imageNet issue

[tengumanc]

Hello, I’m pretty new here so thank you for your help and patience.

I have been trying to run imageNet from the Jetson-Inference Library.

I am using a Jetson Orin Nano Developer kit. I downloaded Jetpack 6.1 to a micro-SD card then updated the image. I followed the Build procedure and tried to run imagenet but get the red error below.

[TRT]    TensorRT 10.3 does not support legacy caffe models
[TRT]    device GPU, failed to load networks/Googlenet/bvlc_googlenet.caffemodel
[TRT]    failed to load networks/Googlenet/bvlc_googlenet.caffemodel
[TRT]    imageNet -- failed to initialize.
imagenet:  failed to initialize imageNet

I’m not sure why this is the case, I have tried multiple SD cards and sometimes it works and sometimes it doesn’t. I’m trying to replicate this issue and I must be doing something different with the setup process.

Is TensorRT unsopported with Jetpack 6.1 or is this error for some other reason?

Thank you for your help.

[AastaLLL]

Hi,

Since we remove caffemodel support in TensorRT 10, please setup your device with JetPack 6.0 GA for jetson-inference.

If JetPack 6.1 is preferred, please check our Jetson AI Lab instead:

Introduction - NVIDIA Jetson AI Lab

Thanks.

[jmakurami909]

Thank you for these insights. As the imagenet image classification is somewhat of a historically foundational model, is there a way to make it work with 6.1? I am also getting the error above when following the tutorial.
Is there a way to use the older version of TensorRT or would it just be easier to flash a JetPack 6.0 GA microsd and proceed?

https://github.com/dusty-nv/jetson-inference/blob/master/docs/imagenet-console-2.md