Bunnii O'Hare from Fortnite — all six of her in-game styles in one file.Base, Hopper's, Auric, Phase, Rimebound and Cosmic Legacy, each on its owntrigger word, plus the character out of costume.WHY I MADE THISThis is only my second LoRA. I do this for fun and because it is a genuinelygood technical puzzle — that is the whole reason, there is no plan behind it.It started as an entry for a bounty here on Civitai. I was too slow. The bountyclosed while my dataset was still unfinished and it went to someone else, fairly.But by then I had looked properly at what already exists for this character, andboth public LoRAs stop at two styles. I wanted all six in one file. Nobody hadmade that, so I finished it anyway, mostly out of stubbornness.THE PART THAT WAS ACTUALLY HARDNot the training. The captions.If a tag appears on EVERY image of a concept, the model stops treating it as atag and folds it into the trigger word. It quietly becomes something you can nolonger turn off. All 114 of my reference images were 3D game captures, so <code>3d</code>sat at 114/114 — meaning the LoRA would have rendered everything as a gamescreenshot forever, whatever you prompted.What does not work is naming the thing you do not want. What works isCONTRADICTING IT WITH IMAGES. I generated 28 extra pictures in flat animeshading and folded them into the set. <code>3d</code> dropped to 80%, and the render stylebecame something you can actually ask for or refuse.Same trick for the suit: 18 out-of-costume images took <code>bodysuit</code> from 100% downto 87%. That is the only reason this LoRA can put her in a sweater and jeansinstead of always armouring her up.Two styles never got that treatment. Auric and Hopper's have no anime referencesanywhere, so <code>3d</code> is still at 100% for those two tokens and they will always lookmore like game capture than the other four. It is an honest asymmetry and it isin the limits below, with everything else I know is wrong with it.THE TRIGGERS bunniibase Bunnii O'Hare navy / white / gold bunniihoppers Hopper's Legacy dark teal / white / pink bunniiauric Auric Legacy black / gold bunniiphase Phase Legacy iridescent / holographic bunniirime Rimebound Legacy cyan / white / silver bunniicosmic Cosmic Legacy starfield / nebula bunnii her out of costume (use this one on its own)You can put <code>bunnii</code> in front of a style token, but you do not have to — onpaired seeds, dropping it moves the image less than simply rerolling does.One thing please keep in your prompt: <code>adult woman, mature female</code>. She has amascot silhouette and that wording is on all 143 training captions. It is therefor a reason.

