When I run the example as the blog says.https://tvm.apache.org/2020/06/04/tinyml-how-tvm-is-taming-tinyI can’t find directory’device’ in tvm/micro,therefore I am not able to import the moudles as the example says.What should I do to find or make the directory?Thank you.
Mike
Hi Mike,
We’re just finishing a rewrite of the µTVM runtime and as part of that we’ll update the example code from the blog post. You can read more in this forum thread. In the meantime, if you want to reproduce the blog post you’ll need to sync the TVM repo back to something that was current around the time of the blog post, say 89359907d.
Apologies for the difficulties–I’ll update this thread when I publish an update to the blog post code.
Andrew
Hi Andrew，
Thank you for your reply and I have successfully worked out that problem with a repository of TVM posted in Mar. However when I test the *.bin after I flashed the document to NUCLEO-F746ZG,something goes wrong like this:
There are sevral errors,but I think the error which makes the program abort probabily is: monitor thread exiting while process still alive; killing process Is there something wrong with my settings?How should I solve this problem？I would very appreciate it if you could give me some suggestions. Mike
hey @qelk123, sorry I missed this. it looks like maybe the script is requiring you to configure the attached board serial number. This is being passed to hla_serial given to openocd.
I think the root cause error is this line: expected exactly one argument to hl_serial <serial-number>
It’s been a while since I’ve run this demo but you might try examining env-config.json in the microtvm-blogpost-eval repo and see if you’ve specified hla_serial. or, you could edit microtvm-dev-config/openocd.cfg.template and remove the hla_serial line.
let me know if this solves your problem.
Andrew
Hi @areusch， Thank you for your reply again.I have already sucessfully run the demo in your repository few days ago.I deleted several lines in this file and finally connected my board,though the runtime improvement of autotuned result wasn’t so appealing compared with your 47%.Anyway,thank you so much for your generous help. Mike