The style of Fonzawa.Trigger word is <code>@fonzawa</code>. Petra (the artist's primary OC) uses <code>petra \(fonzawa\), yellow eyes, blue hair, low twintails, ahoge, elf</code>. Same dataset as my NoobAI <a target="_blank" rel="ugc" href="https://civarchive.com/models/2473482/fonzawa-style-noobvp">version</a>.A Katawa Shoujo charpack redux is well under way (dataset prep is more than halfway done), but it's still a while off. I decided to take a quick break and train this LoRA in the meantime.19.09.2026 - Added regular LoRA version in addition to the DoRA.If you like the style, please support the <a target="_blank" rel="ugc" href="https://www.reddit.com/user/Fonzawa/submitted/">original author</a>.

Fonzawa style AnimaB1

fonzawaStyleAnimab1_epoch10_3224010.safetensors

fonzawaStyleAnimab1_dora.safetensors

DoRA

LoRA

I realized DoRAs had downsides I hadn't previously considered, so now I'm remaking LoRA versions of my DoRA models. In short:<ul><li>DoRA works better alone and on the same model it was trained for</li><li>LoRA works better in combination with other adapters or on checkpoint finetunes</li><li>LoRA also appears to work better on Anima 3.8B</li></ul>

Files with SHA256: bd8628b63b10e8e8b70a5da3de5b57d06365edeb7cec8e730e2a75e3a4058ded