demos/keras_train.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# HiddenLayer Train Demo - Keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import time\n",
    "import random\n",
    "import numpy as np\n",
    "import hiddenlayer as hl"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Basic Use Case\n",
    "\n",
    "To track your training, you need to use two Classes: History to store the metrics, and Canvas to draw them.\n",
    "This example simulates a training loop."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# A History object to store metrics\n",
    "history1 = hl.History()\n",
    "\n",
    "# A Canvas object to draw the metrics\n",
    "canvas1 = hl.Canvas()\n",
    "\n",
    "# Simulate a training loop with two metrics: loss and accuracy\n",
    "loss = 1\n",
    "accuracy = 0\n",
    "for step in range(800):\n",
    "    # Fake loss and accuracy\n",
    "    loss -= loss * np.random.uniform(-.09, 0.1)\n",
    "    accuracy = max(0, accuracy + (1 - accuracy) * np.random.uniform(-.09, 0.1))\n",
    "\n",
    "    # Log metrics and display them at certain intervals\n",
    "    if step % 10 == 0:\n",
    "        # Store metrics in the history object\n",
    "        history1.log(step, loss=loss, accuracy=accuracy)\n",
    "\n",
    "        # Plot the two metrics in one graph\n",
    "        canvas1.draw_plot([history1[\"loss\"], history1[\"accuracy\"]])\n",
    "\n",
    "        time.sleep(0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparing Experiments\n",
    "\n",
    "Often you'd want to compare how experiments compare to each other. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# New history and canvas objects\n",
    "history2 = hl.History()\n",
    "canvas2 = hl.Canvas()\n",
    "\n",
    "# Simulate a training loop with two metrics: loss and accuracy\n",
    "loss = 1\n",
    "accuracy = 0\n",
    "for step in range(800):\n",
    "    # Fake loss and accuracy\n",
    "    loss -= loss * np.random.uniform(-.09, 0.1)\n",
    "    accuracy = max(0, accuracy + (1 - accuracy) * np.random.uniform(-.09, 0.1))\n",
    "\n",
    "    # Log metrics and display them at certain intervals\n",
    "    if step % 10 == 0:\n",
    "        history2.log(step, loss=loss, accuracy=accuracy)\n",
    "\n",
    "        # Draw two plots\n",
    "        # Encluse them in a \"with\" context to ensure they render together\n",
    "        with canvas2:\n",
    "            canvas2.draw_plot([history1[\"loss\"], history2[\"loss\"]],\n",
    "                              labels=[\"Loss 1\", \"Loss 2\"])\n",
    "            canvas2.draw_plot([history1[\"accuracy\"], history2[\"accuracy\"]],\n",
    "                              labels=[\"Accuracy 1\", \"Accuracy 2\"])\n",
    "        time.sleep(0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Saving and Loading Histories\n",
    "\n",
    "The History object store the metrics in RAM, which is often good enough for simple \n",
    "expriments. To keep the history, you can save/load them with."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save experiments 1 and 2\n",
    "history1.save(\"experiment1.pkl\")\n",
    "history2.save(\"experiment2.pkl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load them again. To verify it's working, load them into new objects.\n",
    "h1 = hl.History()\n",
    "h2 = hl.History()\n",
    "h1.load(\"experiment1.pkl\")\n",
    "h2.load(\"experiment2.pkl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Verify the data is loaded"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Last Step: 790\n",
      "Training Time: 0:00:31.288706\n"
     ]
    }
   ],
   "source": [
    "# Show a summary of the experiment\n",
    "h1.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Draw a plot of experiment 2\n",
    "hl.Canvas().draw_plot(h2[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Custom Visualizations\n",
    "\n",
    "Adding new custom visualizations is pretty easy. Derive a new class from `Canvas` and add your new method to it. You can use any of the drawing functions provided by `matplotlib`.\n",
    "\n",
    "Here is an example to display the accuracy metric as a pie chart."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "class MyCanvas(hl.Canvas):\n",
    "    \"\"\"Extending Canvas to add a pie chart method.\"\"\"\n",
    "    \n",
    "    def draw_pie(self, metric):\n",
    "        # Method name must start with 'draw_' for the Canvas to automatically manage it\n",
    "        \n",
    "        # Use the provided matplotlib Axes in self.ax\n",
    "        self.ax.axis('equal')  # set square aspect ratio\n",
    "\n",
    "        # Get latest value of the metric\n",
    "        value = np.clip(metric.data[-1], 0, 1)\n",
    "        \n",
    "        # Draw pie chart\n",
    "        self.ax.pie([value, 1-value], labels=[\"Accuracy\", \"\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In addition to the pie chart, let's use image visualizations (which is built-in)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 8 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "history3 = hl.History()\n",
    "canvas3 = MyCanvas()  # My custom Canvas\n",
    "\n",
    "# Simulate a training loop\n",
    "loss = 1\n",
    "accuracy = 0\n",
    "for step in range(400):\n",
    "    # Fake loss and accuracy\n",
    "    loss -= loss * np.random.uniform(-.09, 0.1)\n",
    "    accuracy = max(0, accuracy + (1 - accuracy) * np.random.uniform(-.09, 0.1))\n",
    "\n",
    "    if step % 10 == 0:\n",
    "        # Log loss and accuracy\n",
    "        history3.log(step, loss=loss, accuracy=accuracy)\n",
