Various small fixes

Author: fchollet

chapter02_mathematical-building-blocks.ipynb

@@ -131,9 +131,9 @@
    },
    "outputs": [],
    "source": [
-    "from tensorflow.keras import models\n",
+    "from tensorflow import keras\n",
     "from tensorflow.keras import layers\n",
-    "model = models.Sequential([\n",
+    "model = keras.Sequential([\n",
     "    layers.Dense(512, activation=\"relu\"),\n",
     "    layers.Dense(10, activation=\"softmax\")\n",
     "])"
@@ -477,8 +477,8 @@
    },
    "outputs": [],
    "source": [
-    "digit = train_images[4]\n",
     "import matplotlib.pyplot as plt\n",
+    "digit = train_images[4]\n",
     "plt.imshow(digit, cmap=plt.cm.binary)\n",
     "plt.show()"
    ]
@@ -1099,7 +1099,7 @@
     "b = tf.Variable(tf.zeros((2,)))\n",
     "x = tf.random.uniform((2, 2))\n",
     "with tf.GradientTape() as tape:\n",
-    "    y = tf.matmul(W, x) + b\n",
+    "    y = tf.matmul(x, W) + b\n",
     "grad_of_y_wrt_W_and_b = tape.gradient(y, [W, b])"
    ]
   },
@@ -1194,7 +1194,6 @@
     "import tensorflow as tf\n",
     "\n",
     "class NaiveDense:\n",
-    "\n",
     "    def __init__(self, input_size, output_size, activation):\n",
     "        self.activation = activation\n",
     "\n",
@@ -1232,7 +1231,6 @@
    "outputs": [],
    "source": [
     "class NaiveSequential:\n",
-    "\n",
     "    def __init__(self, layers):\n",
     "        self.layers = layers\n",
     "\n",
@@ -1283,7 +1281,6 @@
    "outputs": [],
    "source": [
     "class BatchGenerator:\n",
-    "\n",
     "    def __init__(self, images, labels, batch_size=128):\n",
     "        self.index = 0\n",
     "        self.images = images\n",

chapter03_introduction-to-keras-and-tf.ipynb

@@ -610,7 +610,7 @@
    },
    "outputs": [],
    "source": [
-    "for step in range(20):\n",
+    "for step in range(40):\n",
     "    loss = training_step(inputs, targets)\n",
     "    print(f\"Loss at step {step}: {loss:.4f}\")"
    ]
@@ -827,7 +827,7 @@
     "colab_type": "text"
    },
    "source": [
-    "### Understanding the \"fit\" method"
+    "### Understanding the `fit` method"
    ]
   },
   {
@@ -902,10 +902,10 @@
     "shuffled_targets = targets[indices_permutation]\n",
     "\n",
     "num_validation_samples = int(0.3 * len(inputs))\n",
-    "val_inputs = shuffled_inputs[-num_validation_samples:]\n",
-    "val_targets = shuffled_targets[-num_validation_samples:]\n",
-    "training_inputs = shuffled_inputs[:num_validation_samples]\n",
-    "training_targets = shuffled_targets[:num_validation_samples]\n",
+    "val_inputs = shuffled_inputs[:num_validation_samples]\n",
+    "val_targets = shuffled_targets[:num_validation_samples]\n",
+    "training_inputs = shuffled_inputs[num_validation_samples:]\n",
+    "training_targets = shuffled_targets[num_validation_samples:]\n",
     "model.fit(\n",
     "    training_inputs,\n",
     "    training_targets,\n",

