Export YOLO11 on google coral edge-tpu

Support
, ,
1

I follow this article:

In my docker container I use this version of python libraries:

ultralytics 8.3.40
torch 2.7.0

nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2021 NVIDIA Corporation
Built on Thu_Nov_18_09:45:30_PST_2021
Cuda compilation tools, release 11.5, V11.5.119
Build cuda_11.5.r11.5/compiler.30672275_0

when I run exported model on google coral I have this error:

" Model provided has model identifier '"

My enviroment on runtime:

uname -a
Linux 5.4.0-216-generic x86_64 x86_64 x86_64 GNU/Linux

lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 20.04.6 LTS
Release: 20.04
Codename: focal

Thanks
Stefano

2

Please provide more details on how you’re exporting the model to run on your Raspberry Pi.

3

I use:

Export the model

model.export(format=“edgetpu”)

but I use a Intel device with google coral.

I understood that I have to use yolo11n as a starting model but I have the error on “segmantation error”

4

As stated in this post

Meaning myself and others can help you better, if you can help us better understand your set up and the problem you’re facing. The information you’ve shared so far is not sufficient to provide any meaningful guidance.

You said you were following the Raspberry Pi (RPi) guide for setting up a Google Coral EdgeTPU, but you’re using a Intel device, so not a Raspberry Pi. That guide is specific for RPi, and if that’s not what you’re using, then it’s unlikely to be relevant. Most likely you’ll need to seek out documentation/support for EdgeTPU setup for your device. A good place to start is probably the EdgeTPU repository and maybe take a look at this issue.