SKaiNET

SKaiNET is a Kotlin Multiplatform tensor / compile / graph engine. It provides a tensor DSL, execution contexts, a graph IR, model loaders (GGUF, SafeTensors, ONNX), quantization primitives (Q4_K, Q8_0, ternary, TurboQuant), a StableHLO emitter for cross- platform compile targets, and a pluggable backend API that CPU, GPU, and NPU backends can implement independently.

This documentation site is organized following the Diátaxis / Divio framework — pick the section that matches what you’re trying to do:

  • Tutorials

    Learning-oriented walkthroughs. Start here if you are new to SKaiNET — set up a project, build your first tensor, run a forward pass.

  • How-to guides

    Task-oriented recipes. Build from source, load a GGUF model, generate C for Arduino, train a small network from Java.

  • Reference

    Information-oriented lookup: architecture, the operator catalog, the backend coverage matrix, and the Dokka API reference.

  • Explanation

    Understanding-oriented background: SKaiNET for AI/ML, the operator documentation system, mathematical theory, performance notes.

LLM-specific runtimes (Llama, Gemma, Qwen, BERT) live in the sibling SKaiNET-transformers repository and its own documentation site. Mainline SKaiNET covers the engine layer only.