computer_vision/05-custom-convnets.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "d5cd5acf",
   "metadata": {
    "papermill": {
     "duration": 0.014729,
     "end_time": "2022-03-16T21:46:43.837157",
     "exception": false,
     "start_time": "2022-03-16T21:46:43.822428",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "**This notebook is an exercise in the [Computer Vision](https://www.kaggle.com/learn/computer-vision) course.  You can reference the tutorial at [this link](https://www.kaggle.com/ryanholbrook/custom-convnets).**\n",
    "\n",
    "---\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab515f8b",
   "metadata": {
    "papermill": {
     "duration": 0.013513,
     "end_time": "2022-03-16T21:46:43.865657",
     "exception": false,
     "start_time": "2022-03-16T21:46:43.852144",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Introduction #\n",
    "\n",
    "In these exercises, you'll build a custom convnet with performance competitive to the VGG16 model from Lesson 1.\n",
    "\n",
    "Get started by running the code cell below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e06726fe",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-03-16T21:46:43.896939Z",
     "iopub.status.busy": "2022-03-16T21:46:43.895410Z",
     "iopub.status.idle": "2022-03-16T21:46:59.924602Z",
     "shell.execute_reply": "2022-03-16T21:46:59.925240Z",
     "shell.execute_reply.started": "2022-03-16T21:33:40.917813Z"
    },
    "papermill": {
     "duration": 16.047159,
     "end_time": "2022-03-16T21:46:59.925692",
     "exception": false,
     "start_time": "2022-03-16T21:46:43.878533",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 5117 files belonging to 2 classes.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-03-16 21:46:53.543923: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-03-16 21:46:53.633549: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-03-16 21:46:53.634214: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-03-16 21:46:53.636639: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 AVX512F FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2022-03-16 21:46:53.637440: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-03-16 21:46:53.638091: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-03-16 21:46:53.638975: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-03-16 21:46:55.476216: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-03-16 21:46:55.477070: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-03-16 21:46:55.477734: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
      "2022-03-16 21:46:55.478300: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 15403 MB memory:  -> device: 0, name: Tesla P100-PCIE-16GB, pci bus id: 0000:00:04.0, compute capability: 6.0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 5051 files belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "# Setup feedback system\n",
    "from learntools.core import binder\n",
    "binder.bind(globals())\n",
    "from learntools.computer_vision.ex5 import *\n",
    "\n",
    "# Imports\n",
    "import os, warnings\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import gridspec\n",
    "\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.preprocessing import image_dataset_from_directory\n",
    "\n",
    "# Reproducability\n",
    "def set_seed(seed=31415):\n",
    "    np.random.seed(seed)\n",
    "    tf.random.set_seed(seed)\n",
    "    os.environ['PYTHONHASHSEED'] = str(seed)\n",
    "    os.environ['TF_DETERMINISTIC_OPS'] = '1'\n",
    "set_seed()\n",
    "\n",
    "# Set Matplotlib defaults\n",
    "plt.rc('figure', autolayout=True)\n",
    "plt.rc('axes', labelweight='bold', labelsize='large',\n",
    "       titleweight='bold', titlesize=18, titlepad=10)\n",
    "plt.rc('image', cmap='magma')\n",
    "warnings.filterwarnings(\"ignore\") # to clean up output cells\n",
    "\n",
    "\n",
    "# Load training and validation sets\n",
    "ds_train_ = image_dataset_from_directory(\n",
    "    '../input/car-or-truck/train',\n",
    "    labels='inferred',\n",
    "    label_mode='binary',\n",
    "    image_size=[128, 128],\n",
    "    interpolation='nearest',\n",
    "    batch_size=64,\n",
    "    shuffle=True,\n",
    ")\n",
    "ds_valid_ = image_dataset_from_directory(\n",
    "    '../input/car-or-truck/valid',\n",
    "    labels='inferred',\n",
    "    label_mode='binary',\n",
    "    image_size=[128, 128],\n",
    "    interpolation='nearest',\n",
    "    batch_size=64,\n",
    "    shuffle=False,\n",
    ")\n",
    "\n",
    "# Data Pipeline\n",
    "def convert_to_float(image, label):\n",
    "    image = tf.image.convert_image_dtype(image, dtype=tf.float32)\n",
    "    return image, label\n",
    "\n",
    "AUTOTUNE = tf.data.experimental.AUTOTUNE\n",
    "ds_train = (\n",
    "    ds_train_\n",
    "    .map(convert_to_float)\n",
    "    .cache()\n",
    "    .prefetch(buffer_size=AUTOTUNE)\n",
    ")\n",
    "ds_valid = (\n",
    "    ds_valid_\n",
    "    .map(convert_to_float)\n",
    "    .cache()\n",
    "    .prefetch(buffer_size=AUTOTUNE)\n",
    ")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f38af0a",
   "metadata": {
    "papermill": {
     "duration": 0.014433,
     "end_time": "2022-03-16T21:46:59.955527",
     "exception": false,
     "start_time": "2022-03-16T21:46:59.941094",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Design a Convnet #\n",
    "\n",
    "Let's design a convolutional network with a block architecture like we saw in the tutorial. The model from the example had three blocks, each with a single convolutional layer. Its performance on the \"Car or Truck\" problem was okay, but far from what the pretrained VGG16 could achieve. It might be that our simple network lacks the ability to extract sufficiently complex features. We could try improving the model either by adding more blocks or by adding convolutions to the blocks we have.\n",
    "\n",
    "Let's go with the second approach. We'll keep the three block structure, but increase the number of `Conv2D` layer in the second block to two, and in the third block to three.\n",
    "\n",
    "<figure>\n",
    "<!-- <img src=\"./images/2-convmodel-2.png\" width=\"250\" alt=\"Diagram of a convolutional model.\"> -->\n",
    "<img src=\"https://i.imgur.com/Vko6nCK.png\" width=\"250\" alt=\"Diagram of a convolutional model.\">\n",
    "</figure>\n",
    "\n",
    "# 1) Define Model #\n",
    "\n",
    "Given the diagram above, complete the model by defining the layers of the third block."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "07e01183",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-03-16T21:46:59.991214Z",
     "iopub.status.busy": "2022-03-16T21:46:59.990704Z",
     "iopub.status.idle": "2022-03-16T21:47:00.093221Z",
     "shell.execute_reply": "2022-03-16T21:47:00.092794Z",
     "shell.execute_reply.started": "2022-03-16T21:35:28.999757Z"
    },
    "lines_to_next_cell": 0,
    "papermill": {
     "duration": 0.123658,
     "end_time": "2022-03-16T21:47:00.093328",
     "exception": false,
     "start_time": "2022-03-16T21:46:59.969670",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "parent.postMessage({\"jupyterEvent\": \"custom.exercise_interaction\", \"data\": {\"outcomeType\": 1, \"valueTowardsCompletion\": 0.3333333333333333, \"interactionType\": 1, \"questionType\": 2, \"questionId\": \"1_Q1\", \"learnToolsVersion\": \"0.3.4\", \"failureMessage\": \"\", \"exceptionClass\": \"\", \"trace\": \"\"}}, \"*\")"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "<span style=\"color:#33cc33\">Correct</span>"
      ],
      "text/plain": [
       "Correct"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers\n",
    "\n",
    "model = keras.Sequential([\n",
    "    # Block One\n",
    "    layers.Conv2D(filters=32, kernel_size=3, activation='relu', padding='same',\n",
    "                  input_shape=[128, 128, 3]),\n",
    "    layers.MaxPool2D(),\n",
    "\n",
    "    # Block Two\n",
    "    layers.Conv2D(filters=64, kernel_size=3, activation='relu', padding='same'),\n",
    "    layers.MaxPool2D(),\n",
    "\n",
    "    # Block Three\n",
    "    # YOUR CODE HERE\n",
    "    layers.Conv2D(filters=128, kernel_size=3, activation='relu', padding='same'),\n",
    "    layers.Conv2D(filters=128, kernel_size=3, activation='relu', padding='same'),\n",
    "    layers.MaxPool2D(),\n",
    "\n",
    "    # Head\n",
    "    layers.Flatten(),\n",
    "    layers.Dense(6, activation='relu'),\n",
    "    layers.Dropout(0.2),\n",
    "    layers.Dense(1, activation='sigmoid'),\n",
    "])\n",
    "\n",
    "# Check your answer\n",
    "q_1.check()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "4efcdc81",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-03-16T21:47:00.127902Z",
     "iopub.status.busy": "2022-03-16T21:47:00.127190Z",
     "iopub.status.idle": "2022-03-16T21:47:00.129620Z",
     "shell.execute_reply": "2022-03-16T21:47:00.129194Z",
     "shell.execute_reply.started": "2022-03-16T21:35:31.735888Z"
    },
    "papermill": {
     "duration": 0.020696,
     "end_time": "2022-03-16T21:47:00.129738",
     "exception": false,
     "start_time": "2022-03-16T21:47:00.109042",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Lines below will give you a hint or solution code\n",
    "#q_1.hint()\n",
    "#q_1.solution()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "17b8f832",
   "metadata": {
    "papermill": {
     "duration": 0.015406,
     "end_time": "2022-03-16T21:47:00.160222",
     "exception": false,
     "start_time": "2022-03-16T21:47:00.144816",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# 2) Compile #\n",
    "\n",
    "To prepare for training, compile the model with an appropriate loss and accuracy metric for the \"Car or Truck\" dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6058ff8a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-03-16T21:47:00.195502Z",
     "iopub.status.busy": "2022-03-16T21:47:00.194707Z",
     "iopub.status.idle": "2022-03-16T21:47:00.207742Z",
     "shell.execute_reply": "2022-03-16T21:47:00.208141Z",
     "shell.execute_reply.started": "2022-03-16T21:36:32.651815Z"
    },
    "lines_to_next_cell": 0,
    "papermill": {
     "duration": 0.032982,
     "end_time": "2022-03-16T21:47:00.208256",
     "exception": false,
     "start_time": "2022-03-16T21:47:00.175274",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "parent.postMessage({\"jupyterEvent\": \"custom.exercise_interaction\", \"data\": {\"outcomeType\": 1, \"valueTowardsCompletion\": 0.3333333333333333, \"interactionType\": 1, \"questionType\": 2, \"questionId\": \"2_Q2\", \"learnToolsVersion\": \"0.3.4\", \"failureMessage\": \"\", \"exceptionClass\": \"\", \"trace\": \"\"}}, \"*\")"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "<span style=\"color:#33cc33\">Correct</span>"
      ],
      "text/plain": [
       "Correct"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.compile(\n",
    "    optimizer=tf.keras.optimizers.Adam(epsilon=0.01),\n",
    "    # YOUR CODE HERE: Add loss and metric\n",
    "    loss='binary_crossentropy',\n",
    "    metrics=['binary_accuracy']\n",
    ")\n",
    "\n",
    "# Check your answer\n",
    "q_2.check()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "b34d1da9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-03-16T21:47:00.247626Z",
     "iopub.status.busy": "2022-03-16T21:47:00.246839Z",
     "iopub.status.idle": "2022-03-16T21:47:00.254438Z",
     "shell.execute_reply": "2022-03-16T21:47:00.254884Z",
     "shell.execute_reply.started": "2022-03-16T21:36:40.106316Z"
    },
    "lines_to_next_cell": 0,
    "papermill": {
     "duration": 0.030334,
     "end_time": "2022-03-16T21:47:00.254999",
     "exception": false,
