(generation advice below) 👇
is a 24-year-old American actress and model who began her career in the adult industry with the film studio 'Many Vids' in 2019. She was born on June 10, 1999, in the United States. Coco stands at a very petite 147 cm tall, weighs a mere 39 KG, and has a figure size of 32B-24-34, with distinctive features that include her attractive figure, confidence and rejuvenating smile. She has expressed an interest in fashion, favoring clothing brands such as Calvin Klein, LOUIS VUITTON, Tommy Hilfiger, and Levi Strauss & Co. Her hobbies and interests include traveling, shopping, taking selfies, and internet surfing.
Most images have been generated with LORA weighted between 0.65 to 1.
Lower weighting needed often for headshots.
This LORA has captured all her blemishes so unless you want that, place some items in the negative.
Samples done with CyberRealistic v20 with sampling steps above 20, generally between 40 to 80 using Euler a, DPM++ SDE Karras and UniPC.
When testing a prompt, I often do a low lowres gen of multiple images, then choose the seed I like and upscale.
Portrait and close-ups work great using Hires. fix and 8X NMKD-Faces at 0.15 denoise strength.
Sometimes I just send it to Ultimate SD upscale after a quick render and run 2x at a scale of two, but often just the Hires. fix is good enough to capture enough details for the subject matter.
Ultimate SD upscale script used @ ~.15 - .4 denoising using 8x NMKD Faces or 4X UltraSharp redraw upscalers.
This LORA took me ~six tries, I've learned a lot from it about capturing what counts, only up from here.
In contrast to my previous work and going forward, this LORA should be more ready to produce subject matter details when the steps and denoise strength are stepped up during txt2img. Change your workflow accordingly.
The prompts continue to evolve. CyberRealistic is a reasonably flexible model for being realism focused and still capturing abstract concepts. I still think they have trumped the SD 1.5 game. 👑
Feedback appreciated. Please leave a review if you've used this LORA. 👍
Leave a comment if you are having issues getting good output, I'll try to respond.
Dataset gathered and trained on a single concept.
8 hour Kohya_ss training in 15,300 steps across 25 epochs @base dim (rank): 32.
1248x1248px training resolution base
Common bucketed resolutions @ 1024x1472, 1216x1216
1. The rights to reposted models belong to original creators.
2. Original creators should contact SeaArt.AI staff through official channels to claim their models. We are committed to protecting every creator's rights. Click to Claim