    "\n",
    "        # Log a fake image metric (e.g. image generated by a GAN)\n",
    "        image = np.sin(np.sum(((np.indices([32, 32]) - 16) * 0.5 * accuracy) ** 2, 0))\n",
    "        history3.log(step, image=image)\n",
    "        \n",
    "        # Display\n",
    "        with canvas3:\n",
    "            canvas3.draw_pie(history3[\"accuracy\"])\n",
    "            canvas3.draw_plot([history3[\"accuracy\"], history3[\"loss\"]])\n",
    "            canvas3.draw_image(history3[\"image\"])\n",
    "\n",
    "        time.sleep(0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Running without a GUI\n",
    "\n",
    "If the training loop is running on a server without a GUI, then use the `History` `progress()` method to print a text status."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Step 790: loss: 0.0037487527237774275  accuracy: 0.9870165521966132  \n"
     ]
    }
   ],
   "source": [
    "# Print the metrics of the last step.\n",
    "history1.progress()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also periodically saving a snapshot of the graphs to disk to view later. See `demos/history_demo.py` for an example.\n",
    "\n",
    "First, set matplotlib backend to Agg.\n",
    "```Python\n",
    "# Set matplotlib backend to Agg. MUST be done BEFORE importing hiddenlayer\n",
    "import matplotlib\n",
    "matplotlib.use(\"Agg\")\n",
    "```\n",
    "\n",
    "Then, in the training loop:\n",
    "```\n",
    "    # Print a text progress status in the loop\n",
    "    history.progress()\n",
    "\n",
    "    # Occasionally, save a snapshot of the graphs.\n",
    "    canvas.draw_plot([h[\"loss\"], h[\"accuracy\"]])\n",
    "    canvas.save(\"training_graph.png\")\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Real Training Example\n",
    "\n",
    "Below, we train an MNIST classifier using transfer learning with:\n",
    "- a pre-trained VGG16 convnet for feature extraction,\n",
    "- our own 10-output sotfmax layer with cross-entropy loss (log loss) as our classifier.\n",
    "\n",
    "We use:\n",
    "- `Keras` for model training\n",
    "- `sklearn` to generate a confusion matrix on the validation split\n",
    "- `hiddenlayer` to validate our \"before/after\" convnet and check our progress (visualizing the confusion matrix as we go)\n",
    "\n",
    "This model only takes **~3 minutes to train** to reach 97% prediction accuracy on the validation split."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.datasets import mnist\n",
    "from tensorflow.keras.applications.vgg16 import VGG16\n",
    "import tensorflow.keras.backend as K\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.layers import Flatten, Dense, Dropout\n",
    "from tensorflow.keras.callbacks import Callback\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score\n",
    "import cv2\n",
    "import itertools\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Hyperparams and constants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Minimum size of input images Keras' VGG16 will accept\n",
    "input_shape = (48, 48, 3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepare the train/val dataset splits"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the data using the default train/validation split\n",
    "(x_train, y_train), (x_val, y_val) = mnist.load_data()\n",
    "\n",
    "# Extract image size\n",
    "_, x_height, x_width = x_train.shape\n",
    "\n",
    "# Extract number of classes\n",
    "y_classes = len(np.unique(y_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train\tTensor  uint8 (60000, 28, 28)  min: 0.000  max: 255.000\n",
      "x_val\tTensor  uint8 (10000, 28, 28)  min: 0.000  max: 255.000\n",
      "y_train\tTensor  uint8 (60000,)  min: 0.000  max: 9.000\n",
      "y_val\tTensor  uint8 (10000,)  min: 0.000  max: 9.000\n",
      "Sample size: 28x28\n",
      "Number of classes: 10\n"
     ]
    }
   ],
   "source": [
    "# Sanity check\n",
    "hl.write(\"x_train\", x_train)\n",
    "hl.write(\"x_val\", x_val)\n",
    "hl.write(\"y_train\", y_train)\n",
    "hl.write(\"y_val\", y_val)\n",
    "print(f\"Sample size: {x_height}x{x_width}\")\n",
    "print(f\"Number of classes: {y_classes}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x190.08 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visual check\n",
    "hl.show_images(x_train[:5])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Adapt the train/val splits to the convnet input/label format requirements"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The VGG16 backbone was trained on RGB data in the [0.,1.] range\n",
    "# -> convert data to float and perform min-max normalization\n",
    "x_train = x_train.astype('float32') / 255.\n",
    "x_val = x_val.astype('float32') / 255\n",
    "\n",
    "# The VGG16 backbone was trained on color RGB data (imagenet)\n",
    "# -> resize the samples from (#, 28, 28) to (#, input_shape[0], input_shape[1])\n",
    "if input_shape[0:2] != (x_height, x_width):\n",
    "    x_train = [cv2.resize(x, input_shape[0:2]) for x in x_train]\n",
    "    x_val = [cv2.resize(x, input_shape[0:2]) for x in x_val]\n",
    "\n",
    "# -> reshape the samples from (#, H, W) to (#, H, W, 3)\n",
    "#    by copying the \"gray\" MNIST images identically across the 3 RGB channels\n",
    "x_train = np.stack((x_train, x_train, x_train), axis=-1)\n",
    "x_val = np.stack((x_val, x_val, x_val), axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# One-hot encode labels for use by softmax with multiclass cross-entropy loss\n",
    "y_train = to_categorical(y_train, y_classes)\n",
