open-genmoji is an open-source AI image generation project focused on creating high-quality, emoji-style images using diffusion models and LoRA fine-tuning.
It provides tooling for generating, fine-tuning, and customizing emoji-like visuals using modern generative AI workflows.
- Emoji-style image generation
- LoRA-based fine-tuning support
- Prompt engineering utilities
- Custom metaprompt workflows
- Image post-processing (e.g., resizing)
- Model download and setup scripts
open-genmoji leverages diffusion-based generative models and:
- Applies LoRA (Low-Rank Adaptation) fine-tuning
- Uses structured prompts and metaprompts
- Generates stylized emoji-like images
- Applies optional resizing and post-processing
Clone the repository:
git clone https://github.com/<your-username>/open-genmoji.git
cd open-genmojiCreate a virtual environment:
python -m venv venv
source venv/bin/activateInstall dependencies:
pip install -r requirements.txtDownload required models:
python download.pyGenerate an image:
python generateImage.py --prompt "happy robot emoji"Resize output images (optional):
python resize.pyLoRA fine-tuning utilities are available in the lora and finetuning directories.
Adapt the provided scripts to your dataset and model configuration.
- Custom emoji creation
- AI-generated sticker packs
- Stylized avatar generation
- Creative experimentation with diffusion models