Micro Intelligence for Near-term Inference and Models
MINIM is a lightweight, high-performance command-line tool written in Rust for performing polynomial regression (degree 2) on datasets. It enables quick trend analysis and data fitting using gradient descent, without the overhead of massive AI frameworks.
- Efficient: Written in Rust for speed and memory safety.
- Minimalist: Zero heavy dependencies; perfect for lightweight environments or edge devices.
- Gradient Descent: Implements custom learning rate adjustment for stable convergence.
- Interactive CLI: Easily train models, manually adjust parameters, and perform real-time inference.
- Rust installed on your machine.
- Clone the repository.
- Build the project:
cargo build --release