Can I run tlt-converter in a Linux machine and then use the output in a Windows TensorRT application? Assuming that the Linux and Windows machines have the same exact GPU hardware and CUDA version.
Can I run tlt-converter in a Linux machine and then use the output in a Windows TensorRT application? Assuming that the Linux and Windows machines have the same exact GPU hardware and CUDA version.
It is possible. The TRT version should be the same too.