    "y_val = to_categorical(y_val, y_classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x_train\tTensor  float32 (60000, 48, 48, 3)  min: 0.000  max: 1.000\n",
      "x_val\tTensor  float32 (10000, 48, 48, 3)  min: 0.000  max: 1.000\n",
      "y_train\tTensor  float32 (60000, 10)  min: 0.000  max: 1.000\n",
      "y_val\tTensor  float32 (10000, 10)  min: 0.000  max: 1.000\n"
     ]
    }
   ],
   "source": [
    "# Sanity check\n",
    "hl.write(\"x_train\", x_train)\n",
    "hl.write(\"x_val\", x_val)\n",
    "hl.write(\"y_train\", y_train)\n",
    "hl.write(\"y_val\", y_val)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x190.08 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visual check\n",
    "hl.show_images(x_train[:5])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Put the model (convnet base + classifier) together"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set the learning phase to training\n",
    "K.set_learning_phase(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build the VGG16 model as a feature extractor\n",
    "convnet = VGG16(weights='imagenet', include_top=False, input_shape=input_shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\r\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\r\n",
       " \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\r\n",
       "<!-- Generated by graphviz version 2.38.0 (20140413.2041)\r\n",
       " -->\r\n",
       "<!-- Title: %3 Pages: 1 -->\r\n",
       "<svg width=\"241pt\" height=\"978pt\"\r\n",
       " viewBox=\"0.00 0.00 241.00 978.00\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\r\n",
       "<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(72 942)\">\r\n",
       "<title>%3</title>\r\n",
       "<polygon fill=\"#ffffff\" stroke=\"none\" points=\"-72,36 -72,-942 169,-942 169,36 -72,36\"/>\r\n",
       "<!-- input_1 -->\r\n",
       "<g id=\"node1\" class=\"node\"><title>input_1</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"75,-906 9,-906 9,-870 75,-870 75,-906\"/>\r\n",
       "<text text-anchor=\"start\" x=\"17\" y=\"-885\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Placeholder</text>\r\n",
       "</g>\r\n",
       "<!-- 16527022116719804048 -->\r\n",
       "<g id=\"node7\" class=\"node\"><title>16527022116719804048</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"84,-823 0,-823 0,-779 84,-779 84,-823\"/>\r\n",
       "<text text-anchor=\"start\" x=\"8\" y=\"-807\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Conv3x3 &gt; Relu</text>\r\n",
       "<text text-anchor=\"start\" x=\"70\" y=\"-786\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">x2</text>\r\n",
       "</g>\r\n",
       "<!-- input_1&#45;&gt;16527022116719804048 -->\r\n",
       "<g id=\"edge1\" class=\"edge\"><title>input_1&#45;&gt;16527022116719804048</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-869.799C42,-859.369 42,-845.742 42,-833.443\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-833.09 42,-823.09 38.5001,-833.09 45.5001,-833.09\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-844\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x48x48x3</text>\r\n",
       "</g>\r\n",
       "<!-- block1_pool/MaxPool -->\r\n",
       "<g id=\"node2\" class=\"node\"><title>block1_pool/MaxPool</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"82,-732 2,-732 2,-696 82,-696 82,-732\"/>\r\n",
       "<text text-anchor=\"start\" x=\"10\" y=\"-711\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">MaxPool2x2/s2</text>\r\n",
       "</g>\r\n",
       "<!-- 14080021540030993213 -->\r\n",
       "<g id=\"node8\" class=\"node\"><title>14080021540030993213</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"84,-649 0,-649 0,-605 84,-605 84,-649\"/>\r\n",
       "<text text-anchor=\"start\" x=\"8\" y=\"-633\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Conv3x3 &gt; Relu</text>\r\n",
       "<text text-anchor=\"start\" x=\"70\" y=\"-612\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">x2</text>\r\n",
       "</g>\r\n",
       "<!-- block1_pool/MaxPool&#45;&gt;14080021540030993213 -->\r\n",
       "<g id=\"edge3\" class=\"edge\"><title>block1_pool/MaxPool&#45;&gt;14080021540030993213</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-695.799C42,-685.369 42,-671.742 42,-659.443\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-659.09 42,-649.09 38.5001,-659.09 45.5001,-659.09\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-670\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x24x24x64</text>\r\n",
       "</g>\r\n",
       "<!-- block2_pool/MaxPool -->\r\n",
       "<g id=\"node3\" class=\"node\"><title>block2_pool/MaxPool</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"82,-558 2,-558 2,-522 82,-522 82,-558\"/>\r\n",
       "<text text-anchor=\"start\" x=\"10\" y=\"-537\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">MaxPool2x2/s2</text>\r\n",
       "</g>\r\n",
       "<!-- 4750487785616069892 -->\r\n",
       "<g id=\"node9\" class=\"node\"><title>4750487785616069892</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"84,-475 0,-475 0,-431 84,-431 84,-475\"/>\r\n",
       "<text text-anchor=\"start\" x=\"8\" y=\"-459\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Conv3x3 &gt; Relu</text>\r\n",
       "<text text-anchor=\"start\" x=\"70\" y=\"-438\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">x3</text>\r\n",
       "</g>\r\n",
       "<!-- block2_pool/MaxPool&#45;&gt;4750487785616069892 -->\r\n",
