Update evaluation.py

evaluation.py

@@ -109,7 +109,7 @@ def compute_mse(predictions, labels):
 
     # get size of the golden labels list and of
     # the valid predictions array
-    labels_size = len(labels)
+    labels_size = np.array(labels).size
     valid_predictions_size = idx[0].size
 
     # only compute mse if valid score predictions were generated,