bhetherman/live-voice-chat
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Switch to CUDA 12.8 + ONNX-based VAD for RTX 5090 Blackwell support

Browse Source 
Upgrade PyTorch to 2.7+ with cu128 wheels for Blackwell (sm_120) GPU
support. Replace silero-vad (which depends on torchaudio) with a direct
ONNX Runtime implementation of the same Silero VAD model, eliminating
the torchaudio dependency entirely.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Brian
2026-04-08 04:01:27 -04:00
parent 0305f1dccd
commit 263f39e0a3
4 changed files with 77 additions and 45 deletions
Download Patch File Download Diff File Expand all files Collapse all files
requirements.txt
+3 -2
View File
@@ -1,9 +1,10 @@
# torch and auto-gptq are installed in the Dockerfile with GPU-specific index URLs.
# For local dev outside Docker: pip install torch --index-url https://download.pytorch.org/whl/cu121
# For local dev outside Docker: pip install torch --index-url https://download.pytorch.org/whl/cu128
transformers==4.57.6
optimum>=1.19
compressed-tensors>=0.5.0
silero-vad>=5.1
onnxruntime>=1.17.0
huggingface-hub>=0.20.0
qwen-asr==0.0.6
kokoro==0.9.4
fastapi>=0.115.0