diff --git a/notebooks/Block 2/Solutions to Exercises - Block 2.ipynb b/notebooks/Block 2/Solutions to Exercises - Block 2.ipynb
index be50fbbf0b3e62277c638c6744ff055085a9850e..ef392cd3e6d8bc01df4e10218135263f45360388 100644
--- a/notebooks/Block 2/Solutions to Exercises - Block 2.ipynb	
+++ b/notebooks/Block 2/Solutions to Exercises - Block 2.ipynb	
@@ -317,7 +317,7 @@
     {
      "data": {
       "text/plain": [
-       "[<matplotlib.lines.Line2D at 0x7f0c7c049310>]"
+       "[<matplotlib.lines.Line2D at 0x7f6d7b5b6650>]"
       ]
      },
      "execution_count": 8,
@@ -326,7 +326,7 @@
     },
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "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\n",
       "text/plain": [
        "<Figure size 432x288 with 1 Axes>"
       ]
@@ -371,7 +371,7 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[[211.33295]]\n"
+      "[[211.27751]]\n"
      ]
     }
    ],
@@ -424,7 +424,7 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "These are the layer variables: [array([[1.8213079]], dtype=float32), array([29.202171], dtype=float32)]\n"
+      "These are the layer variables: [array([[1.8288891]], dtype=float32), array([28.388597], dtype=float32)]\n"
      ]
     }
    ],
@@ -468,17 +468,17 @@
       "Finished training the model\n",
       "[[211.74745]]\n",
       "Model predicts that 100 degrees Celsius is: [[211.74745]] degrees Fahrenheit\n",
-      "These are the l0 variables: [array([[ 0.5121586 ,  0.04711508, -0.22031644, -0.19786932]],\n",
-      "      dtype=float32), array([ 3.271556 ,  2.696947 , -3.1823864, -3.180513 ], dtype=float32)]\n",
-      "These are the l1 variables: [array([[-0.15210658, -0.66997695, -0.7184656 , -1.2507524 ],\n",
-      "       [ 0.00696711, -0.62529814, -0.03808143, -0.97146195],\n",
-      "       [ 0.32403842,  0.94431067, -0.12347277,  0.8358412 ],\n",
-      "       [ 0.22378582,  1.1214566 , -0.7497232 ,  0.76799536]],\n",
-      "      dtype=float32), array([-2.3398592 , -3.1969426 , -0.42988077, -3.3597116 ], dtype=float32)]\n",
-      "These are the l2 variables: [array([[-0.46421504],\n",
-      "       [-1.0556107 ],\n",
-      "       [-0.13796696],\n",
-      "       [-0.815941  ]], dtype=float32), array([3.033012], dtype=float32)]\n"
+      "These are the l0 variables: [array([[ 0.85532796, -0.12807783, -0.39683163, -0.24233988]],\n",
+      "      dtype=float32), array([ 3.0566595,  2.533055 , -3.0615253,  2.7450564], dtype=float32)]\n",
+      "These are the l1 variables: [array([[-0.42321247,  1.0127366 ,  0.54204386, -1.0158232 ],\n",
+      "       [ 0.05112636,  0.6585605 ,  0.4343624 , -0.05569302],\n",
+      "       [ 0.17409138, -0.56599826, -0.13789305, -0.02274832],\n",
+      "       [-0.15293898,  1.0811394 ,  0.8640427 , -0.8141512 ]],\n",
+      "      dtype=float32), array([-2.6983292,  3.1496687,  3.1004574, -2.8369133], dtype=float32)]\n",
+      "These are the l2 variables: [array([[-0.6664304],\n",
+      "       [ 1.0862367],\n",
+      "       [ 0.5182525],\n",
+      "       [-0.9018049]], dtype=float32), array([3.1196442], dtype=float32)]\n"
      ]
     }
    ],
@@ -528,7 +528,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 13,
+   "execution_count": 12,
    "metadata": {},
    "outputs": [
     {
@@ -537,7 +537,7 @@
        "Text(0, 0.5, 'Broken O-rings')"
       ]
      },
-     "execution_count": 13,
+     "execution_count": 12,
      "metadata": {},
      "output_type": "execute_result"
     },
@@ -572,7 +572,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 14,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -604,7 +604,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": 14,
    "metadata": {},
