fix the op_mapper bug

x2paddle/op_mapper/dygraph/tf2paddle/tf_op_mapper.py

@@ -263,6 +263,7 @@ class TFOpMapper(OpMapper):
             inputs["shape"] = dims.name
         layer_attrs["dtype"] = string(input_value.dtype)
         layer_attrs["fill_value"] = input_value.value
+        
 
         self.paddle_graph.add_layer(
             "paddle.full",