       "<g id=\"edge5\" class=\"edge\"><title>block2_pool/MaxPool&#45;&gt;4750487785616069892</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-521.799C42,-511.369 42,-497.742 42,-485.443\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-485.09 42,-475.09 38.5001,-485.09 45.5001,-485.09\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"69.5\" y=\"-496\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x12x12x128</text>\r\n",
       "</g>\r\n",
       "<!-- block3_pool/MaxPool -->\r\n",
       "<g id=\"node4\" class=\"node\"><title>block3_pool/MaxPool</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"82,-384 2,-384 2,-348 82,-348 82,-384\"/>\r\n",
       "<text text-anchor=\"start\" x=\"10\" y=\"-363\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">MaxPool2x2/s2</text>\r\n",
       "</g>\r\n",
       "<!-- 11119688844565789673 -->\r\n",
       "<g id=\"node10\" class=\"node\"><title>11119688844565789673</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"84,-301 0,-301 0,-257 84,-257 84,-301\"/>\r\n",
       "<text text-anchor=\"start\" x=\"8\" y=\"-285\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Conv3x3 &gt; Relu</text>\r\n",
       "<text text-anchor=\"start\" x=\"70\" y=\"-264\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">x3</text>\r\n",
       "</g>\r\n",
       "<!-- block3_pool/MaxPool&#45;&gt;11119688844565789673 -->\r\n",
       "<g id=\"edge7\" class=\"edge\"><title>block3_pool/MaxPool&#45;&gt;11119688844565789673</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-347.799C42,-337.369 42,-323.742 42,-311.443\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-311.09 42,-301.09 38.5001,-311.09 45.5001,-311.09\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-322\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x6x6x256</text>\r\n",
       "</g>\r\n",
       "<!-- block4_pool/MaxPool -->\r\n",
       "<g id=\"node5\" class=\"node\"><title>block4_pool/MaxPool</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"82,-210 2,-210 2,-174 82,-174 82,-210\"/>\r\n",
       "<text text-anchor=\"start\" x=\"10\" y=\"-189\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">MaxPool2x2/s2</text>\r\n",
       "</g>\r\n",
       "<!-- 3267816487967773741 -->\r\n",
       "<g id=\"node11\" class=\"node\"><title>3267816487967773741</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"84,-127 0,-127 0,-83 84,-83 84,-127\"/>\r\n",
       "<text text-anchor=\"start\" x=\"8\" y=\"-111\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Conv3x3 &gt; Relu</text>\r\n",
       "<text text-anchor=\"start\" x=\"70\" y=\"-90\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">x3</text>\r\n",
       "</g>\r\n",
       "<!-- block4_pool/MaxPool&#45;&gt;3267816487967773741 -->\r\n",
       "<g id=\"edge9\" class=\"edge\"><title>block4_pool/MaxPool&#45;&gt;3267816487967773741</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-173.799C42,-163.369 42,-149.742 42,-137.443\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-137.09 42,-127.09 38.5001,-137.09 45.5001,-137.09\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-148\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x3x3x512</text>\r\n",
       "</g>\r\n",
       "<!-- block5_pool/MaxPool -->\r\n",
       "<g id=\"node6\" class=\"node\"><title>block5_pool/MaxPool</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"82,-36 2,-36 2,-0 82,-0 82,-36\"/>\r\n",
       "<text text-anchor=\"start\" x=\"10\" y=\"-15\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">MaxPool2x2/s2</text>\r\n",
       "</g>\r\n",
       "<!-- 16527022116719804048&#45;&gt;block1_pool/MaxPool -->\r\n",
       "<g id=\"edge2\" class=\"edge\"><title>16527022116719804048&#45;&gt;block1_pool/MaxPool</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-778.989C42,-767.923 42,-754.219 42,-742.336\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-742.062 42,-732.062 38.5001,-742.062 45.5001,-742.062\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-753\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x48x48x64</text>\r\n",
       "</g>\r\n",
       "<!-- 14080021540030993213&#45;&gt;block2_pool/MaxPool -->\r\n",
       "<g id=\"edge4\" class=\"edge\"><title>14080021540030993213&#45;&gt;block2_pool/MaxPool</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-604.989C42,-593.923 42,-580.219 42,-568.336\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-568.062 42,-558.062 38.5001,-568.062 45.5001,-568.062\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"69.5\" y=\"-579\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x24x24x128</text>\r\n",
       "</g>\r\n",
       "<!-- 4750487785616069892&#45;&gt;block3_pool/MaxPool -->\r\n",
       "<g id=\"edge6\" class=\"edge\"><title>4750487785616069892&#45;&gt;block3_pool/MaxPool</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-430.989C42,-419.923 42,-406.219 42,-394.336\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-394.062 42,-384.062 38.5001,-394.062 45.5001,-394.062\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"69.5\" y=\"-405\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x12x12x256</text>\r\n",
       "</g>\r\n",
       "<!-- 11119688844565789673&#45;&gt;block4_pool/MaxPool -->\r\n",
       "<g id=\"edge8\" class=\"edge\"><title>11119688844565789673&#45;&gt;block4_pool/MaxPool</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-256.989C42,-245.923 42,-232.219 42,-220.336\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-220.062 42,-210.062 38.5001,-220.062 45.5001,-220.062\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-231\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x6x6x512</text>\r\n",
       "</g>\r\n",
       "<!-- 3267816487967773741&#45;&gt;block5_pool/MaxPool -->\r\n",
