Timeseries models crash at the end (Jetpack 6.2)

Robotics & Edge Computing Jetson & Embedded Systems Jetson Orin Nano
,
1

Test setup:
Model: NVIDIA Jetson Orin Nano Engineering Reference Developer Kit Super - Jetpack 6.2 [L4T 36.4.3]
NV Power Mode[2]: MAXN_SUPER
Serial Number: [XXX Show with: jetson_release -s XXX]
Hardware:

  • P-Number: p3767-0005
  • Module: NVIDIA Jetson Orin Nano (Developer kit)
    Platform:
  • Distribution: Ubuntu 22.04 Jammy Jellyfish
  • Release: 5.15.148-tegra
    jtop:
  • Version: 4.3.1
  • Service: Active
    Libraries:
  • CUDA: 12.6.68
  • cuDNN: 1.0
  • TensorRT: 10.3.0.30
  • VPI: 3.2.4
  • Vulkan: 1.3.204
  • OpenCV: 4.8.0 - with CUDA: NO

Test environment:
git clone GitHub - dusty-nv/pytorch-timeseries
docker pull nvcr.io/nvidia/l4t-ml:r36.2.0-py3
docker/run.sh -c nvcr.io/nvidia/l4t-ml:r36.2.0-py3
pip3 install matplotlib
python3 train.py --data data/weather.csv --inputs temperature --outputs temperature --horizon 1
python3 train.py --data data/weather.csv --inputs temperature,humidity,pressure --outputs temperature,humidity,pressure --horizon 1
python3 train.py --data data/solar_power.csv --inputs AMBIENT_TEMPERATURE,IRRADIATION --outputs DC_POWER,AC_POWER
python3 train.py --data data/shuttle.csv --inputs 0,1,2,3,4,5,6,7,8 --outputs class --classification --epochs 100

When running the above examples, they all produce expected results but finish with the coredump:
…
corrupted size vs. prev_size
Aborted (core dumped)

3

Hi,

The GitHub hasn’t been updated for a while.
So it might not work with r36 BSP.

But the sample is open-source.
You can try to solve the issue directly.

Thanks.

1 Like
5

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.