MOSS-Image LoRA Targets Uncensored Portraits
This community LoRA adapts MOSS-Image for adult photographic portraiture, poses, and indoor scenes, with compatibility across MOSS-Image-SFT and Base. It is a lightweight adapter rather than a standalone model, designed for local ComfyUI or Diffusers workflows. [Hugging Face](https://huggingface.co/axiomofmind/Moss-image-nsfw-uncensored-lora)
This is a practical specialization layer for MOSS-Image, making the recently released 5B ecosystem more useful for creators seeking fewer content restrictions.
- –The adapter is optimized for realistic adult portraits and close photographic compositions, filling a clear gap in the base model’s community ecosystem.
- –Users should pair it with MOSS-Image-SFT for the preferred results, while Base compatibility broadens experimentation. [Model card](https://huggingface.co/axiomofmind/Moss-image-nsfw-uncensored-lora)
- –The roughly 71 MB adapter keeps distribution easy, but the underlying 5B model and inference stack still determine actual laptop memory requirements.
- –ComfyUI users should use the provided native ComfyUI weights; the standard Diffusers adapter uses different key names and may fail to load directly. [Hugging Face](https://huggingface.co/axiomofmind/Moss-image-nsfw-uncensored-lora)
- –Its Apache-2.0 licensing improves reuse, but users still need to enforce adult-only, consensual, and lawful-generation safeguards.
DISCOVERED
1h ago
2026-10-27
PUBLISHED
1h ago
2026-10-27
RELEVANCE
AUTHOR
0x0SojalSec