In NVIDIA TensorRT Quick Start Guide | NVIDIA Docs, I found file produced by trtexec with different extension, for example ,
trtexec --onnx=resnet50_onnx_model.onnx --saveEngine=resnet_engine.trt
file extension is .trt
trtexec --onnx=fcn-resnet101.onnx --fp16 --workspace=64 --minShapes=input:1x3x256x256
–optShapes=input:1x3x1026x1282 --maxShapes=input:1x3x1440x2560 --buildOnly –
file extension here is .engine
So file extension of output of trtexec can be arbitrary? is it right?
Request you to share the ONNX model and the script if not shared already so that we can assist you better.
Alongside you can try few things:
- validating your model with the below snippet
filename = yourONNXmodel
model = onnx.load(filename)
2) Try running your model with trtexec command.
In case you are still facing issue, request you to share the trtexec “”–verbose"" log for further debugging
My question is that if tensorrt engine file can have different extension, *.trt, *.engine are both OK?
i.e. model.trt == model.engine?
They both are the same. We can give either of the types. TensorRT cares more about the content of the file.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.