    "outputs": [
     {
@@ -652,7 +652,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 15,
    "metadata": {},
    "outputs": [
     {
@@ -702,16 +702,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": 16,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "<tensorflow.python.keras.callbacks.History at 0x7f0c6c301d50>"
+       "<tensorflow.python.keras.callbacks.History at 0x7f6d7a234790>"
       ]
      },
-     "execution_count": 17,
+     "execution_count": 16,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -728,22 +728,22 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 18,
+   "execution_count": 17,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "[<matplotlib.lines.Line2D at 0x7f0c6c368f10>]"
+       "[<matplotlib.lines.Line2D at 0x7f6d7a041110>]"
       ]
      },
-     "execution_count": 18,
+     "execution_count": 17,
      "metadata": {},
      "output_type": "execute_result"
     },
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "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\n",
       "text/plain": [
        "<Figure size 432x288 with 1 Axes>"
       ]
@@ -765,14 +765,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": 18,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[array([[-0.23143539]], dtype=float32), array([15.04461], dtype=float32)]\n"
+      "[array([[-0.23214144]], dtype=float32), array([15.041557], dtype=float32)]\n"
      ]
     }
    ],
@@ -782,21 +782,21 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 20,
+   "execution_count": 19,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "0.4418225\n",
-      "[[0.443 0.239 0.284 0.333 0.387 0.165 0.136 0.239 0.864 0.614 0.239 0.047\n",
-      "  0.387 0.942 0.387 0.09  0.239 0.024 0.073 0.038 0.09  0.073 0.835]]\n"
+      "0.44163468\n",
+      "[[0.431 0.23  0.274 0.322 0.375 0.158 0.13  0.23  0.859 0.603 0.23  0.045\n",
+      "  0.375 0.939 0.375 0.086 0.23  0.023 0.069 0.036 0.086 0.069 0.829]]\n"
      ]
     },
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "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\n",
       "text/plain": [
        "<Figure size 432x288 with 1 Axes>"
       ]
@@ -846,7 +846,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 21,
+   "execution_count": 20,
    "metadata": {},
    "outputs": [
     {
@@ -857,25 +857,25 @@
       "11493376/11490434 [==============================] - 0s 0us/step\n",
       "Train on 60000 samples, validate on 10000 samples\n",
       "Epoch 1/10\n",
-      "60000/60000 [==============================] - 5s 85us/sample - loss: 319.9237 - accuracy: 0.8414 - val_loss: 346.2594 - val_accuracy: 0.8395\n",
+      "60000/60000 [==============================] - 5s 85us/sample - loss: 325.1449 - accuracy: 0.8413 - val_loss: 226.6903 - val_accuracy: 0.8844\n",
       "Epoch 2/10\n",
-      "60000/60000 [==============================] - 5s 81us/sample - loss: 255.3777 - accuracy: 0.8697 - val_loss: 263.5786 - val_accuracy: 0.8611\n",
+      "60000/60000 [==============================] - 5s 81us/sample - loss: 263.3936 - accuracy: 0.8693 - val_loss: 198.8350 - val_accuracy: 0.8932\n",
       "Epoch 3/10\n",
-      "60000/60000 [==============================] - 5s 79us/sample - loss: 259.5848 - accuracy: 0.8706 - val_loss: 256.6105 - val_accuracy: 0.8759\n",
+      "60000/60000 [==============================] - 5s 82us/sample - loss: 251.9340 - accuracy: 0.8727 - val_loss: 233.2403 - val_accuracy: 0.8758\n",
       "Epoch 4/10\n",
-      "60000/60000 [==============================] - 5s 77us/sample - loss: 242.7680 - accuracy: 0.8770 - val_loss: 199.7734 - val_accuracy: 0.9007\n",
