diff --git a/tensorflow/python/eager/benchmarks/resnet50/hvp_test.py b/tensorflow/python/eager/benchmarks/resnet50/hvp_test.py
index c5282174ffa..fa86a98371d 100644
--- a/tensorflow/python/eager/benchmarks/resnet50/hvp_test.py
+++ b/tensorflow/python/eager/benchmarks/resnet50/hvp_test.py
@@ -35,7 +35,7 @@ def _forward_over_back_hvp(model, images, labels, vector):
       model.trainable_variables, vector) as acc:
     with tf.GradientTape() as grad_tape:
       logits = model(images, training=True)
-      loss = tf.losses.softmax_cross_entropy(
+      loss = tf.compat.v1.losses.softmax_cross_entropy(
           logits=logits, onehot_labels=labels)
     grads = grad_tape.gradient(loss, model.trainable_variables)
   return acc.jvp(grads)
@@ -47,7 +47,7 @@ def _back_over_forward_hvp(model, images, labels, vector):
     with forwardprop.ForwardAccumulator(
         model.trainable_variables, vector) as acc:
       logits = model(images, training=True)
-      loss = tf.losses.softmax_cross_entropy(
+      loss = tf.compat.v1.losses.softmax_cross_entropy(
           logits=logits, onehot_labels=labels)
   return grad_tape.gradient(acc.jvp(loss), model.trainable_variables)
 
@@ -55,7 +55,7 @@ def _back_over_forward_hvp(model, images, labels, vector):
 def _tf_gradients_forward_over_back_hvp(model, images, labels, vector):
   with tf.GradientTape() as grad_tape:
     logits = model(images, training=True)
-    loss = tf.losses.softmax_cross_entropy(
+    loss = tf.compat.v1.losses.softmax_cross_entropy(
         logits=logits, onehot_labels=labels)
   variables = model.trainable_variables
   grads = grad_tape.gradient(loss, variables)
@@ -68,7 +68,7 @@ def _back_over_back_hvp(model, images, labels, vector):
   with tf.GradientTape() as outer_tape:
     with tf.GradientTape() as inner_tape:
       logits = model(images, training=True)
-      loss = tf.losses.softmax_cross_entropy(
+      loss = tf.compat.v1.losses.softmax_cross_entropy(
           logits=logits, onehot_labels=labels)
     grads = inner_tape.gradient(loss, model.trainable_variables)
   return outer_tape.gradient(
diff --git a/tensorflow/python/eager/benchmarks/resnet50/resnet50_graph_test.py b/tensorflow/python/eager/benchmarks/resnet50/resnet50_graph_test.py
index 2520df70b44..958756d418b 100644
--- a/tensorflow/python/eager/benchmarks/resnet50/resnet50_graph_test.py
+++ b/tensorflow/python/eager/benchmarks/resnet50/resnet50_graph_test.py
@@ -109,7 +109,7 @@ class ResNet50Benchmarks(tf.test.Benchmark):
 
         model = resnet50.ResNet50(data_format())
         logits = model(images, training=True)
-        loss = tf.losses.softmax_cross_entropy(
+        loss = tf.compat.v1.losses.softmax_cross_entropy(
             logits=logits, onehot_labels=labels)
         optimizer = tf.train.GradientDescentOptimizer(learning_rate=1.0)
         train_op = optimizer.minimize(loss)
diff --git a/tensorflow/python/eager/benchmarks/resnet50/resnet50_test.py b/tensorflow/python/eager/benchmarks/resnet50/resnet50_test.py
index 8be6ab89766..6e8d7b61669 100644
--- a/tensorflow/python/eager/benchmarks/resnet50/resnet50_test.py
+++ b/tensorflow/python/eager/benchmarks/resnet50/resnet50_test.py
@@ -36,7 +36,7 @@ from tensorflow.python.eager.benchmarks.resnet50 import resnet50_test_util
 def compute_gradients(model, images, labels, num_replicas=1):
   with tf.GradientTape() as grad_tape:
     logits = model(images, training=True)
-    loss = tf.losses.softmax_cross_entropy(
+    loss = tf.compat.v1.losses.softmax_cross_entropy(
         logits=logits, onehot_labels=labels)
     tf.compat.v2.summary.write('loss', loss)
     if num_replicas != 1:
diff --git a/tensorflow/python/eager/benchmarks/resnet50/resnet50_test_util.py b/tensorflow/python/eager/benchmarks/resnet50/resnet50_test_util.py
index dacc6fb6a85..8d3b95053b5 100644
--- a/tensorflow/python/eager/benchmarks/resnet50/resnet50_test_util.py
+++ b/tensorflow/python/eager/benchmarks/resnet50/resnet50_test_util.py
@@ -35,8 +35,8 @@ def random_batch(batch_size, data_format):
   shape = (batch_size,) + shape
 
   num_classes = 1000
-  images = tf.random_uniform(shape)
-  labels = tf.random_uniform(
+  images = tf.random.uniform(shape)
+  labels = tf.random.uniform(
       [batch_size], minval=0, maxval=num_classes, dtype=tf.int32)
   one_hot = tf.one_hot(labels, num_classes)