Wednesday, July 10, 2024

Raspberry Pi AI Kit update: Dataflow Compiler now available

Our recent release of the Raspberry Pi AI Kit got quite a bit of attention from the community. At launch we provided a number of computer vision-based AI demos and examples, built on well known state-of-the-art neural network models. However, our power users quickly asked for more — in particular the ability to re-train these models with their own datasets, or even to compile custom models to run on the Hailo AI accelerator. Hailo has been working hard behind the scenes, and we are excited to announce the release of the Hailo Dataflow Compiler (DFC). The DFC will allow our users to extend the ability of the Raspberry Pi AI Kit and fine-tune its performance for their specific use cases.

The image shows a Raspberry Pi 5 with an attached Raspberry Pi M.2 HAT+ board. The Raspberry Pi 5 is the base component, identifiable by its HDMI ports, USB ports, and Ethernet port visible at the bottom right. The M.2 HAT+ board is mounted on top of the Raspberry Pi using four standoffs, which elevate it above the main board. The M.2 HAT+ board has an M.2 module installed, which is secured in place and connected to the HAT+ board. The setup appears to be compact and well-organized, with the M.2 module's connector edge visible and fitted into the HAT+ board. The ribbon cable is connected to the HAT+ board, indicating that it might be used for additional connectivity or power. This configuration is used to enhance the capabilities of the Raspberry Pi 5 by adding support for M.2 devices, which could include high-speed storage solutions or other peripherals, thus expanding the functionality and performance of the Raspberry Pi system.
Everything is held together securely, so it’s easy to embed this in other hardware

Bring Your Own Data (BYOD)

Want to make a wildlife camera to detect certain types of animals? Using the DFC in BYOD mode will allow users to take advantage of some of the most popular neural network models, re-trained on their own custom datasets. Hailo has created an end-to-end tutorial outlining how to re-train an existing neural network model.

Bring Your Own Model (BYOM)

If our existing demos and the neural network models available in Hailo’s model zoo don’t do what you want, you can use the DFC to convert and compile models from ONNX or TensorFlow Lite (TFLite) to Hailo’s HEF format for running on the Hailo AI accelerator. Not for the faint of heart, BYOM requires a deep understanding of the model and the conversion flow — but some will see this as an interesting challenge. Take a look at the DFC tutorials in Hailo’s developer zone.

What’s next?

Users have also asked about Whisper, Stable Diffusion, and so on running on the Hailo AI accelerator.  These very large network models cannot yet run, but Hailo is working hard to port some of them.

The image shows two components commonly used in conjunction with Raspberry Pi devices. The larger component on the top is a Raspberry Pi M.2 HAT+ (Hardware Attached on Top) board. It features an M.2 connector for PCI Express devices and has mounting holes labeled for different M.2 sizes (2230 and 2242). There is also a ribbon cable connector and other electronic components on the board. The smaller component at the bottom is an M.2 module, likely an NVMe SSD or a different type of PCIe device. It has a metallic shield over the main chip and a gold connector edge designed to fit into the M.2 slot on the HAT+ board. The component also shows a few smaller electronic components and traces on its PCB (Printed Circuit Board). These components are used together to expand the functionality of a Raspberry Pi by adding high-speed storage or other peripherals through the M.2 interface.

Also coming soon is Python/Picamera2 integration with the Raspberry Pi AI Kit. We intend to make full support for Python and Picamera2, including demos and examples, available in our next package release.

The post Raspberry Pi AI Kit update: Dataflow Compiler now available appeared first on Raspberry Pi.



from News - Raspberry Pi https://ift.tt/YE3lfRX

Labels: ,

0 Comments:

Post a Comment

Subscribe to Post Comments [Atom]

<< Home