diff --git a/notebooks/multi_modal/labs/image_captioning.ipynb b/notebooks/multi_modal/labs/image_captioning.ipynb
index 19d4c4c7..7b0cd0b1 100644
--- a/notebooks/multi_modal/labs/image_captioning.ipynb
+++ b/notebooks/multi_modal/labs/image_captioning.ipynb
@@ -571,9 +571,9 @@
     "decoder_gru = GRU(...)\n",
     "gru_output, gru_state = decoder_gru(embed_x)\n",
     "\n",
-    "decoder_atention = Attention()\n",
+    "decoder_attention = Attention()\n",
     "# TODO: Define the inputs to the Attention layer\n",
-    "context_vector = decoder_atention([...])\n",
+    "context_vector = decoder_attention([...])\n",
     "\n",
     "addition = Add()([gru_output, context_vector])\n",
     "\n",
@@ -728,7 +728,7 @@
     "gru_output, gru_state = decoder_gru(...)\n",
     "\n",
     "# Reuse other layers as well\n",
-    "context_vector = decoder_atention([...])\n",
+    "context_vector = decoder_attention([...])\n",
     "addition_output = Add()([...])\n",
     "layer_norm_output = layer_norm(...)\n",
     "\n",
diff --git a/notebooks/multi_modal/solutions/image_captioning.ipynb b/notebooks/multi_modal/solutions/image_captioning.ipynb
index b00e2b58..91796eb9 100644
--- a/notebooks/multi_modal/solutions/image_captioning.ipynb
+++ b/notebooks/multi_modal/solutions/image_captioning.ipynb
@@ -716,8 +716,8 @@
     ")\n",
     "gru_output, gru_state = decoder_gru(embed_x)\n",
     "\n",
-    "decoder_atention = Attention()\n",
-    "context_vector = decoder_atention([gru_output, encoder_output])\n",
+    "decoder_attention = Attention()\n",
+    "context_vector = decoder_attention([gru_output, encoder_output])\n",
     "\n",
     "addition = Add()([gru_output, context_vector])\n",
     "\n",
@@ -927,7 +927,7 @@
     "gru_output, gru_state = decoder_gru(embed_x, initial_state=gru_state_input)\n",
     "\n",
     "# Reuse other layers as well\n",
-    "context_vector = decoder_atention([gru_output, encoder_output])\n",
+    "context_vector = decoder_attention([gru_output, encoder_output])\n",
     "addition_output = Add()([gru_output, context_vector])\n",
     "layer_norm_output = layer_norm(addition_output)\n",
     "\n",