All three versions I made of this model came out really great. Seems to have shaved a few years off the age based on the training data, but overall seems fairly flexible in all variants. Usually will depict the intended character accurately. Share the same training data, but use different base models and training parameters:-SpencerLaval-All,-SpencerLaval_CivitAi_allPruned data set that only includes images with 1boy. Less likely to produce weird artifacts and more likely to get the face/body correct. Much less flexible with prompts that include more than 1boy: -SpencerLaval_CivitAi_1boyThis model may include some images that are NSFW. Be careful when generating images as CivitAi does not allow NSFW images of real people.

Files with SHA256: 29f55241240760d39768facf307a291a241e37abd035eb8adccc9a8eb2eeb62c