How to upload model from Roboflow to LL?

Title is pretty self-explanatory. I have a model on RF and want to get it onto the LL. I saw in the docs that this is possible - but nothing explaining how.

Any help will be greatly appreciated. :slight_smile:

You have to train it to a tflite file, the easiest way is probably to train using yolov8 and then convert that trained model to a tflite. Then pretty sure you have to compile it for edge tpu, which is what the google coral is. I havnt had any luck personally though, limelight eventually will train their own model so if it’s just notes you could use theirs

Wow that sounds complicated. Will probably wait, or upload to a coprocessor in an easier way.

Thanks for your help!

I have an update for you. Folks at limelight made some google collab notebooks it looks like, making it so any team can train these models with relative ease. You can take a look here: Training a Custom Detector Model | Limelight Documentation

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.