       "<g id=\"edge10\" class=\"edge\"><title>3267816487967773741&#45;&gt;block5_pool/MaxPool</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-82.9888C42,-71.9227 42,-58.219 42,-46.3357\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-46.062 42,-36.0621 38.5001,-46.0621 45.5001,-46.062\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-57\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x3x3x512</text>\r\n",
       "</g>\r\n",
       "</g>\r\n",
       "</svg>\r\n"
      ],
      "text/plain": [
       "<hiddenlayer.graph.Graph at 0x285f2b13208>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Visual check of the \"before\" model architecture using HiddenLayer\n",
    "hl.build_graph(K.get_session().graph)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build our own classifier on top of the VGG16 feature extractor using\n",
    "# two fully-connected layers with dropout and a final softmax layer\n",
    "classifier = Flatten(input_shape=convnet.output_shape[1:])(convnet.output)\n",
    "classifier = Dense(512, activation='relu')(classifier) # fc1\n",
    "classifier = Dropout(0.5)(classifier)\n",
    "classifier = Dense(512, activation='relu')(classifier) # fc2\n",
    "classifier = Dropout(0.5)(classifier)\n",
    "classifier = Dense(10, activation='softmax')(classifier) # softmax classifier for classes 0..9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Finally, put the model together\n",
    "model = Model(inputs=convnet.input, outputs=classifier)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"no\"?>\r\n",
       "<!DOCTYPE svg PUBLIC \"-//W3C//DTD SVG 1.1//EN\"\r\n",
       " \"http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd\">\r\n",
       "<!-- Generated by graphviz version 2.38.0 (20140413.2041)\r\n",
       " -->\r\n",
       "<!-- Title: %3 Pages: 1 -->\r\n",
       "<svg width=\"241pt\" height=\"1539pt\"\r\n",
       " viewBox=\"0.00 0.00 241.00 1539.00\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\r\n",
       "<g id=\"graph0\" class=\"graph\" transform=\"scale(1 1) rotate(0) translate(72 1503)\">\r\n",
       "<title>%3</title>\r\n",
       "<polygon fill=\"#ffffff\" stroke=\"none\" points=\"-72,36 -72,-1503 169,-1503 169,36 -72,36\"/>\r\n",
       "<!-- input_1 -->\r\n",
       "<g id=\"node1\" class=\"node\"><title>input_1</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"75,-1467 9,-1467 9,-1431 75,-1431 75,-1467\"/>\r\n",
       "<text text-anchor=\"start\" x=\"17\" y=\"-1446\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Placeholder</text>\r\n",
       "</g>\r\n",
       "<!-- 7680877476389183126 -->\r\n",
       "<g id=\"node14\" class=\"node\"><title>7680877476389183126</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"84,-1384 0,-1384 0,-1340 84,-1340 84,-1384\"/>\r\n",
       "<text text-anchor=\"start\" x=\"8\" y=\"-1368\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Conv3x3 &gt; Relu</text>\r\n",
       "<text text-anchor=\"start\" x=\"70\" y=\"-1347\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">x2</text>\r\n",
       "</g>\r\n",
       "<!-- input_1&#45;&gt;7680877476389183126 -->\r\n",
       "<g id=\"edge8\" class=\"edge\"><title>input_1&#45;&gt;7680877476389183126</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-1430.8C42,-1420.37 42,-1406.74 42,-1394.44\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-1394.09 42,-1384.09 38.5001,-1394.09 45.5001,-1394.09\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-1405\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x48x48x3</text>\r\n",
       "</g>\r\n",
       "<!-- block1_pool/MaxPool -->\r\n",
       "<g id=\"node2\" class=\"node\"><title>block1_pool/MaxPool</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"82,-1293 2,-1293 2,-1257 82,-1257 82,-1293\"/>\r\n",
       "<text text-anchor=\"start\" x=\"10\" y=\"-1272\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">MaxPool2x2/s2</text>\r\n",
       "</g>\r\n",
       "<!-- 17434212678539375187 -->\r\n",
       "<g id=\"node15\" class=\"node\"><title>17434212678539375187</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"84,-1210 0,-1210 0,-1166 84,-1166 84,-1210\"/>\r\n",
       "<text text-anchor=\"start\" x=\"8\" y=\"-1194\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Conv3x3 &gt; Relu</text>\r\n",
       "<text text-anchor=\"start\" x=\"70\" y=\"-1173\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">x2</text>\r\n",
       "</g>\r\n",
       "<!-- block1_pool/MaxPool&#45;&gt;17434212678539375187 -->\r\n",
       "<g id=\"edge10\" class=\"edge\"><title>block1_pool/MaxPool&#45;&gt;17434212678539375187</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-1256.8C42,-1246.37 42,-1232.74 42,-1220.44\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-1220.09 42,-1210.09 38.5001,-1220.09 45.5001,-1220.09\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-1231\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x24x24x64</text>\r\n",
       "</g>\r\n",
       "<!-- block2_pool/MaxPool -->\r\n",
       "<g id=\"node3\" class=\"node\"><title>block2_pool/MaxPool</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"82,-1119 2,-1119 2,-1083 82,-1083 82,-1119\"/>\r\n",
       "<text text-anchor=\"start\" x=\"10\" y=\"-1098\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">MaxPool2x2/s2</text>\r\n",
       "</g>\r\n",
       "<!-- 7024157625594113643 -->\r\n",
       "<g id=\"node16\" class=\"node\"><title>7024157625594113643</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"84,-1036 0,-1036 0,-992 84,-992 84,-1036\"/>\r\n",
       "<text text-anchor=\"start\" x=\"8\" y=\"-1020\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Conv3x3 &gt; Relu</text>\r\n",
