TalkEdit/QUICK-FIX.md

🚨 Quick Fix for PyTorch Compatibility Error

If you're seeing the torch.compiler.disable error, here's how to fix it:

Immediate Fix

# Stop the current container
docker-compose down

# Remove the old image to force rebuild with fixed versions
docker rmi $(docker images | grep videotranscriber | awk '{print $3}')

# Rebuild with fixed dependencies
docker-compose up -d --build

Better Solution: Use Prebuilt Images

Once available, use the prebuilt images instead:

# Stop current container
docker-compose down

# Use prebuilt image (no build required)
docker-compose -f docker-compose.prebuilt.yml up -d

What Was Fixed

1. Version Pinning: Updated requirements.txt with compatible versions:

   • torch==2.0.1 (was >=1.7.0)
   • pytorch-lightning==2.0.6 (compatible with torch 2.0.1)
   • pyannote.audio==3.1.1 (updated to compatible version)

2. Build Process: Removed duplicate PyTorch installation that could cause conflicts

3. Prebuilt Images: Created GitHub Actions to build reliable, tested images

Verification

After fixing, you should see the Streamlit app load without errors at http://localhost:8501

If Still Having Issues

1. Clear Docker cache:

   docker system prune -a
   

2. Check logs:

   docker-compose logs -f
   

3. Manual rebuild:

   docker build --no-cache -t videotranscriber .