My Links: twitter, discord, IG
Support my efforts at @ ko-fi.com/fitcorderai
Subscribe for FBB images @ https://patreon.com/fitCorderAI
My new checkpoint is RadianceApex. It is fine tuned for aesthetic, dramatic images.
Fitness checkpoint is FitCorderMix. Realistic checkpoint is fCBlendMix. Anime checkpoint is fCAnimeMix.
Color issues? Download a vae, guide here. Pictures never look sharp? Download a ERSGAN, guide here. Issues getting good portraits? Click here.
Image Generation Tutorial Series
Click here for a detail image guide
Description
FAQ
Comments (17)
Works well. Hopefully the clothed appearance can be improved. The veins seem to mostly disappear when the breasts are not entirely exposed.
Would love a version with the veins but not the bimbo. Bimbo loras are a dime a dozen.
Read the description
There are no bimbos in the training data.
@fitCorder So this was trained with natural tits? If so, why do the preview images mostly look like boobjob tits?
@dustystu If you can point me in the direction of a source for smaller busts with prominent veins i'll train and post an update to the version. The ones I could easily find were the ones I used. They were trained mostly on natural albeit larger/swollen breasts.
@fitCorder I see your point. Perhaps a subreddit of pregnant woman, and pick the smallest looking boobs with veins?
have you tried the update?
@fitCorder was gunna try the loha soon. sorry didn't see your reply earlier.
Do you have a version for the skin?
if you jack up the emphasis (veins:1.4) you can get the effects on the skin overall but it may overdo it.
绝对大胆的尝试，希望可以在此基础上完善对真实皮肤的模拟，期待你更精彩的成果
For the life of me I could get neither 1.1c or 1.1b to work. I do have the lycoris extension, put them in, they were recognized. When I increase the weights to extremes, some of the veins do show up but on completely incorrect areas. Tried on multiple models including ghostmix, revanimated and magicmix fantasy. To troubleshoot I even removed all other prompts besides the trigger words and lycoris/lora themselves, and it still does nothing. Not sure what's going wrong.
So sorry. This was a tough one but there is no inpainting in my examples you should be able to recreate them.
any chance you'd share/p;ost your training data for this? I'd love to keep it going and experiment to really emphasize this. underserved concept here!
Would you be willing to bring this back for XL or Pony? Or the training data, as mentioned by another?
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.