Welcome to Hailo's Developer Zone
Software Downloads, Documentation, Reference code and examplesTake the guided tour
All you need to install and set up Hailo-8™*
*Hailo-15™ documentation will be made available towards the product release
Migrating a model
Machine Learning Frameworks
To run your model on the Hailo device you have to first migrate it to the Hailo environment.
The model build process is now made easy through the support of a wide variety of machine learning frameworks, including: Keras, TensorFlow, Pytorch, TensorFlow-Lite and ONNX.
Bring your own model, either built on If your model is built on one of the supported machine learning frameworks as listed above or built on any other network. In that case, we recommend converting it to ONNX or TF-Lite using the Dataflow Compiler.
You can always use our Hailo Model Zoo to download ready-to-use models.
The Hailo Dataflow Compiler can receive models in 32bit float precision.
The Hailo parser which is the first step of the Dataflow Compiler process will translate the original model into a Hailo internal format.
The second step will run the optimizer which will quantize the weights into a 4/8/16-bit integer format (8-bit as default).
Hailo Model Zoo
The Hailo Model Zoo provides deep learning models for various computer vision tasks. The pre-trained models can be used to create fast prototypes on Hailo devices. Main features include:
- A variety of common and state-of-the-art pre-trained models and tasks in TensorFlow and ONNX
- Model details, including full precision accuracy vs. quantized model accuracy measured on Hailo-8™
- Each model also includes a binary HEF (Hailo Executable Format) file that is fully supported in the Hailo toolchain and Application suite (for registered users only)
- Retraining Dockers (for selected models)
- Ability to go through the whole process – parsing, model optimizer, compilation, inference, accuracy evaluation – either on the pretrained model or on a retrained model
Additionally, the Hailo Model Zoo Github repository provides users with the capability to quickly and easily reproduce the Hailo-8’s published performance on the common models and architectures included in our Model Zoo.Read more
Hailo Dataflow Compiler
The Hailo Dataflow Compiler is responsible for converting the model from the original format all the way to a binary format that can run inference on the Hailo device.
There are three major steps in the compilation process:
- Model Parser - translating the original model into Hailo format
- Model Optimizer - quantizing the model and optimizing the accuracy performance
- Compiler - allocating the needed resources and compiling it into a binary format that can run on the Hailo device.
On top of these mandatory steps in compilation there are two very important tools to help with the process:
- Profiler - profiling the expected model behavior while running on the Hailo hardware and predicting how many hardware resources are required to run the model
- Emulator - emulating the accuracy of the model and supporting in running the optimization process.
After migrating your model to the Hailo environment, you will run inference on the Hailo-8™ device using the different APIs Hailo provides in the HailoRT library.
The most basic ones will be used in order to:
- Load your model
- Send data; and
- Receive data.
The HailoRT includes many more APIs and functions to help you easily integrate your application with the Hailo device.
Use the TAPPAS demo suite to see a set of full application examples by running a single command line or via TAPPAS GUI.
The TAPPAS demonstrates a variety of application examples ranging from classification, detection, and segmentation, all the way up to more complicated pipelines including several models such as LPR and VMS applications.
The TAPPAS code is open and can be viewed to learn how to use the different APIs of the HailoRT for your application.
Other runtime frameworks can be supported and used such as ONNX Runtime, GStreamer and others.
OS IP Stack
Hailo supports PCIe interface as well as Ethernet.
For the PCIe, a driver needs to be installed on your system - either pre-compiled or specifically compiled for your platform.
For the Ethernet interface you have to use the OS IP Stack on your device.
The HailoRT library is your means to communicate with the Hailo device.
From the basic operations like loading models, sending and receiving data, to more complex pipeline management such as scheduling tasks, automatically postprocessing functionality such as NMS support on and off the Hailo device, and more.
The HailoRT supports Python, C and C++ APIs.
To run inference at top performance using Hailo-8™, you have to connect the Hailo device to a board.
Hailo offers both M.2 and mPCIe modules, as well as an M.2 starter kit to help you get started with Hailo.
Hailo AI Software Suite
The Hailo AI Software Suite combines the Compiler and the run time environment into one easy-to-use environment.
This is the most recommended way to get started with Hailo, offering a one-stop-shop for all the Hailo tools in one docker/environment.
