Chroma / README.md
gokaygokay's picture
examples
45ddef4
---
title: Chroma
emoji: 🔥
colorFrom: blue
colorTo: pink
sdk: gradio
sdk_version: 5.14.0
app_file: app.py
pinned: true
license: apache-2.0
---
# Chroma Image Generator
A Gradio app for generating images using the [Chroma](https://huggingface.co/lodestones/Chroma) model by [lodestones](https://huggingface.co/lodestones).
## Features
- **High-quality image generation** using the Chroma model
- **Customizable parameters**: width, height, steps, CFG scale, and seed
- **Pre-cached examples** for instant preview and testing
- **Professional UI** with Gradio interface
## Example Prompts
The app includes 11 pre-configured example prompts with optimized parameters:
1. **High-fashion portrait** with anaglyph 3D effect
2. **Young man** with captain's hat and sunglasses
3. **Shiba Inu** with captain's hat and sunglasses
4. **Impressionistic sailing painting**
5. **Anthro wolf** in forest setting
6. **Dragon with galaxy wings** over castle
7. **Anthro dog** in NYC apartment
8. **Leopard geckos collage**
9. **Anthro wolf plushie collage**
10. **Telegram sticker** anthro fox
11. **Tiger eye** with 'Chroma' text overlay
## Caching
The app uses Gradio's built-in `cache_examples=True` feature to pre-generate and cache all example images. This means:
- **Instant results** when clicking on examples
- **No waiting time** for cached examples
- **Automatic cache management** by Gradio
- **Faster user experience** for trying different prompts
## Usage
1. **Manual input**: Enter your own prompt and adjust parameters
2. **Use examples**: Click on any example to load it instantly
3. **Customize**: Modify example prompts to create variations
4. **Generate**: Click the Generate button to create new images
## Parameters
- **Prompt**: Describe what you want to generate
- **Negative Prompt**: Describe what you want to avoid
- **Width/Height**: Image dimensions (512-2048px)
- **Steps**: Number of denoising steps (1-50)
- **CFG Scale**: How closely to follow the prompt (1-20)
- **Seed**: Random seed for reproducible results (-1 for random)
## Technical Details
- Built with **Gradio** for the web interface
- Uses **ComfyUI** workflow backend
- Powered by **Chroma** diffusion model
- Supports **GPU acceleration** with Spaces
- **Automatic caching** for optimal performance
## Running Locally
```bash
python app.py
```
The app will start and be available at `http://localhost:7860`
## Cache Directory
Examples are cached in the `gradio_cached_examples` directory. To rebuild the cache (if examples change), delete this directory and restart the app.
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference