
                    For tensorrt, we pick up the best solutions provided by github life savers.
                    If google brings you here, you are so lucky! These life savers are angels, they have well solved your problems.
                    So do not waste time scrolling the original issue threads just to dig out solutions... 
                    Here you are, you are welcome. Then, go back to work!
                    

Hi @RizhaoCai
I just had the same issue with TensorRT 7.0 The solution as mentioned in the documentaiton 

TensorRT [TensorRT] ERROR: Network must have at least one output

thanks @rmccorm4
the code snippet really helped to debug the issue After solving the [TensorRT] ERROR: Network must have at least one output
Another error happened
The code is:

The output is:


My pytorch version is 1.3.0

TensorRT virtual nvinfer1::ILayer* nvinfer1::Network::getLayer(int) const: Assertion `layerIndex >= 0' failed.

Hi @gcp Sorry for the delay Description
This is due to mutually incompatible changes in the TRT7 release:
https://docs.nvidia.com/deeplearning/sdk/tensorrt-release-notes/tensorrt-7.html

ONNX parser with dynamic shapes support
The ONNX parser supports full-dimensions mode only

TensorRT ONNX networks can't use INT8 calibration and batching

and further more Description
the result from IResizeLayer(scale_factor=scale_factor 

TensorRT bilinear IResizeLayer setting align_corners=False makes different result to torch interpolate

i already fix issue My mistake i should  replace C3 -> bottleneckCSP  not BottleneckCSP  because in commom.hpp define bottleneckCSP! 