ann/keras.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Introduction to Keras\n",
    "\n",
    "<img align=\"center\" src=\"https://github.com/dfbarrero/dataCourse/raw/master/ann/figs/keras.png\" width=\"300\">\n",
    "\n",
    "This notebook is based on the Chapter 10 *Introduction to Artificial Neural Networks with Keras*, Hands-On Machine Learning with Scikit-Learn, Keras & TensorFlow. concepts, Tools, and Techiniques to Build Intelligent Systems, by Aurélien Géron. Ed. O'Reilly. \n",
    "\n",
    "A more complete notebook based on the same book can be found [here](https://github.com/karthik9001/Hands-On-Machine-Learning-with-Scikit-Learn-Keras-and-TensorFlow-2nd-Edition/blob/master/10_neural_nets_with_keras.ipynb)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Importing the library\n",
    "\n",
    "Keras is a Deep Learning API that can use several backends, i.e., the library that really implements the Deep Learning stuff. Examples of DL libraries are TensorFlow, Microsoft Cognitive Toolkit (CNTK) and Theano. The most popular one is TensorFlow, and actually it has Keras integrated in the library."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow import keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'2.11.0'"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'2.11.0'"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "keras.__version__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading the dataset\n",
    "\n",
    "In this example we will use fashion MNIST, which is a widely known image dataset more challenging that MINIST. It contains 70000 greyscale 28x28 images with 10 classes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
      "29515/29515 [==============================] - 0s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
      "26421880/26421880 [==============================] - 1s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
      "5148/5148 [==============================] - 0s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
      "4422102/4422102 [==============================] - 0s 0us/step\n"
     ]
    }
   ],
   "source": [
    "fashion_mnist = keras.datasets.fashion_mnist\n",
    "\n",
    "(X_train_full, y_train_full), (X_test, y_test) = fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We inspect our data in first place."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 28, 28)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_full.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dtype('uint8')"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_full.dtype"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Max: 255\n",
      "Min: 0\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "\n",
    "print(f\"Max: {np.max(X_train_full)}\")\n",
    "print(f\"Min: {np.min(X_train_full)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1500x800 with 50 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "plt.figure(figsize=(15,8))\n",
    "for i in range(50):\n",
    "    plt.subplot(5,10,i+1)\n",
    "    plt.imshow(X_train_full[i].reshape((28,28)),cmap='binary')\n",
    "    plt.axis(\"off\")\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Normalize and split a validation set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_valid, X_train = X_train_full[:5000]/255, X_train_full[5000:] / 255.0\n",
    "y_valid, y_train = y_train_full[:5000], y_train_full[5000:]\n",
    "X_test = X_test / 255.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Coat'"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class_names = [\"T-shirt/top\", \"Trouser\", \"Pullover\", \"Dress\", \"Coat\", \"Sandal\", \"Shirt\", \"Sneaker\", \"Bag\", \"Angle boot\"]\n",
    "\n",
    "class_names[y_train[0]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Creating the model using the Sequential API\n",
    "\n",
    "Keras comes with two APIs:\n",
    "* Sequential: simple and powerful.\n",
    "* Functional: more flexible but less simple."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-14 17:18:13.031965: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:981] 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",
      "2022-12-14 17:18:13.052592: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:981] 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",
      "2022-12-14 17:18:13.052775: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:981] 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",
      "2022-12-14 17:18:13.053347: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 AVX_VNNI FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2022-12-14 17:18:13.054518: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:981] 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",
