correct TF version in saliency example

examples/Saliency/saliency-maps.py

@@ -11,6 +11,7 @@
 
 import tensorpack as tp
 import tensorpack.utils.viz as viz
+from tensorpack.tfutils import get_tf_version_tuple
 
 IMAGE_SIZE = 224
 
@@ -29,7 +30,7 @@ def guided_relu():
     @tf.RegisterGradient("GuidedReLU")
     def GuidedReluGrad(op, grad):
         return tf.where(0. < grad,
-                        gen_nn_ops._relu_grad(grad, op.outputs[0]),
+                        gen_nn_ops.relu_grad(grad, op.outputs[0]),
                         tf.zeros(grad.get_shape()))
 
     g = tf.get_default_graph()
@@ -59,9 +60,9 @@ def inputs(self):
     def build_graph(self, orig_image):
         mean = tf.get_variable('resnet_v1_50/mean_rgb', shape=[3])
         with guided_relu():
-            with slim.arg_scope(resnet_v1.resnet_arg_scope(is_training=False)):
+            with slim.arg_scope(resnet_v1.resnet_arg_scope()):
                 image = tf.expand_dims(orig_image - mean, 0)
-                logits, _ = resnet_v1.resnet_v1_50(image, 1000)
+                logits, _ = resnet_v1.resnet_v1_50(image, 1000, is_training=False)
             saliency_map(logits, orig_image, name="saliency")
 
 
@@ -103,4 +104,5 @@ def run(model_path, image_path):
     if len(sys.argv) != 2:
         tp.logger.error("Usage: {} image.jpg".format(sys.argv[0]))
         sys.exit(1)
+    assert get_tf_version_tuple() >= (1, 7), "requires TF >= 1.7"
     run("resnet_v1_50.ckpt", sys.argv[1])