     "start_time": "2022-03-16T21:47:00.224665",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "parent.postMessage({\"jupyterEvent\": \"custom.exercise_interaction\", \"data\": {\"outcomeType\": 1, \"valueTowardsCompletion\": 0.3333333333333333, \"interactionType\": 1, \"questionType\": 2, \"questionId\": \"2_Q2\", \"learnToolsVersion\": \"0.3.4\", \"failureMessage\": \"\", \"exceptionClass\": \"\", \"trace\": \"\"}}, \"*\")"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "<span style=\"color:#33cc33\">Correct</span>"
      ],
      "text/plain": [
       "Correct"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.compile(\n",
    "    optimizer=tf.keras.optimizers.Adam(epsilon=0.01),\n",
    "    loss='binary_crossentropy',\n",
    "    metrics=['binary_accuracy'],\n",
    ")\n",
    "q_2.assert_check_passed()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "215b9400",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-03-16T21:47:00.294562Z",
     "iopub.status.busy": "2022-03-16T21:47:00.294024Z",
     "iopub.status.idle": "2022-03-16T21:47:00.297520Z",
     "shell.execute_reply": "2022-03-16T21:47:00.297077Z",
     "shell.execute_reply.started": "2022-03-16T21:36:46.356832Z"
    },
    "lines_to_next_cell": 0,
    "papermill": {
     "duration": 0.025187,
     "end_time": "2022-03-16T21:47:00.297627",
     "exception": false,
     "start_time": "2022-03-16T21:47:00.272440",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Lines below will give you a hint or solution code\n",
    "#q_2.hint()\n",
    "#q_2.solution()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "75328a8f",
   "metadata": {
    "papermill": {
     "duration": 0.01724,
     "end_time": "2022-03-16T21:47:00.332022",
     "exception": false,
     "start_time": "2022-03-16T21:47:00.314782",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Finally, let's test the performance of this new model. First run this cell to fit the model to the training set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "67a54fdf",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-03-16T21:47:00.371169Z",
     "iopub.status.busy": "2022-03-16T21:47:00.370395Z",
     "iopub.status.idle": "2022-03-16T21:51:11.158812Z",
     "shell.execute_reply": "2022-03-16T21:51:11.159272Z",
     "shell.execute_reply.started": "2022-03-16T21:37:06.737307Z"
    },
    "papermill": {
     "duration": 250.809998,
     "end_time": "2022-03-16T21:51:11.159503",
     "exception": false,
     "start_time": "2022-03-16T21:47:00.349505",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-03-16 21:47:00.405702: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-03-16 21:47:03.761997: I tensorflow/stream_executor/cuda/cuda_dnn.cc:369] Loaded cuDNN version 8005\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "80/80 [==============================] - 38s 363ms/step - loss: 0.6816 - binary_accuracy: 0.5744 - val_loss: 0.6691 - val_binary_accuracy: 0.5785\n",
      "Epoch 2/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.6667 - binary_accuracy: 0.5787 - val_loss: 0.6601 - val_binary_accuracy: 0.5785\n",
      "Epoch 3/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.6613 - binary_accuracy: 0.5787 - val_loss: 0.6503 - val_binary_accuracy: 0.5785\n",
      "Epoch 4/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.6517 - binary_accuracy: 0.5787 - val_loss: 0.6447 - val_binary_accuracy: 0.5785\n",
      "Epoch 5/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.6456 - binary_accuracy: 0.6093 - val_loss: 0.6386 - val_binary_accuracy: 0.5961\n",
      "Epoch 6/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.6372 - binary_accuracy: 0.6330 - val_loss: 0.6354 - val_binary_accuracy: 0.5918\n",
      "Epoch 7/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.6354 - binary_accuracy: 0.6334 - val_loss: 0.6317 - val_binary_accuracy: 0.6129\n",
      "Epoch 8/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.6256 - binary_accuracy: 0.6529 - val_loss: 0.6207 - val_binary_accuracy: 0.6434\n",
      "Epoch 9/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.6205 - binary_accuracy: 0.6635 - val_loss: 0.6137 - val_binary_accuracy: 0.6498\n",
      "Epoch 10/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.6141 - binary_accuracy: 0.6729 - val_loss: 0.6074 - val_binary_accuracy: 0.6680\n",
