diff --git a/site/en/tutorials/images/transfer_learning_with_hub.ipynb b/site/en/tutorials/images/transfer_learning_with_hub.ipynb
index 58ebfc2d20..0d5e04fc9b 100644
--- a/site/en/tutorials/images/transfer_learning_with_hub.ipynb
+++ b/site/en/tutorials/images/transfer_learning_with_hub.ipynb
@@ -90,6 +90,7 @@
       },
       "outputs": [],
       "source": [
+        "pip install tf-keras\n",
         "import numpy as np\n",
         "import time\n",
         "\n",
@@ -97,6 +98,7 @@
         "import matplotlib.pylab as plt\n",
         "\n",
         "import tensorflow as tf\n",
+        "import tf_keras as keras\n",
         "import tensorflow_hub as hub\n",
         "\n",
         "import datetime\n",
@@ -150,7 +152,7 @@
       "source": [
         "IMAGE_SHAPE = (224, 224)\n",
         "\n",
-        "classifier = tf.keras.Sequential([\n",
+        "classifier = keras.Sequential([\n",
         "    hub.KerasLayer(classifier_model, input_shape=IMAGE_SHAPE+(3,))\n",
         "])"
       ]
@@ -181,7 +183,7 @@
       },
       "outputs": [],
       "source": [
-        "grace_hopper = tf.keras.utils.get_file('image.jpg','https://storage.googleapis.com/download.tensorflow.org/example_images/grace_hopper.jpg')\n",
+        "grace_hopper = keras.utils.get_file('image.jpg','https://storage.googleapis.com/download.tensorflow.org/example_images/grace_hopper.jpg')\n",
         "grace_hopper = Image.open(grace_hopper).resize(IMAGE_SHAPE)\n",
         "grace_hopper"
       ]
@@ -261,7 +263,7 @@
       },
       "outputs": [],
       "source": [
-        "labels_path = tf.keras.utils.get_file('ImageNetLabels.txt','https://storage.googleapis.com/download.tensorflow.org/data/ImageNetLabels.txt')\n",
+        "labels_path = keras.utils.get_file('ImageNetLabels.txt','https://storage.googleapis.com/download.tensorflow.org/data/ImageNetLabels.txt')\n",
         "imagenet_labels = np.array(open(labels_path).read().splitlines())"
       ]
     },
@@ -323,7 +325,7 @@
       "source": [
         "import pathlib\n",
         "\n",
-        "data_file = tf.keras.utils.get_file(\n",
+        "data_file = keras.utils.get_file(\n",
         "  'flower_photos.tgz',\n",
         "  'https://storage.googleapis.com/download.tensorflow.org/example_images/flower_photos.tgz',\n",
         "  cache_dir='.',\n",
@@ -353,7 +355,7 @@
         "img_height = 224\n",
         "img_width = 224\n",
         "\n",
-        "train_ds = tf.keras.utils.image_dataset_from_directory(\n",
+        "train_ds = keras.utils.image_dataset_from_directory(\n",
         "  str(data_root),\n",
         "  validation_split=0.2,\n",
         "  subset=\"training\",\n",
@@ -362,7 +364,7 @@
         "  batch_size=batch_size\n",
         ")\n",
         "\n",
-        "val_ds = tf.keras.utils.image_dataset_from_directory(\n",
+        "val_ds = keras.utils.image_dataset_from_directory(\n",
         "  str(data_root),\n",
         "  validation_split=0.2,\n",
         "  subset=\"validation\",\n",
@@ -419,7 +421,7 @@
       },
       "outputs": [],
       "source": [
-        "normalization_layer = tf.keras.layers.Rescaling(1./255)\n",
+        "normalization_layer = keras.layers.Rescaling(1./255)\n",
         "train_ds = train_ds.map(lambda x, y: (normalization_layer(x), y)) # Where x—images, y—labels.\n",
         "val_ds = val_ds.map(lambda x, y: (normalization_layer(x), y)) # Where x—images, y—labels."
       ]
@@ -633,9 +635,9 @@
       "source": [
         "num_classes = len(class_names)\n",
         "\n",
-        "model = tf.keras.Sequential([\n",
+        "model = keras.Sequential([\n",
         "  feature_extractor_layer,\n",
-        "  tf.keras.layers.Dense(num_classes)\n",
+        "  keras.layers.Dense(num_classes)\n",
         "])\n",
         "\n",
         "model.summary()"
@@ -683,12 +685,12 @@
       "outputs": [],
       "source": [
         "model.compile(\n",
-        "  optimizer=tf.keras.optimizers.Adam(),\n",
-        "  loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
+        "  optimizer=keras.optimizers.Adam(),\n",
+        "  loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
         "  metrics=['acc'])\n",
         "\n",
         "log_dir = \"logs/fit/\" + datetime.datetime.now().strftime(\"%Y%m%d-%H%M%S\")\n",
-        "tensorboard_callback = tf.keras.callbacks.TensorBoard(\n",
+        "tensorboard_callback = keras.callbacks.TensorBoard(\n",
         "    log_dir=log_dir,\n",
         "    histogram_freq=1) # Enable histogram computation for every epoch."
       ]
@@ -846,7 +848,7 @@
       },
       "outputs": [],
       "source": [
-        "reloaded = tf.keras.models.load_model(export_path)"
+        "reloaded = keras.models.load_model(export_path)"
       ]
     },
     {