I just followed the steps on NVIDIA's website (https://developer.nvidia.com/blog/minimizing-dl-inference-latency-with-mig/), trying to accomplish the flower classification demo. Unfortunately, after setting the trintonserver, I tried to set the client, and stuck on the
perf_client: command not found...
Not sure where is the problem, looking for comments.
question from:https://stackoverflow.com/questions/65831871/perf-client-command-not-found-when-i-follow-the-steps-on-nvidias-website