Hi,
The Resize Layer was fallbacked to GPU.
I am already updated to Jetpack 4.6 and the version of Tensorrt is 8.0. So, I think the resize layer is already supported.
I use the commandtrtexec --onnx=test.onnx --useDLACore=0 --fp16 --allowGPUFallback to run the smallest onnx model. test.onnx (348 Bytes)
The first two elements in scales need to be exactly 1 (for unchanged batch and channel dimensions).
The last two elements in scales, representing the scale values along height and width dimensions, respectively, need to be integer values in the range of [1, 32].
The layer is fallback to GPU for some internal reason.
But the DLA support is added in our future release.
We have confirmed that the model can work with DLA internally.
Please wait for our announcement for the software update.
Thanks and sorry for the inconvenience.