VVS Generator is a cutting-edge web application enabling users to generate short videos from textual inputs. Powered by the formidable GPT-4 for story generation and a sophisticated diffusion model for image creation, it brings textual stories to life in a visually stunning format.
Upon receiving user input, a concise three-sentence story is crafted using GPT-4 and saved as a text file.
- Code File:
main.py - HTML Template:
index.html
Before diving in, ensure Python is installed on your system. Subsequently, the necessary Python packages, such as Flask, torch, and diffusers need to be set up. [Diffusers: https://huggingface.co/runwayml/stable-diffusion-v1-5]
Use the following commands in your terminal or command prompt to install the requisite packages:
pip install Flask torch diffusers
pip install openai moviepy PillowYour OpenAI API key needs to be placed in the config.py file as shown:
API_KEYS = {
'openai': 'PUT_YOUR_API_KEY_HERE'
}
Code File: config.py
Execute the following command to get the app up and running:
python main.py
This kickstarts the Flask application. You can then access the web interface by heading to http://localhost:5000 on your web browser.
Usage 🖱️ Once on the web interface, input your text into the provided area and hit "Generate Video". The application will then weave a story, spawn images, morph them into a video, and showcase it on the webpage.
Sample video
Note: Ensure to substitute 'PUT_YOUR_API_KEY_HERE' with your actual OpenAI API key to ensure the application operates seamlessly.