      "Epoch 11/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.6039 - binary_accuracy: 0.6846 - val_loss: 0.6022 - val_binary_accuracy: 0.6729\n",
      "Epoch 12/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.5969 - binary_accuracy: 0.6998 - val_loss: 0.5973 - val_binary_accuracy: 0.6803\n",
      "Epoch 13/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.5831 - binary_accuracy: 0.7090 - val_loss: 0.5880 - val_binary_accuracy: 0.6967\n",
      "Epoch 14/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.5737 - binary_accuracy: 0.7211 - val_loss: 0.5825 - val_binary_accuracy: 0.7104\n",
      "Epoch 15/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.5616 - binary_accuracy: 0.7299 - val_loss: 0.5712 - val_binary_accuracy: 0.7169\n",
      "Epoch 16/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.5451 - binary_accuracy: 0.7497 - val_loss: 0.5718 - val_binary_accuracy: 0.7090\n",
      "Epoch 17/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.5268 - binary_accuracy: 0.7620 - val_loss: 0.5551 - val_binary_accuracy: 0.7363\n",
      "Epoch 18/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.5020 - binary_accuracy: 0.7880 - val_loss: 0.5333 - val_binary_accuracy: 0.7545\n",
      "Epoch 19/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.4870 - binary_accuracy: 0.7927 - val_loss: 0.5456 - val_binary_accuracy: 0.7355\n",
      "Epoch 20/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.4701 - binary_accuracy: 0.7983 - val_loss: 0.4908 - val_binary_accuracy: 0.7640\n",
      "Epoch 21/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.4308 - binary_accuracy: 0.8085 - val_loss: 0.4861 - val_binary_accuracy: 0.7660\n",
      "Epoch 22/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.3956 - binary_accuracy: 0.8317 - val_loss: 0.4404 - val_binary_accuracy: 0.7989\n",
      "Epoch 23/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.3638 - binary_accuracy: 0.8450 - val_loss: 0.4566 - val_binary_accuracy: 0.7806\n",
      "Epoch 24/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.3382 - binary_accuracy: 0.8607 - val_loss: 0.4534 - val_binary_accuracy: 0.7870\n",
      "Epoch 25/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.3169 - binary_accuracy: 0.8708 - val_loss: 0.4250 - val_binary_accuracy: 0.8127\n",
      "Epoch 26/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.2885 - binary_accuracy: 0.8849 - val_loss: 0.4289 - val_binary_accuracy: 0.8165\n",
      "Epoch 27/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.2668 - binary_accuracy: 0.8904 - val_loss: 0.4419 - val_binary_accuracy: 0.8133\n",
      "Epoch 28/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.2412 - binary_accuracy: 0.9025 - val_loss: 0.4628 - val_binary_accuracy: 0.8070\n",
      "Epoch 29/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.2512 - binary_accuracy: 0.8972 - val_loss: 0.4338 - val_binary_accuracy: 0.7985\n",
      "Epoch 30/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.2250 - binary_accuracy: 0.9150 - val_loss: 0.4218 - val_binary_accuracy: 0.8202\n",
      "Epoch 31/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.2128 - binary_accuracy: 0.9191 - val_loss: 0.4465 - val_binary_accuracy: 0.8377\n",
      "Epoch 32/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.1784 - binary_accuracy: 0.9361 - val_loss: 0.4509 - val_binary_accuracy: 0.8369\n",
      "Epoch 33/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.2013 - binary_accuracy: 0.9234 - val_loss: 0.4551 - val_binary_accuracy: 0.8460\n",
      "Epoch 34/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.1800 - binary_accuracy: 0.9349 - val_loss: 0.4936 - val_binary_accuracy: 0.8414\n",
      "Epoch 35/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.1654 - binary_accuracy: 0.9414 - val_loss: 0.6184 - val_binary_accuracy: 0.8258\n",
      "Epoch 36/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.1439 - binary_accuracy: 0.9478 - val_loss: 0.6884 - val_binary_accuracy: 0.8060\n",
