Introduction

Introduction
Status and Features
Setup and Quick Start
Frequently asked questions
How to Contribute
Release Notes
Citation, Acknowledgments, and Contributors

User Guide

Concepts
Configuration
Command Line Interface (deprecated)
Saving/Loading hls4ml models

Frontends

Keras and its quantized variants
PyTorch and Brevitas
ONNX and QONNX

Backends

Vivado/Vitis
VivadoAccelerator
Coyote
VitisUnified
oneAPI
Catapult
Quartus
SymbolicExpression
XLS

Advanced Features

Profiling
Automatic precision inference
High Granularity Quantization (HGQ2)
Distributed Arithmetic
Model-wise Precision Inference
FIFO Buffer Depth Optimization
Extension API
Spiking Neural Networks (PyTorch/SNN)
Hardware-aware Optimization API
Loading weights from external BRAM
External Backend and Writer Plugins

Internals

Internal representation
ModelGraph Class
MultiModelGraph Class
Optimizer Passes and Flows
Layer attributes

Autogenerated API Reference

hls4ml.backends package
hls4ml.converters package
hls4ml.model package
hls4ml.optimization package
hls4ml.report package
hls4ml.utils package
hls4ml.writer package

hls4ml

Catapult
Edit on GitHub

Catapult

Support for the Siemens Catapult HLS compiler was added in hls4ml version 1.0.0.

TODO expand this section

Previous Next

© Copyright 2026, Fast Machine Learning Lab.

Built with Sphinx using a theme provided by Read the Docs.