+      "60000/60000 [==============================] - 5s 76us/sample - loss: 247.5036 - accuracy: 0.8753 - val_loss: 243.0418 - val_accuracy: 0.8901\n",
       "Epoch 5/10\n",
-      "60000/60000 [==============================] - 5s 76us/sample - loss: 243.7667 - accuracy: 0.8778 - val_loss: 232.4277 - val_accuracy: 0.8872\n",
+      "60000/60000 [==============================] - 5s 80us/sample - loss: 246.6533 - accuracy: 0.8766 - val_loss: 192.8183 - val_accuracy: 0.9034\n",
       "Epoch 6/10\n",
-      "60000/60000 [==============================] - 5s 81us/sample - loss: 236.8930 - accuracy: 0.8798 - val_loss: 255.9200 - val_accuracy: 0.8821\n",
+      "60000/60000 [==============================] - 5s 82us/sample - loss: 239.3516 - accuracy: 0.8789 - val_loss: 224.0492 - val_accuracy: 0.8966\n",
       "Epoch 7/10\n",
-      "60000/60000 [==============================] - 5s 78us/sample - loss: 241.8715 - accuracy: 0.8786 - val_loss: 236.1234 - val_accuracy: 0.8918\n",
+      "60000/60000 [==============================] - 5s 80us/sample - loss: 242.2074 - accuracy: 0.8787 - val_loss: 221.8749 - val_accuracy: 0.8888\n",
       "Epoch 8/10\n",
-      "60000/60000 [==============================] - 5s 81us/sample - loss: 233.4837 - accuracy: 0.8824 - val_loss: 245.5468 - val_accuracy: 0.8893\n",
+      "60000/60000 [==============================] - 5s 79us/sample - loss: 236.2757 - accuracy: 0.8813 - val_loss: 219.4592 - val_accuracy: 0.8895\n",
       "Epoch 9/10\n",
-      "60000/60000 [==============================] - 5s 79us/sample - loss: 231.7262 - accuracy: 0.8824 - val_loss: 237.4862 - val_accuracy: 0.8906\n",
+      "60000/60000 [==============================] - 5s 81us/sample - loss: 234.2579 - accuracy: 0.8813 - val_loss: 198.6786 - val_accuracy: 0.9073\n",
       "Epoch 10/10\n",
-      "60000/60000 [==============================] - 5s 80us/sample - loss: 232.5862 - accuracy: 0.8828 - val_loss: 290.5393 - val_accuracy: 0.8724\n"
+      "60000/60000 [==============================] - 5s 81us/sample - loss: 232.5012 - accuracy: 0.8836 - val_loss: 267.6900 - val_accuracy: 0.8717\n"
      ]
     }
    ],
@@ -915,15 +915,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 22,
+   "execution_count": 21,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "10000/10000 [==============================] - 1s 55us/sample - loss: 290.5393 - accuracy: 0.8724\n",
-      "Accuracy on test dataset: 0.8724\n"
+      "10000/10000 [==============================] - 1s 54us/sample - loss: 267.6900 - accuracy: 0.8717\n",
+      "Accuracy on test dataset: 0.8717\n"
      ]
     }
    ],
@@ -941,7 +941,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 22,
    "metadata": {},
    "outputs": [
     {
@@ -950,25 +950,25 @@
      "text": [
       "Train on 60000 samples, validate on 10000 samples\n",
       "Epoch 1/10\n",
-      "60000/60000 [==============================] - 5s 83us/sample - loss: 328.2818 - accuracy: 0.8390 - val_loss: 206.0383 - val_accuracy: 0.8951\n",
+      "60000/60000 [==============================] - 5s 84us/sample - loss: 337.0724 - accuracy: 0.8372 - val_loss: 239.2647 - val_accuracy: 0.8929\n",
       "Epoch 2/10\n",
-      "60000/60000 [==============================] - 5s 83us/sample - loss: 285.4828 - accuracy: 0.8601 - val_loss: 236.3918 - val_accuracy: 0.8797\n",
+      "60000/60000 [==============================] - 5s 82us/sample - loss: 290.0143 - accuracy: 0.8592 - val_loss: 269.1680 - val_accuracy: 0.8640\n",
       "Epoch 3/10\n",
-      "60000/60000 [==============================] - 5s 81us/sample - loss: 289.9700 - accuracy: 0.8590 - val_loss: 243.5878 - val_accuracy: 0.8860\n",
+      "60000/60000 [==============================] - 5s 83us/sample - loss: 288.0693 - accuracy: 0.8612 - val_loss: 343.6055 - val_accuracy: 0.8224\n",
