Update server.py

server.py

@@ -125,15 +125,31 @@ def health_check():
 #         raise HTTPException(status_code=500, detail=str(e))
 @app.post("/api/v1/inference")
 async def inference(request: AudioRequest) -> AudioResponse:
+    """Run inference with enhanced error handling and logging"""
+    if not INITIALIZATION_STATUS["model_loaded"]:
+        raise HTTPException(
+            status_code=503,
+            detail=f"Model not ready. Status: {INITIALIZATION_STATUS}"
+        )
+        
     try:
-        # Decode audio from base64
-        audio_bytes = base64.b64decode(request.audio_data) 
+        # Log input validation
+        logger.info(f"Received inference request with sample rate: {request.sample_rate}")
+        
+        # Decode audio
+        audio_bytes = base64.b64decode(request.audio_data)
         audio_array = np.load(io.BytesIO(audio_bytes))
+        logger.info(f"Decoded audio array shape: {audio_array.shape}, dtype: {audio_array.dtype}")
         
+        # Validate input format
+        if len(audio_array.shape) != 2:
+            raise ValueError(f"Expected 2D audio array [C,T], got shape {audio_array.shape}")
+            
         # Run inference
         result = model.inference(audio_array, request.sample_rate)
+        logger.info(f"Inference complete. Output shape: {result['audio'].shape}")
         
-        # Encode output audio
+        # Encode output
         buffer = io.BytesIO()
         np.save(buffer, result['audio'])
         audio_b64 = base64.b64encode(buffer.getvalue()).decode()
@@ -142,6 +158,14 @@ async def inference(request: AudioRequest) -> AudioResponse:
             audio_data=audio_b64,
             text=result.get("text", "")
         )
+        
+    except Exception as e:
+        logger.error(f"Inference failed: {str(e)}", exc_info=True)
+        raise HTTPException(
+            status_code=500, 
+            detail=str(e)
+        )
+        )
 if __name__ == "__main__":
     import uvicorn
     uvicorn.run(app, host="0.0.0.0", port=8000)