      "2022-12-14 17:18:13.054652: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:981] 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",
      "2022-12-14 17:18:13.054740: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:981] 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",
      "2022-12-14 17:18:13.550820: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:981] 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",
      "2022-12-14 17:18:13.551325: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:981] 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",
      "2022-12-14 17:18:13.551381: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:981] 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",
      "2022-12-14 17:18:13.551430: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1613] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 5732 MB memory:  -> device: 0, name: NVIDIA GeForce RTX 3070 Ti, pci bus id: 0000:01:00.0, compute capability: 8.6\n"
     ]
    }
   ],
   "source": [
    "model = keras.models.Sequential()\n",
    "\n",
    "model.add(keras.layers.Flatten(input_shape=[28, 28]))\n",
    "model.add(keras.layers.Dense(300, activation=\"relu\"))\n",
    "model.add(keras.layers.Dense(100, activation=\"relu\"))\n",
    "model.add(keras.layers.Dense(10, activation=\"softmax\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observe:\n",
    "- *Flatten layer*, with *input_shape*.\n",
    "- *Dense* layers with relu activation.\n",
    "- The last layer is *Dense* with 10 neurons and softmax activation."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Model can also be defined with a list."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = keras.models.Sequential([\n",
    "    keras.layers.Flatten(input_shape=[28, 28]),\n",
    "    keras.layers.Dense(300, activation=\"relu\"),\n",
    "    keras.layers.Dense(100, activation=\"relu\"),\n",
    "    keras.layers.Dense(10, activation=\"softmax\")\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " flatten_1 (Flatten)         (None, 784)               0         \n",
      "                                                                 \n",
      " dense_3 (Dense)             (None, 300)               235500    \n",
      "                                                                 \n",
      " dense_4 (Dense)             (None, 100)               30100     \n",
      "                                                                 \n",
      " dense_5 (Dense)             (None, 10)                1010      \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 266,610\n",
      "Trainable params: 266,610\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First Dense layer with $235,500 = 784 * 300 + 300$ (bias neurons)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "You must install pydot (`pip install pydot`) and install graphviz (see instructions at https://graphviz.gitlab.io/download/) for plot_model to work.\n"
     ]
    }
   ],
   "source": [
    "# You need pydot and graphviz to run this\n",
    "\n",
    "keras.utils.plot_model(model, \"my_model.png\", show_shapes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<keras.layers.reshaping.flatten.Flatten at 0x7f5c57e05190>,\n",
       " <keras.layers.core.dense.Dense at 0x7f5c59a3bf10>,\n",
       " <keras.layers.core.dense.Dense at 0x7f5c59cc7bd0>,\n",
       " <keras.layers.core.dense.Dense at 0x7f5c582cdd50>]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<keras.layers.core.dense.Dense at 0x7f5c59a3bf10>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hidden1 = model.layers[1]\n",
    "\n",
    "hidden1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.07077837, -0.00283641,  0.06720532, ...,  0.00469161,\n",
       "        -0.04056014,  0.01829311],\n",
       "       [ 0.03221393,  0.06767893,  0.01798223, ...,  0.03763348,\n",
       "         0.02639063,  0.00417218],\n",
       "       [ 0.05584747,  0.05281794, -0.0728555 , ...,  0.02196448,\n",
       "        -0.03511398,  0.07042767],\n",
       "       ...,\n",
       "       [-0.04465934,  0.06900389,  0.05876353, ...,  0.06336838,\n",
       "        -0.06554438, -0.02707247],\n",
       "       [ 0.01439762, -0.03031217,  0.04236083, ...,  0.03631302,\n",
       "         0.05881394,  0.02963777],\n",
       "       [ 0.07124932, -0.02704971,  0.05769567, ..., -0.06873634,\n",
       "         0.04904582, -0.06828022]], dtype=float32)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weights, biases = hidden1.get_weights()\n",
    "\n",
    "weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.], dtype=float32)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "biases"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compiling the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss=\"sparse_categorical_crossentropy\", optimizer=\"sgd\", metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observe:\n",
