Fix RealtimeSTT warmup file and PyTorch CUDA version mismatch

Fixed two build/runtime issues:

1. Windows: Missing warmup_audio.wav file from RealtimeSTT
   - Added RealtimeSTT to collect_data_files() in spec
   - Ensures warmup_audio.wav and other RealtimeSTT data files are bundled
   - Fixes: soundfile.LibsndfileError opening warmup_audio.wav

2. Linux: PyTorch/TorchAudio CUDA version mismatch (12.1 vs 12.4)
   - Added torchaudio>=2.0.0 explicitly to dependencies
   - Ensures torchaudio comes from pytorch-cu121 index (same as torch)
   - Previously RealtimeSTT was pulling torchaudio from PyPI with CUDA 12.4
   - Fixes: RuntimeError about CUDA version mismatch

Both packages now correctly use the pytorch-cu121 index via tool.uv.sources
configuration, ensuring matching CUDA versions.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

local-transcription.spec

@@ -122,7 +122,7 @@ for package in ['fastapi', 'starlette', 'pydantic', 'pydantic_core', 'anyio', 'u
         print(f"  ⚠ Warning: Could not collect {package}: {e}")
 
 # Collect data files for packages that need them
-for package in ['fastapi', 'starlette', 'pydantic', 'uvicorn']:
+for package in ['fastapi', 'starlette', 'pydantic', 'uvicorn', 'RealtimeSTT']:
     try:
         data_files = collect_data_files(package)
         if data_files: