Hi, I’m the Toradex AI Assistant. The Toradex Support team will follow up with you shortly!
Did you like my answer?
Hello,
It seems you’re having trouble accessing the NXP artifactory for a library needed to use the NPU on your Toradex device. While the specific link you mentioned is currently unavailable, you might consider using the NXP eIQ software as an alternative.
NXP eIQ offers a comprehensive suite for machine learning applications optimized for i.MX SoCs, including support for Neural Network acceleration on the GPU or NPU through the OpenVX backend. This could be a suitable alternative for running your YOLO model.
To integrate eIQ with your project, you can use the Yocto Project to build a custom image that includes the necessary AI runtimes. Here’s a brief overview of the steps:
-
Clone the Toradex BSP Repository: Set up your environment by cloning the Toradex BSP repository to provide the base for your custom image.
-
Add eIQ Recipes: Include the eIQ recipes from the meta-imx layer in your Yocto build to add the necessary machine learning libraries and tools to your image.
-
Configure Your Build: Adjust your build configuration to target your specific Toradex module, such as the Verdin iMX8MP, ensuring all dependencies are correctly set up.
-
Build the Image: Compile the image using BitBake, integrating eIQ and any other required components into your custom Linux image.
-
Deploy and Test: Flash the built image onto your Toradex device and test your YOLO model using the NPU.
For Torizon users, you can also leverage Docker containers to deploy machine learning applications. Toradex provides Debian-based containers that can be customized to include eIQ and other necessary libraries.
By following these steps, you should be able to continue your project without relying on the unavailable NXP artifactory link.
Best regards,
Toradex AI Assistant.
Please, also check if these links can help you solve your problem: