Machine learning for Swift devs who’d rather not Python

What it does

This is an awesome list that catalogs machine learning resources through an Apple-centric lens. The maintainer rounds up Core ML models, conversion utilities, sample repositories, and tutorials—plus a smattering of TensorFlow and Keras references for context. It is essentially a curated table of contents for iOS and macOS developers who want to drop pre-trained models into Xcode projects rather than train networks from scratch in another language.

The interesting bit

The list’s bias is its whole point: it treats Python and JavaScript frameworks as reference material to be bridged, not mastered. The author explicitly wants to “feel the outcome first” by using Apple’s native APIs and pre-trained models, then maybe explore underlying mechanisms later. That is a refreshingly pragmatic stance in a field that usually demands you read a dozen papers before touching a compiler.

Key highlights

• Heavy focus on Core ML, Vision, and ARKit resources, with Swift as the default language.
• Links to pre-trained model zoos, including Awesome-CoreML-Models and the Core ML Store.
• Coverage of conversion pipelines: torch2coreml, tf-coreml, onnx-coreml, and Turi Create.
• Curated post list ranges from WWDC session notes to “Not Hotdog” build-alongs and Taylor Swift detectors.
• Includes practical repos like trainer-mac, which generates a complete Xcode project from a trained model without writing code.

Verdict

Worth bookmarking if you are an iOS or macOS developer looking for a quick on-ramp to machine learning without leaving Xcode. Skip it if you are already comfortable in PyTorch or Python ML pipelines and do not need Apple-specific deployment shortcuts.