## Model Introduction

Bento & Love Box is a high-quality SDXL Checkpoint based on the Illustrious architecture.

Unlike standard models, this checkpoint is fine-tuned to produce a specific Romantic & Slice-of-Life aesthetic directly. It is designed to create warm, nostalgic, and detailed anime-style illustrations without needing additional LoRAs.

Key Features:

* Style: Modern Anime / Shojo Manga / Soft Lighting.

* Themes: School life, bento boxes, romantic interactions, and expressive character portraits.

* Architecture: Based on Illustrious SDXL (High adherence to prompts).

## 🔧 Recommended Settings (Usage Guide)

To achieve the best results shown in the showcase, please use the following settings:

* Trigger Word: None required (Just prompt normally!)

* VAE: Built-in / Automatic (or use `sdxl_vae.safetensors` if colors look dull)

* Sampler: `Euler a`, `DPM++ 2M Karras`, or `DPM++ SDE Karras`

* Steps: `25` - `35`

* CFG Scale: `5.5` - `7.0`

* Resolution: `1024x1024`, `832x1216` (Portrait), or other SDXL standard ratios.

* Clip Skip: 2

## 📝 Prompting Tips

Since this is an Illustrious-based model, it understands natural language and danbooru tags well.

Example Prompt 1 (Character Focus):

> `1girl, school uniform, holding a bento box, sweet smile, blushing, rooftop, blue sky, clouds, sunlight, lens flare, anime style, highly detailed, masterpiece`

Example Prompt 2 (Scene Focus):

> `close up shot, a delicious bento box with octopus sausages and rice, love letter on the desk, classroom background, warm afternoon lighting, depth of field`

## ⚠️ Notes

* Negative Prompt: standard SDXL negative prompts (e.g., `low quality, worst quality, bad anatomy, text, watermark`).

* This is a standalone checkpoint. You generally do not need a refiner.

Illustrious Model: Bento & Love Box - DOLL