# Rosita LoRA Description

## Character Description
Rosita is a stunning European woman with sun-kissed skin, wild curly black hair, and an hourglass figure. She radiates playful confidence and elegance, often depicted in sheer lace bodysuits or nude with artistic drapery, posing against velvet backdrops or lounging in dreamy, painterly settings.

## AI Image Generation Model: Z-Image-Turbo
Z-Image-Turbo is a high-performance AI image generation model optimized for fast, efficient creation of high-quality images. It supports various artistic styles, resolutions, and is designed for turbocharged inference, making it ideal for rapid prototyping and detailed image synthesis.

## Training Details
- LoRA generated with approximately 2500 steps
- Network: LoRA (linear: 32, linear_alpha: 32, conv: 16, conv_alpha: 16)
- Training steps: 3000 (configured)
- Batch size: 4
- Optimizer: adamw8bit
- Learning rate: 0.0001

## Optimum KSampler Settintgs for best results
- Base model: Z-Image-Turbo
- Trigger word: rosita
- Sampler: Steps 9, Sampler er_sde, Scheduler beta
- The lora is applied only if the trigger word is part of the prompt.
- SeedVR2 Upscaler for best results.

## Disclaimer
The training images were generated using Juggernaut XL Ragnarok and edited with Qwen Image Edit 202509.