I want to quickly use the trtexec tool to recurrent the jetson orin nx 16G v3.1 version benmarks data in the link (Jetson Benchmarks | NVIDIA Developer).
The method in the following link does not seem to be what I want (inference_results_v3.1/closed/NVIDIA at main · mlcommons/inference_results_v3.1 · GitHub).
Could you please provide me with the corresponding test model and the complete trtexec run command?
I also raised a question here(How to reproduce the official benmarks of jetson orin nx 16G?).
Hi,
The MLPerf needs to be done via the command shared in the above link.
If you want a simpler benchmark, please find our example below:
Thanks.
@AastaLLL
1、I have explained it in the above question and link(How to reproduce the official benmarks of jetson orin nx 16G?).
“I have read this page and have the following questions:
- The model in the csv on the page does not seem to correspond to the model on the Jetson Benchmarks | NVIDIA Developer page;
- I tried to run the program and got the following error:
Please close all other applications and Press Enter to continue…
Setting Jetson orin in max performance mode
Traceback (most recent call last):
File “benchmark.py”, line 130, in
main()
File “benchmark.py”, line 28, in main
system_check.run_set_clocks_withDVFS()
File “/home/jetson/Desktop/benchmark/jetson_benchmarks/utils/utilities.py”, line 47, in run_set_clocks_withDVFS
self.set_clocks_withDVFS(frequency=self.gpu_freq, device=‘gpu’)
File “/home/jetson/Desktop/benchmark/jetson_benchmarks/utils/utilities.py”, line 77, in set_clocks_withDVFS
self.set_frequency(device=device, enable_register=self.enable_register, freq_register=self.freq_register, frequency=frequency, from_freq=from_freq)
File “/home/jetson/Desktop/benchmark/jetson_benchmarks/utils/utilities.py”, line 88, in set_frequency
self.write_internal_register(freq_register1, frequency)
File “/home/jetson/Desktop/benchmark/jetson_benchmarks/utils/utilities.py”, line 108, in write_internal_register
reg_write = open(register, “w”)
FileNotFoundError: [Errno 2] No such file or directory: ‘/sys/kernel/debug/bpmp/debug/clk/nafll_gpc1/rate’”
Hi,
If you want to reproduce the MLPerf result, please follow the steps to set it up.
It is a container environment and the models will be downloaded automatically when running make download_data
.
Thanks.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.