    "- \"sparse_categorical_crossentropy\" loss function because labels are sparce, i.e., indices.\n",
    "- \"sgd\" optimizer.\n",
    "- \"accuracy\" metric."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4, 0, 7, ..., 3, 0, 5], dtype=uint8)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., ..., 0., 0., 0.],\n",
       "       [1., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 1., 0., 0.],\n",
       "       ...,\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [1., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.]], dtype=float32)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "keras.utils.to_categorical(y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training and evaluating the model\n",
    "\n",
    "And here we go!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-12-14 17:19:26.777307: I tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:630] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.\n",
      "2022-12-14 17:19:26.810181: I tensorflow/compiler/xla/service/service.cc:173] XLA service 0x1d469e70 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:\n",
      "2022-12-14 17:19:26.810195: I tensorflow/compiler/xla/service/service.cc:181]   StreamExecutor device (0): NVIDIA GeForce RTX 3070 Ti, Compute Capability 8.6\n",
      "2022-12-14 17:19:26.940135: I tensorflow/compiler/jit/xla_compilation_cache.cc:477] Compiled cluster using XLA!  This line is logged at most once for the lifetime of the process.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1719/1719 [==============================] - 7s 4ms/step - loss: 0.7198 - accuracy: 0.7586 - val_loss: 0.5828 - val_accuracy: 0.7842\n",
      "Epoch 2/30\n",
      "1719/1719 [==============================] - 6s 3ms/step - loss: 0.4882 - accuracy: 0.8302 - val_loss: 0.4474 - val_accuracy: 0.8480\n",
      "Epoch 3/30\n",
      "1719/1719 [==============================] - 6s 3ms/step - loss: 0.4440 - accuracy: 0.8449 - val_loss: 0.4136 - val_accuracy: 0.8626\n",
      "Epoch 4/30\n",
      "1719/1719 [==============================] - 6s 3ms/step - loss: 0.4162 - accuracy: 0.8540 - val_loss: 0.4133 - val_accuracy: 0.8612\n",
      "Epoch 5/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3959 - accuracy: 0.8604 - val_loss: 0.3837 - val_accuracy: 0.8688\n",
      "Epoch 6/30\n",
      "1719/1719 [==============================] - 6s 4ms/step - loss: 0.3811 - accuracy: 0.8665 - val_loss: 0.3898 - val_accuracy: 0.8620\n",
      "Epoch 7/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3670 - accuracy: 0.8712 - val_loss: 0.3704 - val_accuracy: 0.8706\n",
      "Epoch 8/30\n",
      "1719/1719 [==============================] - 6s 4ms/step - loss: 0.3549 - accuracy: 0.8736 - val_loss: 0.3570 - val_accuracy: 0.8740\n",
      "Epoch 9/30\n",
      "1719/1719 [==============================] - 6s 3ms/step - loss: 0.3449 - accuracy: 0.8776 - val_loss: 0.3588 - val_accuracy: 0.8710\n",
      "Epoch 10/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3365 - accuracy: 0.8821 - val_loss: 0.3438 - val_accuracy: 0.8776\n",
      "Epoch 11/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3274 - accuracy: 0.8839 - val_loss: 0.3599 - val_accuracy: 0.8732\n",
      "Epoch 12/30\n",
      "1719/1719 [==============================] - 4s 3ms/step - loss: 0.3202 - accuracy: 0.8861 - val_loss: 0.3418 - val_accuracy: 0.8802\n",
      "Epoch 13/30\n",
      "1719/1719 [==============================] - 4s 3ms/step - loss: 0.3120 - accuracy: 0.8879 - val_loss: 0.3321 - val_accuracy: 0.8820\n",
      "Epoch 14/30\n",
      "1719/1719 [==============================] - 4s 3ms/step - loss: 0.3055 - accuracy: 0.8902 - val_loss: 0.3266 - val_accuracy: 0.8856\n",
      "Epoch 15/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2991 - accuracy: 0.8927 - val_loss: 0.3262 - val_accuracy: 0.8850\n",
      "Epoch 16/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2926 - accuracy: 0.8956 - val_loss: 0.3199 - val_accuracy: 0.8862\n",
      "Epoch 17/30\n",
      "1719/1719 [==============================] - 6s 3ms/step - loss: 0.2875 - accuracy: 0.8970 - val_loss: 0.3289 - val_accuracy: 0.8826\n",
      "Epoch 18/30\n",
      "1719/1719 [==============================] - 4s 2ms/step - loss: 0.2817 - accuracy: 0.8985 - val_loss: 0.3458 - val_accuracy: 0.8778\n",
      "Epoch 19/30\n",
      "1719/1719 [==============================] - 2s 918us/step - loss: 0.2771 - accuracy: 0.8999 - val_loss: 0.3329 - val_accuracy: 0.8762\n",
      "Epoch 20/30\n",
      "1719/1719 [==============================] - 3s 2ms/step - loss: 0.2715 - accuracy: 0.9017 - val_loss: 0.3135 - val_accuracy: 0.8870\n",
