Update Dockerfile

Dockerfile

@@ -31,5 +31,8 @@ COPY --chown=user:user . /app/
 # Expose port 7860 (Hugging Face requirement)
 EXPOSE 7860
 
-# Start the FastAPI application with multiple workers for production
-CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860", "--workers", "2"]
+# Start the FastAPI application with 2 workers.
+# Each worker loads ~1.5GB of ML models (SenseVoice + BERT + VAD).
+# With 16GB RAM this is safe (~3GB total), and 2 workers provides real
+# concurrency benefit for CPU-bound STT when multiple users transcribe simultaneously.
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860", "--workers", "2", "--timeout-keep-alive", "75"]