While converting an .etlt model to a .engine file with tao-converter, my hope was to take the resulting engine file and use the IEngineInspector API that is used in JSON form as an input for TREx in order to visualize the network. However, the output has to correspond to profilingVerbosity=kDETAILED, which is a parameter you can set in trtexec while generating an engine file. It seems that even when profiling the tao-converter output with profilingVerbosity=kDETAILED, because the .engine file was not generated in trtexec with that parameter. Is there any way to enable this with tao-converter?
Tao-converter engine output with profilingVerbosity=kDETAILED
There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks
Currently, the tao-converter does not support this parameter.
For engine inspect , you can find info via below way.
$ python -m pip install colored
$ python -m pip install polygraphy --index-url [https://pypi.ngc.nvidia.com]
$ polygraphy inspect model your.engine
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.