Nosehook - MOSS Image LoRA - K3NK

Specialized LoRA for realistic nose hook imagery on the MOSS-Image base model. Trained on a curated dataset of 40 high-quality close-up photographs featuring diverse subjects with various nose hook styles, materials, and configurations.

Training Parameters

• Base Model: MOSS-Image (Diffusion Pipe)

• Dataset: 40 images

• Resolution: 720p

• LoRA Rank: 32

• Optimizer: AdamW8bitKahan

• Learning Rate: 5e-5

• Aspect Ratio Buckets: 7 (0.5 to 2.0)

Example Prompts

A close-up photograph of a woman with light skin and blonde hair wearing a green plastic nose hook attached to a thin wire pulled up by rope. Soft lighting highlights her skin texture against a dark blurred background.

High-resolution photograph of a woman with medium brown skin and dark brown hair in a bob with bangs. A metal nose hook is attached to her nose with a rope pulling it upwards. Her eyes are closed, dark eyeliner and mascara. Soft lighting, blurred background focusing on her face.

Close-up portrait of a woman with pale skin and black braided hair wearing a black leather collar with metal studs and a nose hook attached to a strap. Plain white background, bright even lighting highlighting her facial features.

A woman with light olive skin and dark brown hair wears a nose hook connected to a black leather strap running across her forehead secured with a metal ring. She has a black leather ball gag in her mouth. Dim lighting creates shadows against an indistinct background.

Close-up photograph of a young woman with pink hair, blue eyes, and light skin wearing a nose hook attached to a black strap. Her tongue is out showing a silver tongue piercing. She wears white earrings and a black choker. Bright lighting highlights her face and piercings against a dark blurry background.

Detailed close-up of a woman with fair skin and dark brown hair wearing a silver nose ring with a chain that pulls her nose upwards. Her lips are slightly parted showing teeth, subtle expression of strain. Soft lighting emphasizes facial features against a blurred indoor setting.

Disclaimer

This LoRA is provided as-is for lawful and responsible use. All faces and identifiable features were masked or anonymized during training. By downloading or using this LoRA, you accept full responsibility for its use and agree to comply with all applicable laws and platform rules.

Have fun! 😉