dynamic lighting (and other lighting related terms) for the best results. Dont overload the prompts though

I will upload a checkpoint merge after this post, the merge works best with this lora. It is a merge of MajicMix_Realistic and epicRealism_naturalSin 1.5 models. Name of the merge is majicRealism_mixSin (just so you dont get confused)

The recommended settings are:

• Steps: 35-40 (40 works best)

• CFG: 7 or 8 (according to the accuracy you want)

• Weight: anywhere between 0.4 to 0.8

• Upscaler: ESRGAN 4x

• Resolution: 768x1024px for portrait (always choose a higher resolution else the results will not be good)

Recommended models:

• epicRealism_nautralSin

• Majicmix_realistic

• majicRealism_mixSin (the merge which I created and attached to this post)

Follow me for more such loras, and message for any problems

Thank You!