API that fetches current news and generates future news predictions using a local LLM.
- Fetches current news from various sources
- Generates hypothetical future news articles based on current trends
- Supports different time frames (day, week, month)
- Supports different news styles (neutral, optimistic, pessimistic)
- Built with FastAPI for high performance and easy API documentation
- Uses local LLM through Ollama for generation
- Python 3.10 or higher
- Poetry (Python package manager)
A FastAPI application that fetches current news from public news APIs, uses it as context for a local LLM running on Ollama, and generates predictive "future news" based on current trends.
- Periodically fetches current news from NewsAPI
- Stores news data in a local cache
- Integrates with Ollama to use local LLM models
- Generates future news predictions based on current news context
- REST API for retrieving news and generating predictions
- Streaming generation support
- Configurable news sources, categories, and update frequency
- Python 3.10 - 3.13
- Poetry for dependency management
- An API key from NewsAPI
- Ollama running locally with models installed
The application is configured to work well with the following models on an NVIDIA 3090 GPU:
llama3(default) - Good balance of quality and performancemistral- Great for general text generationphi3- Efficient model that works well with limited contextmixtral:8x7b- If you want to use a mixture-of-experts modelllama3:70b-q4- Quantized version of larger model for better quality
You can configure the model in the .env file or choose at runtime through the API.
FastAPI application to generate future news based on current news using local LLM.
This project includes built-in CLI commands for managing dependencies:
A mystical web application that reveals news from the future, styled with Masonic themes and symbolism.
- Gaze into the future with customizable time ranges
- Beautifully styled interface with Masonic imagery and symbolism
- Responsive design that works on desktop and mobile devices
- Elegant animations and transitions
- Integration with Ollama for local LLM inference
- Arch Linux (or Arch-based distribution)
- Python 3.13+
- Poetry package manager
- Git
- Ollama (for local LLM inference)
- Clone the repository:
git clone https://github.com/yourusername/news-from-future.git
cd news-from-future