       "<text text-anchor=\"start\" x=\"70\" y=\"-999\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">x3</text>\r\n",
       "</g>\r\n",
       "<!-- block2_pool/MaxPool&#45;&gt;7024157625594113643 -->\r\n",
       "<g id=\"edge12\" class=\"edge\"><title>block2_pool/MaxPool&#45;&gt;7024157625594113643</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-1082.8C42,-1072.37 42,-1058.74 42,-1046.44\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-1046.09 42,-1036.09 38.5001,-1046.09 45.5001,-1046.09\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"69.5\" y=\"-1057\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x12x12x128</text>\r\n",
       "</g>\r\n",
       "<!-- block3_pool/MaxPool -->\r\n",
       "<g id=\"node4\" class=\"node\"><title>block3_pool/MaxPool</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"82,-945 2,-945 2,-909 82,-909 82,-945\"/>\r\n",
       "<text text-anchor=\"start\" x=\"10\" y=\"-924\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">MaxPool2x2/s2</text>\r\n",
       "</g>\r\n",
       "<!-- 8051966429743015524 -->\r\n",
       "<g id=\"node17\" class=\"node\"><title>8051966429743015524</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"84,-862 0,-862 0,-818 84,-818 84,-862\"/>\r\n",
       "<text text-anchor=\"start\" x=\"8\" y=\"-846\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Conv3x3 &gt; Relu</text>\r\n",
       "<text text-anchor=\"start\" x=\"70\" y=\"-825\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">x3</text>\r\n",
       "</g>\r\n",
       "<!-- block3_pool/MaxPool&#45;&gt;8051966429743015524 -->\r\n",
       "<g id=\"edge14\" class=\"edge\"><title>block3_pool/MaxPool&#45;&gt;8051966429743015524</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-908.799C42,-898.369 42,-884.742 42,-872.443\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-872.09 42,-862.09 38.5001,-872.09 45.5001,-872.09\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-883\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x6x6x256</text>\r\n",
       "</g>\r\n",
       "<!-- block4_pool/MaxPool -->\r\n",
       "<g id=\"node5\" class=\"node\"><title>block4_pool/MaxPool</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"82,-771 2,-771 2,-735 82,-735 82,-771\"/>\r\n",
       "<text text-anchor=\"start\" x=\"10\" y=\"-750\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">MaxPool2x2/s2</text>\r\n",
       "</g>\r\n",
       "<!-- 6723214227147554364 -->\r\n",
       "<g id=\"node18\" class=\"node\"><title>6723214227147554364</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"84,-688 0,-688 0,-644 84,-644 84,-688\"/>\r\n",
       "<text text-anchor=\"start\" x=\"8\" y=\"-672\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Conv3x3 &gt; Relu</text>\r\n",
       "<text text-anchor=\"start\" x=\"70\" y=\"-651\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">x3</text>\r\n",
       "</g>\r\n",
       "<!-- block4_pool/MaxPool&#45;&gt;6723214227147554364 -->\r\n",
       "<g id=\"edge16\" class=\"edge\"><title>block4_pool/MaxPool&#45;&gt;6723214227147554364</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-734.799C42,-724.369 42,-710.742 42,-698.443\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-698.09 42,-688.09 38.5001,-698.09 45.5001,-698.09\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-709\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x3x3x512</text>\r\n",
       "</g>\r\n",
       "<!-- block5_pool/MaxPool -->\r\n",
       "<g id=\"node6\" class=\"node\"><title>block5_pool/MaxPool</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"82,-597 2,-597 2,-561 82,-561 82,-597\"/>\r\n",
       "<text text-anchor=\"start\" x=\"10\" y=\"-576\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">MaxPool2x2/s2</text>\r\n",
       "</g>\r\n",
       "<!-- flatten/Reshape -->\r\n",
       "<g id=\"node7\" class=\"node\"><title>flatten/Reshape</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"69,-514 15,-514 15,-478 69,-478 69,-514\"/>\r\n",
       "<text text-anchor=\"start\" x=\"24\" y=\"-493\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Reshape</text>\r\n",
       "</g>\r\n",
       "<!-- block5_pool/MaxPool&#45;&gt;flatten/Reshape -->\r\n",
       "<g id=\"edge1\" class=\"edge\"><title>block5_pool/MaxPool&#45;&gt;flatten/Reshape</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-560.822C42,-550.19 42,-536.306 42,-524.204\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-524.153 42,-514.153 38.5001,-524.153 45.5001,-524.153\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-535\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x1x1x512</text>\r\n",
       "</g>\r\n",
       "<!-- 11167034092546913880 -->\r\n",
       "<g id=\"node12\" class=\"node\"><title>11167034092546913880</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"79,-431 5,-431 5,-395 79,-395 79,-431\"/>\r\n",
       "<text text-anchor=\"start\" x=\"13\" y=\"-410\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Linear &gt; Relu</text>\r\n",
       "</g>\r\n",
       "<!-- flatten/Reshape&#45;&gt;11167034092546913880 -->\r\n",
       "<g id=\"edge4\" class=\"edge\"><title>flatten/Reshape&#45;&gt;11167034092546913880</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-477.822C42,-467.19 42,-453.306 42,-441.204\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-441.153 42,-431.153 38.5001,-441.153 45.5001,-441.153\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"54.5\" y=\"-452\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x512</text>\r\n",
       "</g>\r\n",
       "<!-- dense_2/MatMul -->\r\n",
