You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Windows is supported (Linux is the primary target, but Windows works)
GPU strongly recommended for training
2. Install PyTorch first
Go to https://pytorch.org/get-started/locally/ and install the build that matches your hardware (cuda for NVIDIA, cpu otherwise). Install PyTorch before nnU-Net.
3. Install nnU-Net
pip install nnunetv2
For a local editable install (development):
git clone https://github.com/MIC-DKFZ/nnUNet.git
cd nnUNet
pip install -e .
Activity
FabianIsensee commentedon Apr 8, 2026
@claude
claude commentedon Apr 8, 2026
🤖 nnU-Net Assistant
Hi @FabianIsensee! Here are the installation instructions for Windows, pulled from
documentation/getting-started/installation-and-setup.md:1. Requirements
2. Install PyTorch first
Go to https://pytorch.org/get-started/locally/ and install the build that matches your hardware (
cudafor NVIDIA,cpuotherwise). Install PyTorch before nnU-Net.3. Install nnU-Net
For a local editable install (development):
4. Create the three storage folders
5. Set environment variables
PowerShell:
Command Prompt:
6. Verify
7. Optional extras
Once set up, follow the Getting Started guide for your first dataset run.