RKNPU2

onnx2rknn convert

I) Docker instruction

1. Install Docker

Please install Docker according to the official manual: https://docs.docker.com/install/linux/docker-ce/ubuntu/

2. docker pull

docker pull belotserkovskiyva/rknn_toolkit2_py38:1.4.0

3. Run image

Execute the following command to run the docker image. After running, it will enter the bash environment.

docker run -t -i --privileged -v /dev/bus/usb:/dev/bus/usb <IMAGE_ID> /bin/bash

4. Run demo (ONNX2RKNN convert)

cd /examples/onnx/yolov5/
python3 test.py

II) Git instruction

install the RKNN-Toolkit2 to convert the model. (https://wiki.t-firefly.com/en/ROC-RK3568-PC/usage_npu.html)

git clone https://github.com/rockchip-linux/rknn-toolkit2

1. Install virtualenv、Python3.8 and pip3

sudo apt install virtualenv 
sudo apt-get install python3 python3-dev python3-pip

2. Install dependent libraries

sudo apt-get update
sudo apt-get install libxslt1-dev zlib1g zlib1g-dev libglib2.0-0 libsm6 \
libgl1-mesa-glx libprotobuf-dev gcc

3. Use virtualenv and install Python dependency, such as requirements_cp38-1.4.0.txt

virtualenv -p /usr/bin/python3 venv
source venv/bin/activate
#python3 -m pip install --upgrade pip
cd rknn-toolkit2/
pip3 install -r doc/requirements_cp38-*.txt

3. If conda activate

virtualenv -p /opt/conda/bin/python3 venv
source venv/bin/activate
#python3 -m pip install --upgrade pip
cd rknn-toolkit2/
pip3 install -r doc/requirements_cp38-*.txt

4.Install RKNN-Toolkit2，such as rknn_toolkit2-1.4.0_22dcfef4-cp38-cp38-linux_x86_64.whl

sudo pip3 install packages/rknn_toolkit2*cp38*.whl

5. Check if RKNN-Toolkit2 is installed successfully or not，press key ctrl+d to exit

(venv) user@linuxx:~/rknn-toolkit2$ python3 from rknn.api import RKNN

ONNX2RKNN convert

then use onnx2rknn.py (rknpu2/onnx2rknn.py)

#to run inference see: -> rknpu2/examples/rknn_yolov5_demo/