Hi,
I am trying to follow the instructions outlined here Auto-tuning a Convolutional Network for ARM CPU — tvm 0.9.dev182+ge718f5a8a documentation to autotune CNN on my android mobile phone (Samsung)
I have the following questions.
-
I am currently building tvm on Amazon AWS Ubuntu machine by "ssh"ing into it from my Windows 11 laptop. Is it even possible to run rpc_tracker on Ubuntu that would connect to my Android phone? The Android phone is connected via USB cable to my Windows laptop.
-
To add more complexity, I am building tvm compiler on the Ubuntu system on a docker as I don’t have sudo permissions. I really don’t know if rpc_tracker works with dockers. I did use "docker -p 127.0.0.1:8080 " to keep the container visible for connection.
I really don’t know if building tvm on Ubuntu that is “ssh” ed from my Windows and connecting my Android phone to Windows is going to work. So far I was not successful in registering my Android device to rpc_tracker.
If not what are the other tvm build options where I could make rpc_tracker work with my host and target?
Is it -
- Build tvm on my Windows laptop? The instructions in Auto-tuning a Convolutional Network for ARM CPU — tvm 0.9.dev182+ge718f5a8a documentation seem to be tailored to build tvm on Linux. OR
- Use Windows Subsystem for Linux (WSL) on my Windows 11 machine to build tvm? OR
- Is there any other feasible solution with my current build setup?
Thanks in advance for your kind suggestions,