Allwinner V853 development board builds YOLOV5-V6.0 environment

aleksib

Allwinner V853 development board builds YOLOV5-V6.0 environment [Copy link]

 

# Conda builds YOLOV5-V6.0 environment## 0. Preface This chapter mainly records how to build the yolov5-v6.0 environment and support CUDA acceleration. The hardware requirements of this article are as follows. If your computer does not have a graphics card device, this article cannot be used as a reference for you. **Hardware and software list:** * CPU: Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40GHz 2.40 GHz * Graphics card: RTX 3060 * System: Windows 10 Professional Workstation Edition## 1. Build Python environment Open the Conda terminal, create a Conda environment for Python3.7, and enter ```lua conda create -n py37_yolov5 python=3.7 ``` After the creation is completed, pytorch needs to be installed. ## 2. Install Pytorch and cuda There is no requirement for the installation order of the two, but there are version requirements. In short: the versions of the two need to correspond, a high version and a low version will cause CUDA to be unusable. You need to check the graphics driver version. The way to check is as follows:  If the customer's driver version is too low, you can upgrade it by updating the graphics driver. You can search for the update method on the Internet. Below I installed the CUDA11.3 version to meet the needs of more customers. Paste the copied command into the terminal and enter: ```r conda install pytorch==1.10.0 torchvision==0.11.0 torchaudio==0.10.0 cudatoolkit=11.3 -c pytorch -c conda-forge ``` Test whether the installed pytorch is available. After entering `python` in the Conda terminal, load the torch module and print whether cuda is available. ```python (py37_yolov5) D:\Programmers\ModelDeployment\2.yolov5\yolov5-train\yolov5-6.0\yolov5-6.0>python Python 3.7.16 (default, Jan 17 2023, 16:06:28) [MSC v.1916 64 bit (AMD64)] :: Anaconda, Inc. on win32 Type "help", "copyright", "credits" or "license" for more information. >>> import torch >>> torch.cuda.is_available() True >>> exit() ``` ## 3. Install YOLOV5-V6.0 version dependency package Before installation, check the installed dependency packages and enter `pip list` ```diff (py37_yolov5) D:\Programmers\ModelDeployment\2.yolov5\yolov5-train\yolov5-6.0\yolov5-6.0>pip list Package Version ----------------- -------- certifi 2023.5.7 cycler 0.11.0 fonttools 4.38.0 kiwisolver 1.4.4 numpy 1.21.6 packaging 23.1 Pillow 9.2.0 pip 22.3.1 pyparsing 3.1.0 python-dateutil 2.8.2 setuptools 65.6.3 six 1.16.0 torch 1.10.0 torchaudio 0.10.0 torchvision 0.11.0 typing_extensions 4.6.3 wheel 0.38.4 wincertstore 0.2 ```` 1.8.0 torchvision>=0.8.1 tqdm>=4.41.0 # Logging ------------------------------------- tensorboard>=2.7.1 # wandb # Plotting ---------------------------------------------------- pandas>=1.23.4 seaborn>=0.11.0 # Export -------------------------------------- # coremltools>=4.0 # CoreML export # onnx>=1.9.0 # ONNX export # onnx-simplifier>=0.3.6 # ONNX simplifyer # scikit-learn==0.19.2 # CoreML quantization # tensorflow>=2.4.1 # TFLite export # tensorflowjs>=3.9.0 # TF.js export # Extras ----------------------------------------- # albumentations>=1.0.3 # Cython # for pycocotools https://github.com/cocodataset/cocoapi/issues/ 172 # pycocotools>=2.0 # COCO mAP # roboflow thop # FLOPs computation ``` Here I suggest installing the required dependencies separately in the Conda command, checking for omissions and filling in the gaps, and just add the missing dependency package. ```bash pip install opencv-python==4.7.0.68 -i https://pypi.doubanio.com/simple/ pip install matplotlib==3.5.3 -i https://pypi.doubanio.com/simple/ pip install PyYAML==6.0 -i https://pypi.doubanio.com/simple/ pip install requests==2.28.2 -i https://pypi.doubanio.com/ simple/ pip install