       "<g id=\"node8\" class=\"node\"><title>dense_2/MatMul</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"69,-119 15,-119 15,-83 69,-83 69,-119\"/>\r\n",
       "<text text-anchor=\"start\" x=\"29\" y=\"-98\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Linear</text>\r\n",
       "</g>\r\n",
       "<!-- dense_2/Softmax -->\r\n",
       "<g id=\"node9\" class=\"node\"><title>dense_2/Softmax</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"69,-36 15,-36 15,-0 69,-0 69,-36\"/>\r\n",
       "<text text-anchor=\"start\" x=\"24\" y=\"-15\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Softmax</text>\r\n",
       "</g>\r\n",
       "<!-- dense_2/MatMul&#45;&gt;dense_2/Softmax -->\r\n",
       "<g id=\"edge2\" class=\"edge\"><title>dense_2/MatMul&#45;&gt;dense_2/Softmax</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-82.822C42,-72.1903 42,-58.306 42,-46.2035\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-46.1532 42,-36.1533 38.5001,-46.1533 45.5001,-46.1532\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"52\" y=\"-57\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x10</text>\r\n",
       "</g>\r\n",
       "<!-- dropout -->\r\n",
       "<g id=\"node10\" class=\"node\"><title>dropout</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"69,-348 15,-348 15,-312 69,-312 69,-348\"/>\r\n",
       "<text text-anchor=\"start\" x=\"25\" y=\"-327\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Dropout</text>\r\n",
       "</g>\r\n",
       "<!-- 3834672188120457531 -->\r\n",
       "<g id=\"node13\" class=\"node\"><title>3834672188120457531</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"79,-275 5,-275 5,-239 79,-239 79,-275\"/>\r\n",
       "<text text-anchor=\"start\" x=\"13\" y=\"-254\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Linear &gt; Relu</text>\r\n",
       "</g>\r\n",
       "<!-- dropout&#45;&gt;3834672188120457531 -->\r\n",
       "<g id=\"edge6\" class=\"edge\"><title>dropout&#45;&gt;3834672188120457531</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-311.813C42,-303.789 42,-294.047 42,-285.069\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-285.029 42,-275.029 38.5001,-285.029 45.5001,-285.029\"/>\r\n",
       "</g>\r\n",
       "<!-- dropout_1 -->\r\n",
       "<g id=\"node11\" class=\"node\"><title>dropout_1</title>\r\n",
       "<polygon fill=\"#e8e8e8\" stroke=\"#000000\" points=\"69,-192 15,-192 15,-156 69,-156 69,-192\"/>\r\n",
       "<text text-anchor=\"start\" x=\"25\" y=\"-171\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">Dropout</text>\r\n",
       "</g>\r\n",
       "<!-- dropout_1&#45;&gt;dense_2/MatMul -->\r\n",
       "<g id=\"edge3\" class=\"edge\"><title>dropout_1&#45;&gt;dense_2/MatMul</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-155.813C42,-147.789 42,-138.047 42,-129.069\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-129.029 42,-119.029 38.5001,-129.029 45.5001,-129.029\"/>\r\n",
       "</g>\r\n",
       "<!-- 11167034092546913880&#45;&gt;dropout -->\r\n",
       "<g id=\"edge5\" class=\"edge\"><title>11167034092546913880&#45;&gt;dropout</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-394.822C42,-384.19 42,-370.306 42,-358.204\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-358.153 42,-348.153 38.5001,-358.153 45.5001,-358.153\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"54.5\" y=\"-369\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x512</text>\r\n",
       "</g>\r\n",
       "<!-- 3834672188120457531&#45;&gt;dropout_1 -->\r\n",
       "<g id=\"edge7\" class=\"edge\"><title>3834672188120457531&#45;&gt;dropout_1</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-238.822C42,-228.19 42,-214.306 42,-202.204\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-202.153 42,-192.153 38.5001,-202.153 45.5001,-202.153\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"54.5\" y=\"-213\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x512</text>\r\n",
       "</g>\r\n",
       "<!-- 7680877476389183126&#45;&gt;block1_pool/MaxPool -->\r\n",
       "<g id=\"edge9\" class=\"edge\"><title>7680877476389183126&#45;&gt;block1_pool/MaxPool</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-1339.99C42,-1328.92 42,-1315.22 42,-1303.34\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-1303.06 42,-1293.06 38.5001,-1303.06 45.5001,-1303.06\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"67\" y=\"-1314\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x48x48x64</text>\r\n",
       "</g>\r\n",
       "<!-- 17434212678539375187&#45;&gt;block2_pool/MaxPool -->\r\n",
       "<g id=\"edge11\" class=\"edge\"><title>17434212678539375187&#45;&gt;block2_pool/MaxPool</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-1165.99C42,-1154.92 42,-1141.22 42,-1129.34\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-1129.06 42,-1119.06 38.5001,-1129.06 45.5001,-1129.06\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"69.5\" y=\"-1140\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x24x24x128</text>\r\n",
       "</g>\r\n",
       "<!-- 7024157625594113643&#45;&gt;block3_pool/MaxPool -->\r\n",
       "<g id=\"edge13\" class=\"edge\"><title>7024157625594113643&#45;&gt;block3_pool/MaxPool</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-991.989C42,-980.923 42,-967.219 42,-955.336\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-955.062 42,-945.062 38.5001,-955.062 45.5001,-955.062\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"69.5\" y=\"-966\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x12x12x256</text>\r\n",
       "</g>\r\n",
       "<!-- 8051966429743015524&#45;&gt;block4_pool/MaxPool -->\r\n",
