MOSS-Image-SFT-lora-Anime

Note: All model files below must be downloaded via the LM Downloader client. If you do not have the client, Download >>>

MOSS‑Image‑SFT‑lora‑Anime is an anime-style LoRA adapter tailored for the MOSS‑Image‑SFT version. It specially enhances the rendering of cartoon and anime visuals. There is no need to retrain the base model. Once loaded, it can quickly produce consistent 2D anime art, ideal for anime characters and sci-fi cartoon scenes. It must be used together with the MOSS‑Image‑SFT base model. Trigger word: anima, catgirl MOSS‑Image is an open‑source text‑to‑image model released by OpenMOSS. It supports both Chinese and English prompts. Its core strength is precise spatial control over objects and text within frames, which fits poster and graphic‑image production. It comes in two variants: Base and SFT. The Base version is the raw foundation for developers to perform further modification and training. It yields unsatisfactory outputs for ordinary end‑users and has limited instruction‑following performance. The SFT version is supervised fine‑tuned from Base. It is ready‑to‑use for regular users with improved instruction comprehension, more stable layout and text rendering. SFT is recommended for daily generation tasks. mossqiansuyan is a community‑built portrait‑focused LoRA for the SFT variant. While preserving the base model’s layout‑control feature, it enhances camera‑shot bare‑face effects to produce realistic portraits with light‑makeup and authentic skin texture.