tqdm==4.64.1 -i https://pypi.doubanio.com/simple/ pip install tensorboard==2.11.2 -i https://pypi.doubanio.com/simple/ pip install pandas==1.1.5 -i https://pypi.doubanio.com/simple/ pip install seaborn==0.12.2 -i https://pypi.doubanio.com/simple/ p ip install coremltools==4.1 -i https://pypi.doubanio.com/simple/ pip install onnx==1.13.0 -i https://pypi.doubanio.com/simple/ pip install onnxsim==0.4.13 -i https://pypi.doubanio.com/simple/ pip install scikit-learn==0.19.2 -i https://pypi.doubanio.com/simple/ ```` Or use Tsinghua source ```` bash p ip install opencv-python==4.7.0.68 -i https://pypi.tuna.tsinghua.edu.cn/simple pip install matplotlib==3.5.3 -i https://pypi.tuna.tsinghua.edu.cn/simple pip install PyYAML==6.0 -i https://pypi.tuna.tsinghua.edu.cn/simple pip install requests==2.28.2 -i https://pypi.tuna.tsinghua.edu.cn/simple pip install tqdm==4.64.1 -i https://pypi.tuna.tsinghua.edu.cn/simple pip install tensorboard==2.11.2 -i https://pypi.tuna.tsinghua.edu.cn/simple pip install pandas==1.1.5 -i https://pypi.tuna.tsinghua.edu.cn /simple pip install seaborn==0.12.2 -i https://pypi.tuna.tsinghua.edu.cn/simple pip install coremltools==4.1 -i https://pypi.tuna.tsinghua.edu.cn/simple pip install onnx==1.13.0 -i https://pypi.tuna.tsinghua.edu.cn/simple pip install onnxsim==0.4.13 -i 23.1.0 cachetools 5.3.0 certifi 2023.5.7 charset-normalizer 3.1.0 colorama 0.4.6 coremltools 4.1 cycler 0.11.0 fonttools 4.38.0 google-auth 2.21.0 google-auth-oauthlib 0.4.6 grpcio 1.56.0 idna 3.4 importlib-metadata 6.6.0 kiwisolver 2.6.0 ioctl 2.0 scikit-learn 0.19.2 -i https://pypi.tuna.tsinghua.edu.cn/simple ``` After the installation is complete, the list of dependency packages is as follows: ```yaml Package Version ----------------------- -------- absl-py 1.4.0 attr 0.3.2 attrs 23.1.0 cachetools 5.3.0 certifi 2023.5.7 charset-normalizer 3.1.0 colorama 0.4.6 coremltools 4.1 cycler 1.4.4 Markdown 3.4.3 markdown-it-py 2.2.0 MarkupSafe 2.1.2 matplotlib 3.5.3 mdurl 0.1.2 mpmath 1.3.0 numpy 1.19.5 oauthlib 3.2.2 onnx 1.13.0 onnxsim 0.4.13 opencv-python 4.7.0.68 packaging 2 3.1 pandas 1.1.5 Pillow 9.2.0 pip 22.3.1 protobuf 3.20.3 pyasn1 0.5.0 pyasn1-modules 0.3.0 Pygments 2.15.1 pyparsing 3.1.0 python-dateutil 2.8.2 pytz 2023.3 PyYAML 6.0 requests 2.28.2 requests-oauthlib 1.3.1 rich 13.3.5 rsa 4.9 scikit-learn 0.19.2 scipy 1.7.3 seaborn 0.12.2 setuptools 65.6.3 six 1.16.0 sympy 1.10.1 tensorboard 2.11.2 tensorboard-data-server 0.6.1 tensorboard-plugin-wit 1.8.1 torch 1.10.0 torchaudio 0.10.0 torchvision 0.11.0 tqdm 4.64.1 typing_extensions 4.6.3 urllib3 1.26.15 Werkzeug 2.2. 3 wheels 0.38.4 wincertstore 0.2 zipp 3.15.0 ``` ## 4. Test YOLOV5 environment### 4.1 Test detection model After entering the YOLOV5 source code directory, enter `python detect.py`  You can see that the CUDA graphics card device number 0 has been detected. ```yaml YOLOv5 2021-10-12 torch 1.10.0 CUDA:0 (NVIDIA GeForce RTX 3060, 12287.5MB) ``` ### 4.2 Test export model In the terminal, enter ```css python export.py --weights yolov5s.pt --include onnx --dynamic ```  ### 4.3 Test training Model In the terminal, enter ``` python train.py ``` [](https://forums.100ask.net/uploads/default/original/ 2X/4/4a23b06259c7f85883a69f8aab465f93c0e85910.png "image-20230627182900796")3 Test the training model. Enter ``` python train.py ``` in the terminal. [](https://forums.100ask.net/uploads/default/original/2X/4/4a23b06259c7f85883a69f8aab465f93c0e85910.png "image-20230627182900796")3 Test the training model. Enter ``` python train.py ``` in the terminal. [](https://forums.100ask.net/uploads/default/original/2X/4/4a23b06259c7f85883a69f8aab465f93c0e85910.png "image-20230627182900796")

humancat01

I strongly support domestic chips

lugl4313820

• CPU: Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40GHz 2.40 GHz
• The boss is using a server, awesome!