Tutorials

• Quick Start — Llama-v3

  ---

  A minimal end-to-end quantization run on a real LLM.

  Quick start

• Getting Started

  ---

  Quantize an LLM from scratch — the concepts, step by step.

  Getting started

• Save & Load

  ---

  Persist and restore quantization state across runs.

  Save & Load Tutorial

• Autoquant

  ---

  Convert PyTorch models to quantized ones automatically.

  Autoquant Tutorial

• MPath

  ---

  Select submodules and quantizers with filesystem-like queries.

  MPath Tutorial

• Export

  ---

  Turn a quantized model into a deployment artifact.

  Export Tutorial

• Autoquant Qwen

  ---

  Automatically convert a complex HuggingFace model into a quantized one.

  Autoquantize Qwen