      "Epoch 37/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.1291 - binary_accuracy: 0.9580 - val_loss: 0.7367 - val_binary_accuracy: 0.8155\n",
      "Epoch 38/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.1257 - binary_accuracy: 0.9582 - val_loss: 0.7216 - val_binary_accuracy: 0.8175\n",
      "Epoch 39/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.1313 - binary_accuracy: 0.9541 - val_loss: 0.5901 - val_binary_accuracy: 0.8282\n",
      "Epoch 40/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.1164 - binary_accuracy: 0.9580 - val_loss: 0.7537 - val_binary_accuracy: 0.8167\n",
      "Epoch 41/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.1020 - binary_accuracy: 0.9699 - val_loss: 0.7372 - val_binary_accuracy: 0.8293\n",
      "Epoch 42/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.0985 - binary_accuracy: 0.9687 - val_loss: 0.5475 - val_binary_accuracy: 0.8470\n",
      "Epoch 43/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.1001 - binary_accuracy: 0.9679 - val_loss: 0.6405 - val_binary_accuracy: 0.8386\n",
      "Epoch 44/50\n",
      "80/80 [==============================] - 4s 46ms/step - loss: 0.1064 - binary_accuracy: 0.9613 - val_loss: 0.5989 - val_binary_accuracy: 0.8284\n",
      "Epoch 45/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.0996 - binary_accuracy: 0.9695 - val_loss: 0.5213 - val_binary_accuracy: 0.8311\n",
      "Epoch 46/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.0823 - binary_accuracy: 0.9740 - val_loss: 0.6159 - val_binary_accuracy: 0.8052\n",
      "Epoch 47/50\n",
      "80/80 [==============================] - 3s 43ms/step - loss: 0.0737 - binary_accuracy: 0.9773 - val_loss: 0.6454 - val_binary_accuracy: 0.8359\n",
      "Epoch 48/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.0721 - binary_accuracy: 0.9791 - val_loss: 0.6538 - val_binary_accuracy: 0.8491\n",
      "Epoch 49/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.0684 - binary_accuracy: 0.9797 - val_loss: 0.6213 - val_binary_accuracy: 0.8418\n",
      "Epoch 50/50\n",
      "80/80 [==============================] - 3s 42ms/step - loss: 0.0623 - binary_accuracy: 0.9803 - val_loss: 0.7014 - val_binary_accuracy: 0.8517\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(\n",
    "    ds_train,\n",
    "    validation_data=ds_valid,\n",
    "    epochs=50,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "def23a41",
   "metadata": {
    "papermill": {
     "duration": 0.658843,
     "end_time": "2022-03-16T21:51:12.473439",
     "exception": false,
     "start_time": "2022-03-16T21:51:11.814596",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "And now run the cell below to plot the loss and metric curves for this training run."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "fa307a47",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-03-16T21:51:13.828148Z",
     "iopub.status.busy": "2022-03-16T21:51:13.827473Z",
     "iopub.status.idle": "2022-03-16T21:51:14.347226Z",
     "shell.execute_reply": "2022-03-16T21:51:14.347697Z",
     "shell.execute_reply.started": "2022-03-16T21:44:35.111315Z"
    },
    "papermill": {
     "duration": 1.204267,
     "end_time": "2022-03-16T21:51:14.347843",
     "exception": false,
     "start_time": "2022-03-16T21:51:13.143576",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "history_frame = pd.DataFrame(history.history)\n",
    "history_frame.loc[:, ['loss', 'val_loss']].plot()\n",
    "history_frame.loc[:, ['binary_accuracy', 'val_binary_accuracy']].plot();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a3beb38b",
   "metadata": {
    "papermill": {
     "duration": 0.66007,
     "end_time": "2022-03-16T21:51:15.666943",
     "exception": false,
     "start_time": "2022-03-16T21:51:15.006873",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# 3) Train the Model #\n",
    "\n",
    "How would you interpret these training curves? Did this model improve upon the model from the tutorial?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "a518b7f2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-03-16T21:51:17.488324Z",
