I tried to profile tacotron2 (GitHub - NVIDIA/tacotron2: Tacotron 2 - PyTorch implementation with faster-than-realtime inference) inference by Nsight Systems (latest version).
However, I cannot profile CPU utilization in Docker. By comparison, I can get the information outside of Docker.
The profiled qdrep file is here:
https://www.dropbox.com/s/ha8bobpd8rqn4qh/report1.qdrep?dl=0
Your qdrep shows several errors, the first of which is:
“Collection of IP samples, backtraces, and scheduling data disabled. Ubuntu distribution requires kernel version 4.3.0-0 or later. Detected kernel version is 3.10.0-862.”
What is the OS running inside and outside your Docker?
In a Docker, when a system’s host utilizes a kernel older than v4.3, it is not possible for NVIDIA Nsight Systems to collect sampling data unless both the host and Docker are running a RHEL or CentOS operating system utilizing kernel version 3.10.1-693 or newer. A user override for this will be made available in a future version.