Hi-onara! This is a lora for Angie Yonaga from Danganronpa V3.Unique features: The main point I try to stress in my Danganronpa loras is authenticity to the source material. It's fairly easy to get the basics of a character's style or clothing down. Characterization is a lot harder. Models have their own concepts of what a scared, or happy, or sad, or angry expression or stance would be. But a character in something like Dananronpa has a lot of careful attention put into ensuring their designs make all of those states unique. I go through every sprite, every CG, every official image I can find to caption and include them in the dataset in hope of best representing that. Whether it works out is another question, but it's what I'm aiming for and why these might be a little odd.Prompt style: I trained using natural language for the captions rather than tags. So natural language prompts will give you the best results. Tags will generally still work. However, natural language usually works best, followed by natural language with some additional tags,and finally just tags.<h1 id="specific-prompts-utca9pgna">Specific prompts</h1>I trained on a few specific image styles which you can replicate, to varying levels of success. I tried to include a range of the prompt options in the images so you can easily compare and contrast. But these are the major options.Danganronpa the Animation: Use 'This is an anime screencap from Danganronpa The Animation' in the prompt to emulate the art style from the anime series Danganronpa the animation.Danganronpa 3 (anime): Use "This is an anime screencap from Danganronpa 3."Sprites from Danganronpa: Mention 'danganronpa sprite style' in the prompt to try emulating the art style of Danganronpa. For example, 'this image is in the danganronpa sprite style'. Despite the name, I trained on full screenshots, with and without the GUI, with this description along with character sprite rips. So in theory it should be able to 'danganronpafy' a full image.Danganronpa S GUI: You can give it a try with this. "The image uses danganronpa S GUI style with the danganronpa sprite style. A speech bubble is in the bottom of the image and it says "This is test text!" The speech bubble is in front of Angie Yonaga. There's a box with a blue background with white stars, and a grey arrow pointing right, beside the speech bubble. The text in the box says "Angie Yonaga" and it is next to the speech bubble."Danganronpa V3 GUI: This will be less reliable than the Danganronpa 3 GUI: To the left of the image there is a yellow bar containing the text 'Music Track Name'. The digital text in yellow on the bar is in danganronpa V3 GUI style. There is a black bar with white text reading 'Chapter 3 Level 72'. In the bottom left corner of the image, there is another yellow bar with an angled black line down the left side, which reads 'Angie Yonaga' using a danganronpa V3 GUI style font. Below that, there is a white speech bubble using the danganronpa V3 GUI style containing speech that reads "Miu's Text" In the corner of the image, there is another danganronpa V3 GUI style text button saying 'AUTO.' The image is in Danganronpa sprite style with Danganronpa V3 GUI elements.<h1 id="clothing-itnb32orm">Clothing</h1>You can try different clothing options with prompts like the following:Standard outfit: Angie Yonaga wearing her uniform from Danganronpa V3. She has tanned skin. Her platinum blond hair is pulled back into two loose, wavy ponytails. She's wearing a white bikini top with pale blue frills and a white beaded necklace with a seashell pendant. She's also wearing a yellow smock over her white bikini top and pale blue skirt, with her school's crest on her shoulders. She has a white beaded bracelet on her left wrist.Standard outfit with closed smock: Angie Yonaga wearing her uniform from Danganronpa V3. She has tanned skin. Her platinum blond hair is pulled back into two loose, wavy ponytails. She's wearing a closed and buttoned yellow smock that reaches to her knees with her school's crest on her shoulders. The smock has loose sleeves She has a white beaded bracelet on her left wrist.Swimsuit: Angie Yonaga wearing her swimsuit from Danganronpa S. She's standing with her feet slightly apart. Both of her arms are straight by her side. Her platnum blond hair is in two loose wavy ponytails. She has a white, frilly bikini top with a blue bow on the front. She has tanned skin. Angie wears a light blue skirt and has a pink belt around her waist.Locations: I didn't try to train on any specific locations, but some of them appeared enough in the training data that you might be able to improve your results by specifying areas from Hope's Peak, Jabberwock island, or to a lesser extent the ultimate academy. Even just mentioning them might be enough to help push things in the right direction.<h1 id="misc-prjezr9i9">Misc</h1>Training material: Every official Angie Yonaga sprite and image from Danganronpa V3 and Danganronpa S. Additionally I found any additional official images and higher quality fan art. I also used a large amount of the backgrounds from Danganronpa 1,2, S and to a lesser extent V3. Finally, I used images from Danganronpa the Animation and Danganronpa 3 to allow for the style choices. However, since Angie isn't in those shows it's less exact than it might be for other characters like Makoto. Nevertheless, the anime styles seem to take quite well when used.Additional models and loras: While the lora works fine on its own, I really like how well it meshes with the Best of Flux: Style Enhancing LoRA, Flux-Ghibli_Style, osanai0804style, Mistoon_Anime, and FLUX_Black_Magic_XL anime style loras. I should have the gallery filled up with examples fairly soon after getting this lora uploaded.Special notes: The biggest problem areas I've noticed is the logo on her smock shoulders and limiting her bracelet to just her left wrist rather than both. It 'can' work perfectly fine, but it's a bit of a coin flip.Angie the bobblehead: I tried to train in those extreme left and right tilts of her body using the trigger phrase "Angie left lean" and "Angie right lean" but it doesn't seem to come out very reliably. Still, you might be able to get them using those phrases. I haven't had a huge amount of time to test this. But I think that you should probably be able to trigger it by just using descriptive language. Eventually, some examples should appear in the gallery.

Files with SHA256: 8af4a8f96b4525d434344a8f31f50a0ed28164aba6882d1404741f51a86e3cad