     "iopub.status.busy": "2022-03-16T21:51:17.487497Z",
     "iopub.status.idle": "2022-03-16T21:51:17.492008Z",
     "shell.execute_reply": "2022-03-16T21:51:17.492395Z",
     "shell.execute_reply.started": "2022-03-16T21:44:52.257569Z"
    },
    "lines_to_next_cell": 0,
    "papermill": {
     "duration": 1.054019,
     "end_time": "2022-03-16T21:51:17.492566",
     "exception": false,
     "start_time": "2022-03-16T21:51:16.438547",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "parent.postMessage({\"jupyterEvent\": \"custom.exercise_interaction\", \"data\": {\"outcomeType\": 1, \"valueTowardsCompletion\": 0.3333333333333333, \"interactionType\": 1, \"questionType\": 4, \"questionId\": \"3_Q3\", \"learnToolsVersion\": \"0.3.4\", \"failureMessage\": \"\", \"exceptionClass\": \"\", \"trace\": \"\"}}, \"*\")"
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/markdown": [
       "<span style=\"color:#33cc33\">Correct:</span> \n",
       "\n",
       "\n",
       "The learning curves for the model from the tutorial diverged fairly rapidly. This would indicate that it was prone to overfitting and in need of some regularization. The additional layer in our new model would make it even more prone to overfitting. However, adding some regularization with the `Dropout` layer helped prevent this. These changes improved the validation accuracy of the model by several points.\n"
      ],
      "text/plain": [
       "Correct: \n",
       "\n",
       "\n",
       "The learning curves for the model from the tutorial diverged fairly rapidly. This would indicate that it was prone to overfitting and in need of some regularization. The additional layer in our new model would make it even more prone to overfitting. However, adding some regularization with the `Dropout` layer helped prevent this. These changes improved the validation accuracy of the model by several points."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# View the solution (Run this code cell to receive credit!)\n",
    "q_3.check()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "855aba22",
   "metadata": {
    "papermill": {
     "duration": 0.669993,
     "end_time": "2022-03-16T21:51:18.827709",
     "exception": false,
     "start_time": "2022-03-16T21:51:18.157716",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Conclusion #\n",
    "\n",
    "These exercises showed you how to design a custom convolutional network to solve a specific classification problem. Though most models these days will be built on top of a pretrained base, it certain circumstances a smaller custom convnet might still be preferable -- such as with a smaller or unusual dataset or when computing resources are very limited. As you saw here, for certain problems they can perform just as well as a pretrained model.\n",
    "\n",
    "# Keep Going #\n",
    "\n",
    "Continue on to [**Lesson 6**](https://www.kaggle.com/ryanholbrook/data-augmentation), where you'll learn a widely-used technique that can give a boost to your training data: **data augmentation**."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "270992d5",
   "metadata": {
    "papermill": {
     "duration": 0.923831,
     "end_time": "2022-03-16T21:51:20.418169",
     "exception": false,
     "start_time": "2022-03-16T21:51:19.494338",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "---\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "*Have questions or comments? Visit the [course discussion forum](https://www.kaggle.com/learn/computer-vision/discussion) to chat with other learners.*"
   ]
  }
 ],
 "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": "3.7.12"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 288.618241,
   "end_time": "2022-03-16T21:51:24.126440",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2022-03-16T21:46:35.508199",
   "version": "2.3.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}