Take me to the tutorial
- What’s new in Hailo AI SW Suite 2023_04
- Hailo Integration Tool – sources
- Hailo Integration Tool – Ubuntu package (deb) for arm64
- Hailo Integration Tool – Ubuntu package (deb) for amd64
- Hailo-8™ M.2 User Guide
- Hailo MVTec HALCON plugin-184.108.40.206 armv7a
- Hailo MVTec HALCON plugin-220.127.116.11 x86
- Hailo MVTec HALCON plugin-18.104.22.168 example
- Hailo Integration Tool User Guide
- Hailo Software Suite – Self extractable 2023-04
- Hailo Software Suite – Docker 2023-04
- HailoRT – Docker V4.13.0
- HailoRT Windows Installer v4.13.0
- HailoRT – Ubuntu package (deb) for armhf V4.13.0
- HailoRT – Ubuntu package (deb) for armel V4.13.0
- HailoRT – Ubuntu package (deb) for arm64 V4.13.0
- HailoRT – Ubuntu package (deb) for amd64 V4.13.0
- HailoRT – PCIe driver Ubuntu package (deb) V4.13.0
- HailoRT – Python package (whl) for Python 3.9, x86_64 V4.13.0
- HailoRT – Python package (whl) for Python 3.8, x86_64 V4.13.0
- HailoRT – Python package (whl) for Python 3.9, aarch64 V4.13.0
- HailoRT – Python package (whl) for Python 3.10, x86_64 V4.13.0
- HailoRT – Python package (whl) for Python 3.8, aarch64 V4.13.0
- HailoRT – Python package (whl) for Python 3.10, aarch64 V4.13.0
- Hailo TAPPAS User Guide v3.24.0
- Hailo Model Zoo User Guide v2.7.0
- HailoRT User Guide
- Hailo Dataflow Compiler User Guide
- Hailo Model Zoo – Python package (whl) V2.7.0
- Hailo Suite User Guide
- Hailo Dataflow Compiler – Python package V3.23.0
- TAPPAS – Docker aarch64 V3.24.0
- TAPPAS – Linux installer 3.24.0
- TAPPAS – Docker x86_64 Ubuntu 20 V3.24.0
- TAPPAS – Docker x86_64 Ubuntu 22 V3.24.0
- Hailo-15™ Product Brief
- Hailo-8™ Modules Integration Tool User Guide
- Hailo Software Suite - Release Update
- HailoRT – Log collector tool (whl) V1.0.3
- HailoRT – Log Collector tool (bash) v1.0.3
- HailoRT Log collector tool User Guide v1.0.3
- Hailo-8™ M.2 Key B+M Extended Temperature Datasheet
- 3D Model Files for Hailo-8R™ mPCIe ET Module Configuration MPA
- 3D Model Files for Hailo-8™ M.2 Key-M ET Module Configuration M2B
- 3D Model Files for Hailo-8™ M.2 Key-M ET Module Configuration M2A
- 3D Model Files for Hailo-8™ M.2 Key-A+E ET Module Configuration MEB
- 3D Model Files for Hailo-8™ M.2 Key-B+M ET Module Configuration MBA
- 3D Model Files for Hailo-8™ M.2 Key-A+E ET Module Configuration MEA
- Hailo-8-M.2 Modules Configurations Identification v1.1
- Hailo-8R mPCIe AI Acceleration Module Thermal Considerations v1.0
- Hailo-8R™ mPCIe Extended Temperature Getting Started
- Hailo SW suite User Guide - HTML
- Hailo-8™ IBIS Model
- TAPPAS User Guide
- Hailo-8™ PCIe Reference Design
- Hailo-8 Board Design Guidelines v2.0
- Hailo-8™ M.2 Key M Extended Temperature Datasheet
- Hailo-8R™ mPCIe Extended Temperature Datasheet
- Hailo-8™ M.2 Key A+E Extended Temperature Datasheet
- Hailo-8™ Datasheet Automotive
- 3D Model Files for Hailo-8™ Mini PCIe Module
- HailoRT User Guide - HTML
- Hailo Dataflow Compiler User Guide - HTML
- Hailo-8™ M.2 Key B+M Datasheet
- Hailo-8™ Datasheet
- Hailo-8™ Reference design
- 3D Model Files for Hailo-8™ M.2 A+E -Key Module
- 3D Model Files for Hailo-8™ M.2 B+M-Key Module
- Thermal Simulation Design File for Hailo-8™ M.2 M-Key Module
- 3D Model Files for Hailo-8™ M.2 M-Key Module
- Errata E21001
- Hailo-8™ M.2 Thermal Considerations Application Note
- Hailo-8™ M.2 Key M Datasheet
- Hailo-8R™ mPCIe Datasheet
- Hailo-8™ M.2 Key A+E Datasheet
- Hailo-8R™ mPCIe Extended Temperature Product Brief
- Hailo-8™ M.2 Extended Temperature Product Brief
- Hailo-8™ Product Brief