chapter04_getting-started-with-neural-networks.ipynb

@@ -130,7 +130,7 @@
     "colab_type": "text"
    },
    "source": [
-    "**Encoding the integer sequences via one-hot encoding**"
+    "**Encoding the integer sequences via multi-hot encoding**"
    ]
   },
   {
@@ -145,7 +145,8 @@
     "def vectorize_sequences(sequences, dimension=10000):\n",
     "    results = np.zeros((len(sequences), dimension))\n",
     "    for i, sequence in enumerate(sequences):\n",
-    "        results[i, sequence] = 1.\n",
+    "        for j in sequence:\n",
+    "            results[i, j] = 1.\n",
     "    return results\n",
     "x_train = vectorize_sequences(train_data)\n",
     "x_test = vectorize_sequences(test_data)"
@@ -281,9 +282,6 @@
    },
    "outputs": [],
    "source": [
-    "model.compile(optimizer=\"rmsprop\",\n",
-    "              loss=\"binary_crossentropy\",\n",
-    "              metrics=[\"accuracy\"])\n",
     "history = model.fit(partial_x_train,\n",
     "                    partial_y_train,\n",
     "                    epochs=20,\n",
@@ -574,11 +572,6 @@
    },
    "outputs": [],
    "source": [
-    "def vectorize_sequences(sequences, dimension=10000):\n",
-    "    results = np.zeros((len(sequences), dimension))\n",
-    "    for i, sequence in enumerate(sequences):\n",
-    "        results[i, sequence] = 1.\n",
-    "    return results\n",
     "x_train = vectorize_sequences(train_data)\n",
     "x_test = vectorize_sequences(test_data)"
    ]
@@ -605,8 +598,8 @@
     "    for i, label in enumerate(labels):\n",
     "        results[i, label] = 1.\n",
     "    return results\n",
-    "one_hot_train_labels = to_one_hot(train_labels)\n",
-    "one_hot_test_labels = to_one_hot(test_labels)"
+    "y_train = to_one_hot(train_labels)\n",
+    "y_test = to_one_hot(test_labels)"
    ]
   },
   {
@@ -618,8 +611,8 @@
    "outputs": [],
    "source": [
     "from tensorflow.keras.utils import to_categorical\n",
-    "one_hot_train_labels = to_categorical(train_labels)\n",
-    "one_hot_test_labels = to_categorical(test_labels)"
+    "y_train = to_categorical(train_labels)\n",
+    "y_test = to_categorical(test_labels)"
    ]
   },
   {
@@ -705,8 +698,8 @@
    "source": [
     "x_val = x_train[:1000]\n",
     "partial_x_train = x_train[1000:]\n",
-    "y_val = one_hot_train_labels[:1000]\n",
-    "partial_y_train = one_hot_train_labels[1000:]"
+    "y_val = y_train[:1000]\n",
+    "partial_y_train = y_train[1000:]"
    ]
   },
   {
@@ -816,12 +809,11 @@
     "model.compile(optimizer=\"rmsprop\",\n",
     "              loss=\"categorical_crossentropy\",\n",
     "              metrics=[\"accuracy\"])\n",
-    "model.fit(partial_x_train,\n",
-    "          partial_y_train,\n",
+    "model.fit(x_train,\n",
+    "          y_train,\n",
     "          epochs=9,\n",
-    "          batch_size=512,\n",
-    "          validation_data=(x_val, y_val))\n",
-    "results = model.evaluate(x_test, one_hot_test_labels)"
+    "          batch_size=512)\n",
+    "results = model.evaluate(x_test, y_test)"
    ]
   },
   {
@@ -847,7 +839,7 @@
     "test_labels_copy = copy.copy(test_labels)\n",
     "np.random.shuffle(test_labels_copy)\n",
     "hits_array = np.array(test_labels) == np.array(test_labels_copy)\n",
-    "float(np.sum(hits_array)) / len(test_labels)"
+    "hits_array.mean()"
    ]
   },
   {
@@ -1182,7 +1174,7 @@
     "        axis=0)\n",
     "    model = build_model()\n",
     "    model.fit(partial_train_data, partial_train_targets,\n",
-    "              epochs=num_epochs, batch_size=1, verbose=0)\n",
+    "              epochs=num_epochs, batch_size=16, verbose=0)\n",
     "    val_mse, val_mae = model.evaluate(val_data, val_targets, verbose=0)\n",
     "    all_scores.append(val_mae)"
    ]
@@ -1243,7 +1235,7 @@
     "    model = build_model()\n",
     "    history = model.fit(partial_train_data, partial_train_targets,\n",
     "                        validation_data=(val_data, val_targets),\n",
-    "                        epochs=num_epochs, batch_size=1, verbose=0)\n",
+    "                        epochs=num_epochs, batch_size=16, verbose=0)\n",
     "    mae_history = history.history[\"val_mae\"]\n",
     "    all_mae_histories.append(mae_history)"
    ]
@@ -1298,7 +1290,7 @@
     "colab_type": "text"
    },
    "source": [
-    "**Plotting smoothed validation scores, excluding the first 10 data points**"
+    "**Plotting validation scores, excluding the first 10 data points**"
    ]
   },
   {
@@ -1309,17 +1301,8 @@
    },
    "outputs": [],
    "source": [
-    "def smooth_curve(points, factor=0.9):\n",
-    "    smoothed_points = []\n",
-    "    for point in points:\n",
-    "        if smoothed_points:\n",
-    "            previous = smoothed_points[-1]\n",
-    "            smoothed_points.append(previous * factor + point * (1 - factor))\n",
-    "        else:\n",
-    "            smoothed_points.append(point)\n",
-    "    return smoothed_points\n",
-    "smooth_mae_history = smooth_curve(average_mae_history[10:])\n",
-    "plt.plot(range(1, len(smooth_mae_history) + 1), smooth_mae_history)\n",
+    "truncated_mae_history = average_mae_history[10:]\n",
+    "plt.plot(range(1, len(truncated_mae_history) + 1), truncated_mae_history)\n",
     "plt.xlabel(\"Epochs\")\n",
     "plt.ylabel(\"Validation MAE\")\n",
     "plt.show()"
@@ -1344,7 +1327,7 @@
    "source": [
     "model = build_model()\n",
     "model.fit(train_data, train_targets,\n",
-    "          epochs=80, batch_size=16, verbose=0)\n",
+    "          epochs=130, batch_size=16, verbose=0)\n",
     "test_mse_score, test_mae_score = model.evaluate(test_data, test_targets)"
    ]
   },

chapter08_intro-to-dl-for-computer-vision.ipynb

@@ -507,7 +507,7 @@
     "colab_type": "text"
    },
    "source": [
-    "**Applying a transformation to Dataset elements using `map()``**"
+    "**Applying a transformation to Dataset elements using `map()`**"
    ]
   },
   {