This is my personal-use style LoRA that I created so I wouldn't need an <a target="_blank" rel="ugc" href="https://civarchive.com/articles/26565/bvga-styles-models-i-use-to-generate-my-images">8-style-LoRA stack</a> to get the style I liked. With the release of Anima it's even more important to me, as it's the only way I'm able to transfer that style over (and seems to work pretty well at doing that!). My aim with this is to address the generic-looking oversaturated, flat, glossy, and plastic look of the anime checkpoint models. It does that in a number of ways, such as improving textures, depth of field, and lighting, with a painterly and semi-realistic feel. Anima version notes:No LoRA trigger tag is needed.Has a bit of an issue with fingers (and blurry features at a distance) under certain situations, which I'm currently researching the cause of (my dataset images are clean, so it's something else). Fixing this is my main focus with the next version. Until then, using detailers should help. Illustrious version notes:My intention was to have a single LoRA that both strongly captured that style while being cleaner and more flexible to use. But after weeks of experimentation, I've come to the conclusion that I need to split this into 2 separate LoRAs.Comparison Chart: <img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/34c61af1-e289-490a-9c49-8d7446096b1c/original=true/34c61af1-e289-490a-9c49-8d7446096b1c.jpeg" />v2.0 is a typically trained style LoRA, which has a softer style and doesn't alter prompt understanding, ideal for flexible use and mixing with other style LoRAs. (v2.0 is what most people will want to use)v1.0 is an intentionally "overtrained" model that aims to capture that style as strongly as possible, at the expense of prompt sensitivity. As a result, with v1.0 you'll need to experiment with prompt weights, e.g. (character name:1.3). (v1.0 was also an experimental proof of concept, so the quality of the dataset wasn't perfect.)

Core :: bvga Style

bvga_Core_v2.0-Rank96.safetensors

IL v2.0

Anima v1.0

IL v1.0 (Concept)

A style LoRA I made to address the oversaturated, flat, glossy, and plastic look of the anime checkpoint models. It does that in a number of ways, such as improving textures, depth of field, and lighting, with a semi-realistic and painterly feel.

Sampler Recommendations:
(for standard style)
30 Steps
5 CFG
DPM++ 2m karras
Checkpoint: Wai Illustrious v16

v2.0

v2.0 is a typically trained style LoRA, which has a softer style and doesn't alter prompt understanding, ideal for flexible use and mixing with other style LoRAs. (v2.0 is what most people will want to use)v1.0 is an intentionally "overtrained" model that aims to capture that style as strongly as possible, at the expense of prompt sensitivity. As a result, with v1.0 you'll need to experiment with prompt weights, e.g. (character name:1.3). (v1.0 was also an experimental proof of concept, so the quality of the dataset wasn't perfect.)

Files with SHA256: 258ded01691cff68c9f8ff30fb69ce0ac7e735249368ca1ffff8f2c0d07804f7