Madison Beer SDXL (2018-2020)

Madison Beer is an American singer, songwriter, and social media influencer. She gained initial fame in 2012 when pop star Justin Bieber shared a link to one of her cover songs on his Twitter account. This exposure led to her signing with Island Records.

VERSION 2

I had made this so I could make images for a buddy of mine but he didn't like how some of the images turned out so I decided to further train this. Trained on 174 more images of Madison Beer leading to 203 images in total. I also used images from 2018 instead of just 2019-2020. This version should be better at just about everything. Better likeness and flexibility. It may still try to make hoop earrings if you use a higher strength so try putting "hoop earrings" in positive or negative prompt if needed. Keep strength at 0.9. 0.8 for more flexibility, 1 for more likeness. The trigger word is still "m4d1s0n" but I still recommend doing "m4d1s0n a woman". Keep Clip Skip at 1. I recommend starting off with this negative prompt: "ugly, deformed, airbrushed, photoshop, rendered, (multiple people), blurry, child, red lips, mouth open, makeup". The sample images are generated on 1024x1024 with Adetailer. This lora was trained on SDXL base. Should be able to do side views, back views, front views, etc. Should be pretty good at full body, upper body, and closeups. A big shoutout to steffangund, over half of the new images used were from his dataset which saved me some time, go check him out, he makes Loras of celebrities.:https://civitai.com/user/steffangund.

VERSION 1

Trained on 29 images of Madison Beer ranging from 2019 to 2020. This lora can do a good amount of angles and stuff. Pretty good at both closeup and full body photos. Keep the Clip Skip at 1. Strength of 0.9 is recommended, 0.8 for more flexibility, 1 for more likeness. Trigger word is "m4d1s0n". You can do this which is what I usually do: " m4d1s0n a woman". I recommend starting off with this negative prompt: "ugly, deformed, airbrushed, photoshop, rendered, (multiple people), blurry, child, red lips, mouth open, makeup". Sometimes it may try to blend hair with earrings, haven't had problems with it except one image. You can try putting earrings in negative prompt if it gives you a problem. Or even add it to your positive prompt. If you want a certain type of earring I'd go for hoop earrings, the one she wears the most.