Trained on the work of Nononononoway. It pushes images toward a glossy 2.5D style and excels with arthropods, monsters, post-apocalyptic settings, and machines. It will often be able to push anime based character loras into a more 2.5D presentation as well without forcing too much of the dystopian properties.The Flux 1 version is best used with narrative style prompts. The Illustrious and Pony models work better with tags but should support a mixture of tags and natural language. The Chroma and Anima lora should work equally well with tags, narrative prompts, or a combination of both. Likewise the z-image and Flux 2 loras should work equally well with tags, narrative prompts, or both. Though narrative heavy with a few keywords tends to work especially well. Though with z-image prompts if you're using both then it's best to begin with the trigger phrase, then the natural language description, THEN the tags. The chroma and z-image lora also benefit more from a description of the art style you want. With something like "in the Nononononoway style with a combination of x and y that promotes z" or the like.<h1 id="prompting">Prompting</h1>Just using a tag of 86_(nononononoway) should be enough for this style with all of the loras. However, you can often strengthen it a bit with "86_(nononononoway), This nononononoway style digital illustration is done in a style showcasing detailed shading to evoke gothic urban horror." especially with the more narrative focused models like z-image.<h1 id="examples">Examples</h1>Scorpion girl: 86_(nononononoway), nonoscorpiongirl, A detailed anime-style illustration of nonoscorpiongirl a pale young woman with long flowing silver-white hair and grey eyes. Her face and neck show intricate crack-like markings on her humanlike exoskeleton resembling porcelain damage or scars. She wears a mechanical mouth restraint with visible metallic teeth elements across her lower face. She's dressed in a flowing white Victorian-style dress with corset lacing at the chest. Her dark amber colored scorpion tail can be seen behind her extending from her lower back with her tail's stinger raised to waist level. Rendered in the nononononoway style the artwork blends hyper-detailed almost photorealistic textures with the clean expressive lines of anime character design creating a surreal high-contrast scene of ethereal dark fantasy paired with dramatic cinematic lighting. , braces, cracked cheeks, cracks on neck, 1girl, arthropod_girl, cracks on skin, cross-laced_clothes, cross-laced_dress, dress, face cracks, girl made of porcelain, grey_eyes, half-closed_eyes, long_hair, ringed_eyes, scorpion_girl, scorpion_tail, tail, white skin, white_dress, white_hairMoth girl: 86_(nononononoway), mothgirlnono, The pale-skinned figure of Mothgirlnono stands with long flowing silver-blue hair and piercing black eyes. Handcuffs dangle from her wrists. Multiple pairs of enormous iridescent butterfly wings extend from her back. Her butterfly wings are detailed with intricate veining and shimmering with holographic hues. She wears a dark blue short sleeved blue shirt paired with blue cargo pants. , 1girl, antennae, arthropod_girl, barefoot, blue_eyes, blue_hair, butterfly_wings, cuffs, expressionless, handcuffs, insect_wings, long_hair, moth_girl, multiple_wings, pants, cargo_pants, blue_pants, short_sleeves, solo, wings<h1 id="versioning">Versioning</h1>The 1.5 branch is an experiment with a fraction of synthetic data on top of the prior dataset used with 1.0. I'm still not really sure what I think of it in comparison. Though it did provide a few new concepts (Tsumugi Shirogane from Danganronpa V3 is more or less promptable from the 1.5 z-image lora for example).

Nononononoway Style

Just retrained the same dataset as the previous anima preview model on the anima 1.0 base model.

Files with SHA256: a979a3bd42887e66b52dab61fc2db51df09deb72eef8455d9cca807c5d204da5