📌 Webio-women | Realistic Women LoRA

Base model: runwayml/stable-diffusion-v1-5
Format: LoRA (.safetensors)
File size: 72 MB
Date trained: 2025-08-23
Kohya module: networks.lora
Clip skip: 2
Resolutions used: 576×512, ??0×320, ??0×512, 512×512, …
Dataset size: 2292 images

🔹 About this LoRA

Webio-women is a LoRA trained to generate realistic portraits of women with natural detail and lifelike quality.
It was trained on a curated dataset of 2,292 images across multiple resolutions, ensuring strong versatility in poses, lighting, and environments.

This LoRA is designed for photorealistic results while keeping flexibility for different outfits, moods, and backgrounds.

🔹 How to Use

• Recommended weight: 0.7 – 1.0

• Optimized for SD 1.5 checkpoints (works well with popular realism models such as ChilloutMix, RealisticVision, etc.)

• Use Clip skip 2 for best results.

🔹 Example Prompt

<lora:Webio-women:0.8>, photo of a beautiful woman, cinematic portrait, natural skin texture, soft lighting, detailed eyes, ultra realistic photography

Negative Prompt:

blurry, low quality, distorted, bad ???????, extra limbs, watermark, text

🔹 Notes

• No trigger words required.

• Works out-of-the-box for a wide range of photorealistic female portraits.

• No major issues observed during testing.