       "<g id=\"edge15\" class=\"edge\"><title>8051966429743015524&#45;&gt;block4_pool/MaxPool</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-817.989C42,-806.923 42,-793.219 42,-781.336\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-781.062 42,-771.062 38.5001,-781.062 45.5001,-781.062\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-792\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x6x6x512</text>\r\n",
       "</g>\r\n",
       "<!-- 6723214227147554364&#45;&gt;block5_pool/MaxPool -->\r\n",
       "<g id=\"edge17\" class=\"edge\"><title>6723214227147554364&#45;&gt;block5_pool/MaxPool</title>\r\n",
       "<path fill=\"none\" stroke=\"#000000\" d=\"M42,-643.989C42,-632.923 42,-619.219 42,-607.336\"/>\r\n",
       "<polygon fill=\"#000000\" stroke=\"#000000\" points=\"45.5001,-607.062 42,-597.062 38.5001,-607.062 45.5001,-607.062\"/>\r\n",
       "<text text-anchor=\"middle\" x=\"64.5\" y=\"-618\" font-family=\"Times\" font-size=\"10.00\" fill=\"#000000\">?x3x3x512</text>\r\n",
       "</g>\r\n",
       "</g>\r\n",
       "</svg>\r\n"
      ],
      "text/plain": [
       "<hiddenlayer.graph.Graph at 0x285f2b234a8>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Visual check of the \"after\" model architecture using HiddenLayer\n",
    "hl.build_graph(K.get_session().graph)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Subclass Canvas and add draw_confusion_matrix()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ConfusionMatrixCanvas(hl.Canvas):\n",
    "    \"\"\"Extending HiddenLayer's Canvas to plot a confusion matrix.\"\"\"\n",
    "\n",
    "    def __init__(self, classes):\n",
    "        super().__init__()\n",
    "        self.classes = range(0, classes)\n",
    "    \n",
    "    def draw_confusion_matrix(self, metric):\n",
    "        \"\"\"This function plots the confusion matrix.\n",
    "        Ref:\n",
    "            https://scikit-learn.org/stable/auto_examples/model_selection/plot_confusion_matrix.html\n",
    "        \"\"\"\n",
    "        # Get confusion matrix of the last step\n",
    "        cm = metric.data[-1]\n",
    "\n",
    "        # Build the plot\n",
    "        plt.imshow(cm, interpolation='nearest', cmap=plt.cm.Blues)\n",
    "        tick_marks = np.arange(len(self.classes))\n",
    "        plt.xticks(tick_marks, self.classes, rotation=45)\n",
    "        plt.yticks(tick_marks, self.classes)\n",
    "\n",
    "        thresh = cm.max() / 2.\n",
    "        for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "            plt.text(j, i, format(cm[i, j], 'd'),\n",
    "                     horizontalalignment=\"center\",\n",
    "                     color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "        plt.ylabel('True label')\n",
    "        plt.xlabel('Predicted label')\n",
    "        plt.tight_layout()\n",
    "        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create a Keras callback to compute and visualize a confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ConfusionMatrixCallback(Callback):\n",
    "    \"\"\"Extending Keras' Callback to compute a confusion matrix.\"\"\"\n",
    "    \n",
    "    def __init__(self, x_val, y_val, y_classes):\n",
    "        # Keep a reference to the validation data\n",
    "        self.x_val = x_val\n",
    "        self.y_val = [np.argmax(y) for y in y_val]\n",
    "        \n",
    "        # Use a HiddenLayer History object to store metrics\n",
    "        self.history = hl.History()\n",
    "\n",
    "        # Use our custom HiddenLayer Canvas object to draw the metrics\n",
    "        self.canvas = ConfusionMatrixCanvas(y_classes)        \n",
    "        \n",
    "    def on_train_begin(self, logs={}):\n",
    "        self.epoch = 0\n",
    "              \n",
    "    def on_epoch_end(self, batch, logs={}):\n",
    "        # Generate predictions for validation data\n",
    "        y_val_hat = self.model.predict(self.x_val)\n",
    "        y_val_hat = [np.argmax(y) for y in y_val_hat]\n",
    "              \n",
    "        # Compute the accuracy and generate a confusion matrix using the validation data\n",
    "        val_acc = accuracy_score(self.y_val, y_val_hat)\n",
    "        cm = confusion_matrix(self.y_val, y_val_hat)\n",
    "              \n",
    "        # Store the metrics in the history object\n",
    "        self.history.log(self.epoch, acc=logs.get('acc'), val_acc=val_acc, cm=cm)\n",
    "        \n",
    "        # Plot the metrics\n",
    "        with self.canvas:\n",
    "            self.canvas.draw_confusion_matrix(self.history[\"cm\"])\n",
    "            self.canvas.draw_plot([self.history[\"acc\"], self.history[\"val_acc\"]])\n",
    "              \n",
    "        # Onto the next epoch...\n",
    "        self.epoch += 1\n",
    "\n",
    "callbacks = [ConfusionMatrixCallback(x_val, y_val, y_classes)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compile the model so we'll only train the classifier head"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Since we only want to use the convnet as a feature extractor, \n",
    "# freeze the convnet layers and only train the classifier head\n",
    "for layer in convnet.layers:\n",
    "    layer.trainable = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile the model using the Adam optimizer\n",
    "model.compile(loss='categorical_crossentropy', optimizer=Adam(1e-3), metrics=['acc'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Train the model\n",
    "model.fit(x_train, y_train, batch_size=128, epochs=20, verbose=0, callbacks=callbacks);"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}