Questions

Docs · Questions

StreamDiffusionTDInstallation

How do I verify or repair my StreamDiffusionTD installation?

Installer CLI Commands (v0.3.1+)

v0.3.1 includes a built-in installer CLI with three commands for managing your installation. Activate your venv from the Install page first.

verify

Runs a 13-point environment check:

cd StreamDiffusion-installer
python -m sd_installer verify

Checks include: Python version, CUDA availability, torch installation, TensorRT packages, numpy/protobuf/onnx pins, diffusers fork version, mediapipe, peft, and more. Each check reports pass/fail with details.

repair

Auto-fixes the most common dependency issues:

python -m sd_installer repair

Handles: onnx version pin, diffusers fork reinstallation, protobuf downgrade, numpy pin, mediapipe/opencv conflicts, and TensorRT sub-package installation. Run this after any failed install or if verify reports problems.

diagnose

Generates a full system info dump for bug reports:

python -m sd_installer diagnose

Outputs your GPU info, driver version, Python version, CUDA version, and the version of every relevant package in your venv. Copy-paste the output when posting in Discord #troubleshoot-central.

When to Use Each

  • Just installed? Run verify to confirm everything is correct
  • Something broke? Run repair to auto-fix, then verify to confirm
  • Posting a bug report? Run diagnose and include the output
  • Upgraded from v0.3.0? Run repair to ensure dependencies are correct