Bring your photos to life with MiniMe SD, a finely tuned Stable Diffusion 1.5 LoRA trained to generate fun, expressive, and stylized caricature portraits. Whether you want playful character avatars, cartoon-style exaggerations, or unique profile illustrations, this LoRA is designed to capture personality with a bold and vibrant twist.

+VE:

chibi style, short legs, short feet, short hands, top-heavy, mini body, small body, big head, big head doll, chibi, tiny human being, minimesdfinal, 1girl_minime, 1girl, solo, looking at viewer, <lora:MiniMeSDFINAL:0.6>. detailed hair strands, detailed iris.

OR

+VE:

chibi style, short legs, short feet, short hands, top-heavy, mini body, small body, big head, big head doll, chibi, tiny human being, minimesdfinal, 1girl_minime, 1girl, solo, looking at viewer, <lora:MiniMeSDFINAL:0.6>. detailed hair strands, detailed iris.

-VE:

worst quality, low quality, normal quality, lowres, watermark, logo, username, artist name, signature, jpeg blemish, jpg artifacts

✨ Features

• 🎭 Caricature Style – Converts real faces into exaggerated, yet recognizable cartoon-like versions.

• 🖼️ High Versatility – Works across portraits, busts, and full-body renders.

• 🎨 Compatible with SD 1.5 – Optimized for Automatic1111, ComfyUI, and other SD pipelines.

• 🔧 Easy to Control – Adjust weight (0.6–0.9 recommended) to fine-tune caricature intensity.

• ⚡ Creative Ready – Great for profile pics, comics, posters, merch mockups, or just for fun.

📥 Model Details

• Base Model: Stable Diffusion 1.5

• Format: .safetensors

• Training: Custom dataset of caricature references (final refined version)

• Recommended Sampler: DPM++ 2M Karras

• CFG Scale: 6–8

• Resolution: Works best at 512x512–768x768

🚀 Use Cases

• Social Media Avatars

• Comics & Storyboards

• Poster & Merch Concepts

• Lighthearted Branding

• Fun Experiments

⚠️ Notes

• This model is meant for artistic/entertainment use only.

• Always respect personal consent when using real-life likenesses.