All these command are expected to be started using CMD with admin priveleges (requirement for cvat-cli)
- Install Python 3.9.0 (later versions will probably work as well)
Do not forget to tick "Add to PATH" / "Add Python to environment variables", if installer will offer you to do so
for Windows x64:
https://www.python.org/ftp/python/3.9.0/python-3.9.0-amd64.exe
- Verify that Python has installed by running version check, it should print "Python 3.9.0" (or other version)
python --version
- Verify that Python package manager (PIP) is installed and working by running PIP version check, it should print something like "pip 20.2.3 from c:\users\gpuvm\appdata\local\programs\python\python39\lib\site-packages\pip (python 3.9)"
pip --version
- Install Ultralytics which is the model we're going to use for our training. Installation may take a while as it weights quite a lot (few gigabytes)
pip install ultralytics==8.2.78
- Now let's make sure that Ultralytics have been installed correctly. Run that command, which will use the default model to find a bus on an image pulled from internet:
yolo predict model=yolov8n.pt source='https://ultralytics.com/images/bus.jpg'
- By default, "yolo" saves predicted results into the same folder that you've currently been in: "runs/detect/predict", open Explorer to navigate to the current folder, then you can double-check that the folders are there and the detection results are fine.
explorer .
7. Install Python dependencies
pip install opencv-python numpy matplotlib shapely onnxruntime==1.18.0
- Install CVAT Command Line Interface
pip install cvat-cli==2.5.0 cvat_sdk==2.5.0
- Verify that CVAT CLI has been installed correctly by running version check, should print the version number to console
cvat-cli --version
- If EyeAuras or YoloEase were running, you'll have to restart them, otherwise they won't be able to find the tools that you've installed