🔧 Fix predict endpoint - Use proper fairseq_signals syntax for model inference

server.py

@@ -268,7 +268,12 @@ async def predict_ecg(payload: ECGPayload):
         with torch.no_grad():
             if fairseq_available:
                 print("🚀 Using fairseq_signals for ECG-FM inference")
-                result = pretrained_model(signal)
+                result = pretrained_model(
+                    source=signal,
+                    padding_mask=None,
+                    mask=False,
+                    features_only=True
+                )
             else:
                 print("⚠️  Using fallback PyTorch inference")
                 result = pretrained_model(signal)