This project leverages the power of web scraping, generative AI, automated editing, and Python scripting to transform scraped short stories into engaging scripts, along with accompanying visuals. The key components include:
- Short Story Scraping: Collecting a rich dataset of short stories from online sources.
- Generative AI Integration: Using ChatGPT and similar AI models to generate compelling scripts.
- Audio content : Use of pretained voice/sound models to generate audio content for short stories.
- Automated Editing: Implementing techniques to streamline editing, ensuring coherence and quality in the generated outputs.
- Publication: Developing a process for publishing the final scripts and images (formats, platforms to be determined).
-
Prerequisites:
- Python 3.11
- A ChatGPT API key (obtain from OpenAI)
- Additional libraries (will be listed as the project develops)
-
Installation: Clone this repository: git clone https://github.com/Dan-Irl/GenAI-Stories.git
Navigate to the project directory: cd short-story-generator Install dependencies: pip install -r requirements.txt'
- Tasks
- Identify target websites for short story scraping (Currently focusing on integrating reddit API).
- Develop automatic web scraping scripts.
- Create a structured data storage solution (e.g., database or structured text files).
- Tasks
- Research and experiment with suitable generative AI models (ChatGPT, Midjourned, local diffusion models, Hugging face models).
- Develop prompts and workflows for interacting with the AI models for script and image creation.
- Tasks
- Combine script reading,
- Explore natural language processing tools and techniques for automated editing.
- Tasks
- Research and select platforms for publication. Currently focusing on youtube shorts and Tiktok.
- Implement an automated or semi-automated publication process.