The Hailo 8 is an inference engine. AI model training is done separately then converted and written to the Hailo 8. This blog seeks to organize some of the many resources relevant to the Hailo 8.
The process for using the Hailo 8 in an application is:
- Use a pre-trained model or train an AI model
- Convert and write it to the Hailo 8
HAILO 8
- Hailo 8 website, https://hailo.ai/products/ai-accelerators/hailo-8-ai-accelerator/
- Hailo 8 community, https://community.hailo.ai/tag/hailo8
- Youtube videos by Hailo, https://www.youtube.com/@hailo2062
HAILO 8 SOFTWARE TOOLS
- Hailo Model Zoo: A repository of pre-trained models optimized for Hailo devices.
- Hailo Model Explorer, https://hailo.ai/products/hailo-software/model-explorer/
- Github Hailo Model Zoo, https://github.com/hailo-ai/hailo_model_zoo
- HailoRT: The runtime environment for executing models on Hailo hardware, https://github.com/hailo-ai/hailort.
- TAPPAS: Template Applications and Solutions that demonstrate real-world use cases and integrations, https://github.com/hailo-ai/tappas.
MODELS & TRAINING (A MODEL)
For information about this, a simple prompt such as "tell me about training an AI model for use with a Hailo 8" in ChatGPT, Claude or Gemini will give you the start to understanding the process.
- ONNX, the open standard for AI model exchange. The Hailo 8 compiler accepts ONNX models. https://onnx.ai/index.html
- Tensorflow, a well establish machine learning system, https://www.tensorflow.org/.
- Pytorch, an open-source machine learning framework that allows developers to build, train, and deploy deep learning models using a flexible and intuitive Python interface. https://pytorch.org/
OTHER
- Raspberry Pi AI pages