If I’ve trained my data to operate on Jetson Xaver NX, is that data transferrable to a different device/module? i.e. Same machine type or even different machine like Orin
Hi,
The TensorRT engine is not portable as it is built according to hardware information.
But the model like .ONNX or .pth should be okay.
Thanks.
If if I used Tensor Flow to train for object detection and then I buy an Orin after training for Xavier NX can I use the same training or do I have to restart training from the beginning?
Hi,
We recommended dGPU for training jobs.
If the model is in a portable format (ex. onnx, pth), you can use it cross different platforms.
Thanks.