Bunnii O'Hare - all 6 styles (Anima) - character LoRA

SETTINGS Base Anima-Base v1.0. Trained against Base, so it carries over to Anima-Turbo, JANIMA and the all-in-one; the reverse is not true. Strength 1.0 Sampler er_sde Scheduler beta CFG 4.5 Steps 30 Size 832x1216 portrait, or 1216x832 There is no clip skip. Anima encodes with a Qwen3 text encoder and an adapter, not CLIP. Any clip-skip value is ignored.TAG LISTS — one per style Each style has a costume block that was trained on every single one of its images. The trigger already carries it: writing it out changes nothing, and it is reproduced here so you know what the token contains. bunniibase purple hair, blue eyes, purple lips, white hood, gold trim, navy blue bodysuit, white chest plate, white gloves, white armor, gold armor, gold emblem, white boots, rabbit tail, bodysuit, high heels bunniihoppers blonde hair, blue eyes, pink lips, white hood, pink trim, dark teal bodysuit, white chest plate, white gloves, pink armor, white emblem, pink boots, rabbit tail, bodysuit, high heels bunniiauric blonde hair, brown eyes, gold lips, gold hood, orange lining, black bodysuit, gold chest plate, black gloves, gold armor, gold emblem, gold boots, rabbit tail, bodysuit, high heels bunniiphase pink hair, pale lips, iridescent hood, iridescent bodysuit, holographic, pearlescent, opalescent, iridescent armor, iridescent gloves, white boots, rabbit tail, bodysuit, high heels bunniirime purple hair, cyan lips, white hood, silver trim, cyan bodysuit, white chest plate, cyan gloves, silver armor, white emblem, white boots, rabbit tail, bodysuit, high heels bunniicosmic purple hair, dark purple lips, starfield, galaxy print, nebula bodysuit, cyan gradient, magenta gradient, dark hood, dark gloves, dark armor, dark boots, rabbit tail, bodysuit, high heels bunnii purple hair, blue eyes, purple lips (out of costume — no style token, no bodysuit)WHAT IT DOES — tested, with the pass and fail conditions written down first Ten conditions, two seeds each, criteria fixed BEFORE generating. Eight passed, one failed, one works poorly. Works action and motion; complex environments (night streets, rain, crowds); civilian clothes with no costume piece bleeding through; forced anime rendering; stacking with a style LoRA; two characters in one frame; nudity with correct anatomy; hands holding an object; alternate styles under scene lighting rather than flat backdrops. Fails hood down. 1 success in 4 attempts. See below. Poorly two characters in DIFFERENT styles — you get two figures, but the colourways bleed into each other and the composition drifts toward a character sheet.WHAT IT DOES NOT DO — known defects, named rather than hidden HOOD DOWN IS UNRELIABLE — 1 success in 4. <code>hood down</code> appeared on 43 of the 125 in-costume captions against 82 for <code>hood up</code>. It is not absorbed, it is outvoted. Two things help: - name NO hood in your prompt. <code>white hood</code> together with <code>hood down</code> gave 0 successes in 2; removing it gave 1 in 2. - expect to reroll the seed. THE EAR LINING COMES OUT CYAN ON AURIC AND COSMIC. Auric's lining should be orange and Cosmic's should be a starfield. Five of the six styles share a cyan lining in the reference captures, so the minority cases had no contrast to separate against. No tag fixes this. <code>bunniiauric</code> AND <code>bunniihoppers</code> CARRY 3D RENDERING AND CANNOT SHED IT. <code>3d</code> appears on 23/23 auric captions and 18/18 hoppers captions, against 26/29 for base. At 100% a tag is absorbed by the token, so these two styles will look more like game capture than the other four. This is asymmetric and it is a limitation, not a preference. DEFAULT RENDERING IS ANIME, NOT GAME CAPTURE. Despite 114 of the 143 training images being 3D game captures, the LoRA renders in anime by default. 28 counter-examples were generated on purpose to make the render style detachable, and they over-corrected. Whether it can produce the game look on demand has NOT been tested. OUT OF COSTUME, IDENTITY IS THIN. Without the suit the character is carried by hair, eyes and lips only. There is no <code>bodysuit</code>-free variant of the five Legacy styles: <code>bodysuit</code> sits at 100% in all six style concepts, so it is absorbed by the token. For higher ratings, OPEN the suit <code>unzipped bodysuit</code>, <code>bodysuit pulled down</code>) rather than removing it.HOW IT WAS MADE Trained with kohya-ss/sd-scripts (anima_train_network) against Anima-Base v1.0, on a single 8 GB card. 1 h 29, peak VRAM 7997 MB of 8192. unet-only, dim/alpha 32, CAME (weight_decay 0.01), lr 7e-5 / unet 2e-5, cosine with 5 warmup steps, sigmoid timestep sampling, l2 loss, 1024 with aspect-ratio bucketing and no upscaling, batch 1, bf16, 1600 steps, seed 42. Checkpointed every 200 steps so the epoch could be chosen by looking rather than by taking the last one. Dataset: 114 in-game reference captures across the six styles, plus 28 generated images whose only job was to introduce VARIATION the captures did not have. Without them <code>3d</code> sat at 114/114 and <code>bodysuit</code> at 114/114 — constant, therefore absorbed, therefore imposed on every output. With them: 80% and 87%, therefore detachable. Verified afterwards, not assumed. Seven concepts: the six styles plus one out-of-costume concept that carries no style token, because there is no style to designate on a woman in a t-shirt and attaching one would teach a false association. Captions were built by merging a hand-written fact sheet per style with what a WD14 tagger actually saw in each image, so the vocabulary describes the pixels rather than a marketing blurb.

Files with SHA256: 4ebf8af8fd51a3392a0454a98da659e0c282816630a025cb3b12359dbca57af7