       "Epoch 4/10\n",
-      "60000/60000 [==============================] - 5s 82us/sample - loss: 281.4834 - accuracy: 0.8624 - val_loss: 255.3606 - val_accuracy: 0.8745\n",
+      "60000/60000 [==============================] - 5s 80us/sample - loss: 292.7700 - accuracy: 0.8601 - val_loss: 245.0807 - val_accuracy: 0.8847\n",
       "Epoch 5/10\n",
-      "60000/60000 [==============================] - 5s 78us/sample - loss: 289.1926 - accuracy: 0.8604 - val_loss: 230.2264 - val_accuracy: 0.8903\n",
+      "60000/60000 [==============================] - 5s 84us/sample - loss: 294.5471 - accuracy: 0.8604 - val_loss: 308.2878 - val_accuracy: 0.8575\n",
       "Epoch 6/10\n",
-      "60000/60000 [==============================] - 5s 82us/sample - loss: 288.9536 - accuracy: 0.8629 - val_loss: 237.8040 - val_accuracy: 0.8687\n",
+      "60000/60000 [==============================] - 5s 81us/sample - loss: 291.4390 - accuracy: 0.8600 - val_loss: 256.9222 - val_accuracy: 0.8781\n",
       "Epoch 7/10\n",
-      "60000/60000 [==============================] - 5s 83us/sample - loss: 284.3899 - accuracy: 0.8610 - val_loss: 285.2041 - val_accuracy: 0.8610\n",
+      "60000/60000 [==============================] - 5s 79us/sample - loss: 284.8989 - accuracy: 0.8613 - val_loss: 289.8903 - val_accuracy: 0.8623\n",
       "Epoch 8/10\n",
-      "60000/60000 [==============================] - 5s 83us/sample - loss: 283.2107 - accuracy: 0.8633 - val_loss: 617.0699 - val_accuracy: 0.7375\n",
+      "60000/60000 [==============================] - 5s 82us/sample - loss: 285.4971 - accuracy: 0.8608 - val_loss: 219.4322 - val_accuracy: 0.8888\n",
       "Epoch 9/10\n",
-      "60000/60000 [==============================] - 5s 83us/sample - loss: 281.6904 - accuracy: 0.8613 - val_loss: 259.5067 - val_accuracy: 0.8739\n",
+      "60000/60000 [==============================] - 5s 82us/sample - loss: 283.5283 - accuracy: 0.8622 - val_loss: 233.1485 - val_accuracy: 0.8888\n",
       "Epoch 10/10\n",
-      "59872/60000 [============================>.] - ETA: 0s - loss: 283.1054 - accuracy: 0.8623"
+      "60000/60000 [==============================] - 5s 82us/sample - loss: 277.7306 - accuracy: 0.8627 - val_loss: 221.5486 - val_accuracy: 0.8934\n"
      ]
     }
    ],
@@ -1010,9 +1010,18 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 23,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "10000/10000 [==============================] - 0s 46us/sample - loss: 221.5486 - accuracy: 0.8934\n",
+      "Accuracy on test dataset: 0.8934\n"
+     ]
+    }
+   ],
    "source": [
     "# Evaluate Network\n",
     "test_loss, test_accuracy = model.evaluate(X_test, y_test_cat)\n",
diff --git a/notebooks/Block 4/Solutions to Exercises Block 4 - Convolutional Neural Networks.ipynb b/notebooks/Block 4/Solutions to Exercises Block 4 - Convolutional Neural Networks.ipynb
index f2e4ee1a00f08f520417387f9567f6e0ad6255f7..0269a2f6f74face9a7313b224c50abba9a920d14 100644
--- a/notebooks/Block 4/Solutions to Exercises Block 4 - Convolutional Neural Networks.ipynb	
+++ b/notebooks/Block 4/Solutions to Exercises Block 4 - Convolutional Neural Networks.ipynb	
@@ -983,7 +983,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.6.8"
+   "version": "3.7.6"
   }
  },
  "nbformat": 4,
diff --git a/requirements.txt b/requirements.txt
index e4a7c50fda8a743930e5181852e583a1ec651178..b8730c3c9e170881a255fbd16a6cc64eb7497468 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -6,4 +6,5 @@ tensorflow==2.1.0
 seaborn==0.11.0
 scikit-learn==0.23.2
 vega_datasets==0.8.0
+altair==4.1.0
 mrcnn==0.2
\ No newline at end of file