Trained on Frieren Cosplay style photos. Hunyuan really wants to force it into anime style so you have to prompt it specifically not to get anime. Additionally I did not caption her normal adventure outfit specifically in hopes the model would learn it but it likes to go off the rails so it's best to describe the clothing you want. Trigger word is HyVid_Frieren to try avoid know tags within Hunyuan.Should understand these triggers fairly well too: Holding staff, Black choker, partially naked.You can get the clothing dissolved concept inconsistently with: "Skirt and shirt have holes in them", "Pouring a white liquid from a purple potion bottle on herself" This is the best base prompt I've found to get consistent character appearance. "real life, real iphone video of young woman, HyVid_Frieren, white hair in long pig tails. elf ears. button nose. green eyes. wearing white shoulder cape with gold trim over a black and white striped shirt, white skirt, black leggings, brown knee high boots, Holding staff"Trained on 55 images, 30 epochs, 8250 steps.See my latest article about my video comparison tool I'm developing to test Loras. <a target="_blank" rel="ugc" href="https://civarchive.com/posts/11389777">https://civarchive.com/posts/11389777</a>

Frieren - Beyond Journey's End - Cosplay Lora

Trigger word is HyVid_Frieren to try avoid know tags within Hunyuan.Should understand these triggers fairly well too: Holding staff, Black choker, partially naked.You can get the clothing dissolved concept inconsistently with: "Skirt and shirt have holes in them", "Pouring a white liquid from a purple potion bottle on herself" This is the best base prompt I've found to get consistent character appearance. "real life, real iphone video of young woman, HyVid_Frieren, white hair in long pig tails. elf ears. button nose. green eyes. wearing white shoulder cape with gold trim over a black and white striped shirt, white skirt, black leggings, brown knee high boots, Holding staff"Trained on 55 images, 30 epochs, 8250 steps.

Files with SHA256: c39039aba204e1ea64895ee9c4614ca532d183523db3b6da93e2be0376733ba7