diff --git a/paddleseg/models/sfnet.py b/paddleseg/models/sfnet.py
index 43b34baa72..fe48911719 100644
--- a/paddleseg/models/sfnet.py
+++ b/paddleseg/models/sfnet.py
@@ -80,7 +80,7 @@ def forward(self, x):
         logit_list = [
             F.interpolate(
                 logit,
-                x.shape[2:],
+                paddle.shape(x)[2:],
                 mode='bilinear',
                 align_corners=self.align_corners) for logit in logit_list
         ]
@@ -165,7 +165,7 @@ def forward(self, conv_out):
                 out.append(self.dsn[i](f))
 
         fpn_feature_list.reverse()
-        output_size = fpn_feature_list[0].shape[2:]
+        output_size = paddle.shape(fpn_feature_list[0])[2:]
         fusion_list = [fpn_feature_list[0]]
 
         for i in range(1, len(fpn_feature_list)):
@@ -205,24 +205,25 @@ def __init__(self, inplane, outplane, kernel_size=3):
             padding=1,
             bias_attr=False)
 
-    def flow_warp(self, inputs, flow, size):
-        out_h, out_w = size
-        n, c, h, w = inputs.shape
-        norm = paddle.to_tensor([[[[out_w, out_h]]]]).astype('float32')
-        h = paddle.linspace(-1.0, 1.0, out_h).reshape([-1, 1]).tile([1, out_w])
-        w = paddle.linspace(-1.0, 1.0, out_w).tile([out_h, 1])
-        grid = paddle.concat([paddle.unsqueeze(w, 2),
-                              paddle.unsqueeze(h, 2)], 2)
-        grid = grid.tile([n, 1, 1, 1]).astype('float32')
-        grid = grid + flow.transpose([0, 2, 3, 1]) / norm
-        output = F.grid_sample(inputs, grid)
+    def flow_warp(self, input, flow, size):
+        input_shape = paddle.shape(input)
+        norm = size[::-1].reshape([1, 1, 1, -1])
+        norm.stop_gradient = True
+        h_grid = paddle.linspace(-1.0, 1.0, size[0]).reshape([-1, 1])
+        h_grid = h_grid.tile([size[1]])
+        w_grid = paddle.linspace(-1.0, 1.0, size[1]).reshape([-1, 1])
+        w_grid = w_grid.tile([size[0]]).transpose([1, 0])
+        grid = paddle.concat([w_grid.unsqueeze(2), h_grid.unsqueeze(2)], axis=2)
+        grid.unsqueeze(0).tile([input_shape[0], 1, 1, 1])
+        grid = grid + paddle.transpose(flow, (0, 2, 3, 1)) / norm
+
+        output = F.grid_sample(input, grid)
         return output
 
     def forward(self, x):
         low_feature, h_feature = x
         h_feature_orign = h_feature
-        h, w = low_feature.shape[2:]
-        size = (h, w)
+        size = paddle.shape(low_feature)[2:]
         low_feature = self.down_l(low_feature)
         h_feature = self.down_h(h_feature)
         h_feature = F.interpolate(