Dream Factory is a multi-threaded GUI manager designed for the mass creation of AI-generated art, targeting users who need to produce numerous images efficiently. It leverages the Automatic1111 Stable Diffusion backend, offering extensive customization and automation for batch image generation.

How It Works

Dream Factory acts as a sophisticated front-end and automation layer for the Automatic1111 Stable Diffusion repository. It manages multiple GPUs concurrently, allowing users to define complex generation workflows through a custom prompt file format. This format supports templating, random combinations, and conditional logic, enabling the creation of thousands of image variations from a single configuration. The system integrates directly with Automatic1111's features, including ControlNet, ADetailer, and various upscaling extensions.

Quick Start & Requirements

• Install: Requires Anaconda, then conda create --name dream-factory python=3.10, conda activate dream-factory, conda install -c anaconda git urllib3, git clone https://github.com/rbbrdckybk/dream-factory, cd dream-factory, python setup.py.
• Prerequisites: A working Automatic1111 Stable Diffusion webui installation is mandatory. Requires at least one NVIDIA GPU with sufficient VRAM (3GB minimum for 512x512, more for larger images/faster generation). Tested on Windows 10 and Ubuntu Server 20.04.
• Setup: Initial setup involves cloning the repo, running setup.py, and configuring config.txt with the path to the Automatic1111 installation. First run downloads ~2GB of files.
• Docs: Usage Instructions, Prompt File Reference

Highlighted Details

• Supports Stable Diffusion models SD 1.4 through SD3, including SDXL.
• Direct integration with ControlNet, ADetailer, and Ultimate SD Upscale extensions.
• Advanced prompt file format for complex batch generation, including templating and random combinations.
• Automatic insertion of model trigger words and embedding of EXIF metadata with generation settings.
• Remote management via a web UI, with an option to disable it for command-line operation.
• Civitai.com integration for model information within the prompt editor.

Maintenance & Community

• The project is actively maintained, with recent updates noted for compatibility with Automatic1111.
• Troubleshooting and update instructions are provided.

Licensing & Compatibility

• The repository itself does not explicitly state a license in the README.
• It relies on the Automatic1111 Stable Diffusion webui, which has an unclear license status, potentially impacting commercial use or derivative works. The author notes this lack of clarity.

Limitations & Caveats

• Compatibility is tied to specific versions of the Automatic1111 repository; significant changes there may break Dream Factory. A specific git checkout hash is provided for a known-compatible version.
• The web UI is not optimized for mobile devices.
• ControlNet functionality is noted as "WIP" with potentially rough editor references and bugs.