FableForge generates illustrated children's picture books from a single text prompt. It targets users interested in AI-driven content creation, offering a streamlined workflow for producing visual stories. The primary benefit is the rapid generation of complete, multi-page narratives with accompanying images.

How It Works

The system leverages OpenAI's function calling to parse a user's prompt and generate a story outline. This outline is then fed to Replicate's API, which utilizes Stable Diffusion for image generation based on detailed prompts derived from the story. LangChain orchestrates these components, managing the flow of information. Generated images and their associated text prompts are stored in Deep Lake, enabling multi-modal data management and a web-based UI for review.

Quick Start & Requirements

• Install dependencies via requirements.txt.
• Set OpenAI and Replicate API keys, along with Deep Lake token and dataset path, in keys.env.
• Run the application using streamlit run main.py.
• Requires Python, OpenAI API key, Replicate API key (with free credits available for new users), and an Activeloop Deep Lake account and token.

Highlighted Details

• Utilizes OpenAI function calling for prompt parsing and story generation.
• Employs Replicate API for Stable Diffusion image generation.
• Stores images and prompts in Deep Lake for multi-modal data management.
• Orchestrated using LangChain.

Maintenance & Community

The project is maintained by e-johnstonn. Further community or roadmap details are not specified in the README.

Licensing & Compatibility

Licensed under the MIT License, permitting commercial use and integration with closed-source projects.

Limitations & Caveats

The project relies on external APIs (OpenAI, Replicate) which may incur costs beyond initial free tiers. The README suggests connecting to a local or cloud-based Stable Diffusion setup for improved results, implying the current Replicate integration might have limitations in output quality or customization.