      "Epoch 21/30\n",
      "1719/1719 [==============================] - 7s 4ms/step - loss: 0.2664 - accuracy: 0.9049 - val_loss: 0.3158 - val_accuracy: 0.8868\n",
      "Epoch 22/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2619 - accuracy: 0.9049 - val_loss: 0.3240 - val_accuracy: 0.8850\n",
      "Epoch 23/30\n",
      "1719/1719 [==============================] - 6s 3ms/step - loss: 0.2569 - accuracy: 0.9078 - val_loss: 0.3160 - val_accuracy: 0.8850\n",
      "Epoch 24/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2530 - accuracy: 0.9082 - val_loss: 0.3090 - val_accuracy: 0.8896\n",
      "Epoch 25/30\n",
      "1719/1719 [==============================] - 6s 4ms/step - loss: 0.2493 - accuracy: 0.9102 - val_loss: 0.3307 - val_accuracy: 0.8780\n",
      "Epoch 26/30\n",
      "1719/1719 [==============================] - 6s 3ms/step - loss: 0.2438 - accuracy: 0.9120 - val_loss: 0.3192 - val_accuracy: 0.8832\n",
      "Epoch 27/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2411 - accuracy: 0.9119 - val_loss: 0.2957 - val_accuracy: 0.8938\n",
      "Epoch 28/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2371 - accuracy: 0.9146 - val_loss: 0.3006 - val_accuracy: 0.8932\n",
      "Epoch 29/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2341 - accuracy: 0.9153 - val_loss: 0.2895 - val_accuracy: 0.8924\n",
      "Epoch 30/30\n",
      "1719/1719 [==============================] - 6s 3ms/step - loss: 0.2299 - accuracy: 0.9171 - val_loss: 0.3093 - val_accuracy: 0.8872\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, y_train, epochs=30, validation_data=(X_valid, y_valid))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Alternatively:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# history = model.fit(X_train, y_train, epochs=30, validation_split=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "pd.DataFrame(history.history).plot(figsize=(8, 5))\n",
    "plt.grid(True)\n",
    "plt.gca().set_ylim(0, 1)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We finally evaluate the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 0s 844us/step - loss: 0.3455 - accuracy: 0.8778\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.34552648663520813, 0.8777999877929688]"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using the model to make predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 38ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[0.  , 0.  , 0.  , 0.  , 0.  , 0.01, 0.  , 0.06, 0.  , 0.93],\n",
       "       [0.  , 0.  , 1.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ],\n",
       "       [0.  , 1.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ]],\n",
       "      dtype=float32)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_new = X_test[:3]\n",
    "y_proba = model.predict(X_new)\n",
    "y_proba.round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 11ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([9, 2, 1])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predict_x=model.predict(X_new) \n",
    "y_pred =np.argmax(predict_x, axis=1)\n",
    "\n",
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Angle boot', 'Pullover', 'Trouser'], dtype='<U11')"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(class_names)[y_pred]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([9, 2, 1], dtype=uint8)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_new = y_test[:3]\n",
    "y_new"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.4       , 1.        , 0.90588235, 0.89411765, 0.89019608,\n",
       "        0.89411765, 0.91372549, 0.90196078, 0.90196078, 0.89803922,\n",
       "        0.89411765, 0.90980392, 0.90980392, 0.90588235, 0.89019608,\n",
       "        0.87843137, 0.98823529, 0.70196078, 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ],\n",
       "       [0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        1.        , 0.92156863, 0.90588235, 0.91372549, 0.88627451,\n",
       "        0.88235294, 0.89803922, 0.87058824, 1.        , 0.56862745,\n",
       "        0.        , 0.        , 0.        , 0.        , 0.        ,\n",
       "        0.        , 0.        , 0.        ]])"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_new[:,3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(3):\n",
    "    plt.subplot(1, 3, i+1)\n",
    "    plt.imshow(X_new[i].reshape((28,28)),cmap='binary')\n",
    "    plt.title(class_names[y_pred[i]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Regression with Sequential API"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import fetch_california_housing\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "housing = fetch_california_housing()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_full, X_test, y_train_full, y_test = train_test_split(housing.data, housing.target)\n",
