Another experimental model with a slightly modified clip from dhxl.
How run:
cfg 1-5, step 8-20, sempler euler_a, sheduler beta.
Description
FAQ
Comments (10)
very good Checkpoint V1, good understanding, very fast, top picture. Only low-waist leggings are always highwaist ;) can you please try to fix in V2? :) The more I test it, the more I like it. Brilliant colors. Understands positions. It's easy to use. I love it. Thank you for this, it's for me the best checkpoint at the moment.
and many Fetish out of the Box.. :) i love it so much, thank you thank you thank you :)
@Gorean ;)
Unfortunately, distant faces, such as full-body photos while standing, are very poor. The closer the face gets, the better it gets. Close-ups are incredibly perfect. Do you have a solution for this? I've tried it with many LoRa systems but haven't found anything. But all the LoRa systems I've used so far, whether Pony, IL, or SDXL, have worked with your checkpoint. :) That's fantastic. The only problem with the faces out of the box is not so good.
Edit: i had the wrong word translated, English is not my native :) Im sorry
@Gorean Yes, the faces need some work, I know this problem, I will try to solve it in future versions.
I found a solution for the low-waist leggings problem, "highleg leggings", then it is the way I want it :)
One of the most responsive/understanding checkpoints I've tried. Understands different races, sizes, clothing almost perfectly. Lighting is consistent and solid. I haven't had to use any LORA's either so far... great job overall!
Thanks for your feedback
Very cool realistic model, I was pleasantly surprised, because this is my first NAI model, I compared it with a popular realistic model and the difference is colossal, I downloaded it to iPad via 8step lora and dnd lora via tcd pretty good result.
Glad you liked it, hope the next versions will only get better.