This repository provides a collection of Rust examples and templates for MLOps, aiming to explore workflows outside the traditional Python-centric stack (Jupyter, Conda, Pandas, NumPy, Scikit-learn). It targets engineers and developers interested in leveraging Rust's performance, efficiency, and safety for machine learning operations.

How It Works

The project functions as a "cookbook" demonstrating various MLOps tasks using Rust. It showcases Rust's capabilities in areas like data processing (with Polars), web services (Actix), command-line tools, Hugging Face integrations, and even GPU acceleration with PyTorch bindings. The core approach emphasizes building robust, performant, and memory-efficient ML systems without relying on Python's ecosystem.

Quick Start & Requirements

• Installation: Clone the repository and use make all for local setup, which includes formatting, linting, and testing. Rust and Cargo installation is recommended via the official Rust install guide.
• Prerequisites: Rust toolchain (rustc, cargo, rustfmt, clippy, rustup). Some examples may require specific libraries or GPU support (CUDA).
• Resources: GitHub Codespaces are pre-configured. Local setup involves standard Rust toolchain installation.
• Links: Official Install Guide for Rust, Hugging Face Candle.

Highlighted Details

• Demonstrates building CLI tools for tasks like deduplication and data analysis.
• Includes examples of microservices using Actix for API endpoints.
• Showcases integration with Hugging Face models for NLP tasks and translation.
• Features examples of GPU-accelerated computations with PyTorch bindings and Rayon for parallelism.

Maintenance & Community

The repository is maintained by nogibjj. Further community engagement and learning resources are available through linked Coursera courses and other GitHub projects by the author.

Licensing & Compatibility

The repository's licensing is not explicitly stated in the README. Compatibility for commercial use or closed-source linking would require clarification on the specific licenses of the included examples and dependencies.

Limitations & Caveats

This repository is described as a "work in progress," with many examples marked as "aspirational demos" or "hopefully almost every day/weekly" to be solved. Some advanced features, like PyTorch binary embedding, are still under investigation.