Such examples are needed
Any chance the DFC will be available on Apple M series in the future?
3 Likes
You can find such examples in our tutorials. The tutorials are available both in our documentation and with the hailo CLI tool hailo tutorials
1 Like
is there any information about when would LLM be executable?
I am also interested. M1 Max and higher/newer M-chips are powerful enough to train models.
1 Like
Which version of DFC should we be using for use with the RPI5 Hailo-8L AI KIT ?
If there a reference to the planned version support for Hailo-8L from Raspberry Pi ?
1 Like
@giladn Hello, I am trying to convert custom yolo trained model to .hef. I am able to create quantize .har but unable to compile to .hef.
Is there a guide for converting an ONNX model to .hef using a cloud VM?
This is really important for a lot of us Hailo AI Kit users who don’t have an x86 computer! I ran into a variety of problems trying to use a Jupyter notebook running on a cloud PC to do the conversion. Please advise! Thank you!!
1 Like
Error creating hef file after succeseful quantization- mapping fail - General - Hailo Community
This could help you
1 Like
Thank you I was able to compile using hailomz in docker env.
Could you explain the difference between BYOD and BYOM.
What is the data that I bring, and what is the model that I bring?
BYOD (Bring Your Own Data) allows you to fine-tune or retrain existing models from the Hailo Model Zoo, such as YOLO or ResNet, using your custom datasets to adapt them for specific applications. On the other hand, BYOM (Bring Your Own Model) enables you to bring your pre-trained or custom-designed models (e.g., TensorFlow, PyTorch, or ONNX) and deploy them on Hailo’s hardware, like the Hailo-8 or Hailo-15, using the Hailo toolchain. This flexibility lets you leverage Hailo’s optimized AI acceleration while tailoring the solution to your unique requirements.
1 Like