How to Setup Environment for WSL2

Windows Side

1. Install Windows GPU Drivers
   https://www.nvidia.com/Download/index.aspx

2. Install CUDA Toolkit in WSL2
   https://developer.nvidia.com/cuda-11-8-0-download-archive?target_os=Linux&target_arch=x86_64&Distribution=WSL-Ubuntu&target_version=2.0&target_type=deb_local

3. Add .wslconfig file
   Go to %USERPROFILE%\.wslconfig. Create it if it doesn't exist.
   Add the following:

[wsl2]
networkingMode=mirrored

4. Allow inbound connections through the Hyper-V firewall
   https://superuser.com/questions/1717753/how-to-connect-to-windows-subsystem-for-linux-from-another-machine-within-networ Run Powershell with administrator privileges and input the following command:

Set-NetFirewallHyperVVMSetting -Name "{40E0AC32-46A5-438A-A0B2-2B479E8F2E90}" -DefaultInboundAction Allow

Ubuntu Side

1. Install Python 3.10

sudo add-apt-repository ppa:deadsnakes/ppa
sudo apt install python3.10

2. Create .venv

3. Install Dependencies

pip install -r requirements.txt

This will return errors.

4. Install OpenCv

pip install opencv-python

5. Install PyTorch https://pytorch.org/get-started/locally/

6. Install YOLOv8 https://docs.ultralytics.com/quickstart/

pip install ultralytics