0
0

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?

Jetson Orin NXとJetPack 6.0でYOLOv8.3.11を動かす

Posted at

Jetson Orin NXをお借りできたので、YOLOv8をインストールしてみました。

  • Ubuntu 22.04.4 LTS (Let's note RZ4)
  • Jetson Orin NX 16GB
  • Ediloca EN600 PRO SSD 512GB PCle 3.0x4 NVMe M.2 2280
  • logicool C270N

JetPack 6.0のインストール

※JetsonにHDMIディスプレイを接続し、Force Recovery Modeで起動します。

Target Hardware:Jetson Orin NX 16GB
SDK VERSION:JetPack 6.0(rev.2)

OEM Configuration:Pre-Config
Username:jetson
Password:jetson
Storage Device:NVMe

OSのFlashが完了したら、JetsonのUbuntuが起動します。ログイン画面が表示されたら、Installを実行します。

YOLOv8.3.11のインストール

参考:
https://docs.ultralytics.com/ja/guides/nvidia-jetson/#install-onnxruntime-gpu

sudo apt install python3-pip
pip3 install ultralytics
reboot

Successfully installed filelock-3.16.1 fsspec-2024.9.0 jinja2-3.1.4 networkx-3.4.1 numpy-2.1.2 opencv-python-4.10.0.84 psutil-6.0.0 py-cpuinfo-9.0.0 seaborn-0.13.2 torch-2.4.1 torchvision-0.19.1 tqdm-4.66.5 typing-extensions-4.12.2 ultralytics-8.3.11 ultralytics-thop-2.0.9

sudo apt-get install libopenmpi-dev libopenblas-base libomp-dev -y
pip3 install --no-deps torch-2.3.0-cp310-cp310-linux_aarch64.whl 

Successfully installed torch-2.3.0

pip3 install torchvision-0.18.0a0+6043bc2-cp310-cp310-linux_aarch64.whl 

Successfully installed torchvision-0.18.0a0+6043bc2

wget https://nvidia.box.com/shared/static/48dtuob7meiw6ebgfsfqakc9vse62sg4.whl -O onnxruntime_gpu-1.18.0-cp310-cp310-linux_aarch64.whl
pip3 install onnxruntime_gpu-1.18.0-cp310-cp310-linux_aarch64.whl 

Successfully installed coloredlogs-15.0.1 flatbuffers-24.3.25 humanfriendly-10.0 onnxruntime-gpu-1.18.0

pip3 install numpy==1.23.5

Successfully installed numpy-1.23.5

sudo apt-get install nvidia-jetpack

参考:
https://forums.developer.nvidia.com/t/orin-pytorch-cuda-issue/262858/11

YOLOv8の動作確認

yolo predict model=yolov8n.pt source="https://ultralytics.com/images/zidane.jpg"

検出結果は「Home/runs/detect/predict」内に保存されます(zidane.jpg)。

yolo task=detect mode=predict model=yolov8n.pt source=0 show=True imgsz=256

Jetson Orin NX 16GB:1 frame:22.4ms

0
0
0

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
0
0

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?