    "X_train, X_valid, y_train, y_valid = train_test_split(X_train_full, y_train_full)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Standarize."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "X_train = scaler.fit_transform(X_train)\n",
    "X_valid = scaler.transform(X_valid)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = keras.models.Sequential([\n",
    "    keras.layers.Dense(30, activation=\"relu\", input_shape=X_train.shape[1:]),\n",
    "    keras.layers.Dense(1)\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss=\"mean_squared_error\", optimizer=\"sgd\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.7382 - val_loss: 2.8659\n",
      "Epoch 2/20\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 1.3813 - val_loss: 0.8254\n",
      "Epoch 3/20\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.5059 - val_loss: 0.4566\n",
      "Epoch 4/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4412 - val_loss: 0.4302\n",
      "Epoch 5/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4046 - val_loss: 0.4248\n",
      "Epoch 6/20\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.4017 - val_loss: 0.4140\n",
      "Epoch 7/20\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3903 - val_loss: 0.4104\n",
      "Epoch 8/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3777 - val_loss: 0.4060\n",
      "Epoch 9/20\n",
      "363/363 [==============================] - 1s 1ms/step - loss: 0.3720 - val_loss: 0.3976\n",
      "Epoch 10/20\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3658 - val_loss: 0.3987\n",
      "Epoch 11/20\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3620 - val_loss: 0.3878\n",
      "Epoch 12/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3586 - val_loss: 0.3882\n",
      "Epoch 13/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3557 - val_loss: 0.3816\n",
      "Epoch 14/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3525 - val_loss: 0.3859\n",
      "Epoch 15/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3498 - val_loss: 0.3860\n",
      "Epoch 16/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3488 - val_loss: 0.3752\n",
      "Epoch 17/20\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3462 - val_loss: 0.3772\n",
      "Epoch 18/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3442 - val_loss: 0.3755\n",
      "Epoch 19/20\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3429 - val_loss: 0.3765\n",
      "Epoch 20/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3409 - val_loss: 0.3707\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, y_train, epochs=20, validation_data=(X_valid, y_valid))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "162/162 [==============================] - 0s 479us/step - loss: 0.3414\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.34135323762893677"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mse_test = model.evaluate(X_test, y_test)\n",
    "\n",
    "mse_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 10ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[1.0068436 ],\n",
       "       [0.62224364],\n",
       "       [2.0187678 ]], dtype=float32)"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_new = X_test[:3]\n",
    "y_pred = model.predict(X_new)\n",
    "\n",
    "y_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Saving and restoring"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save(\"my_model.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "anothermodel = keras.models.load_model(\"my_model.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 26ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[1.3958093e-05, 3.6013796e-07, 1.2022042e-05, 5.6356629e-07,\n",
       "        8.3547616e-07, 1.1620520e-02, 3.0096080e-05, 5.7869151e-02,\n",
       "        2.8582674e-04, 9.3016660e-01]], dtype=float32)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anothermodel.predict(X_new[:1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using callbacks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3388 - val_loss: 0.3730\n",
      "Epoch 2/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3377 - val_loss: 0.3675\n",
      "Epoch 3/10\n",
      "363/363 [==============================] - 0s 1ms/step - loss: 0.3353 - val_loss: 0.3676\n",
      "Epoch 4/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3340 - val_loss: 0.3665\n",
      "Epoch 5/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3334 - val_loss: 0.3654\n",
      "Epoch 6/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3320 - val_loss: 0.3650\n",
      "Epoch 7/10\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3306 - val_loss: 0.3617\n",
      "Epoch 8/10\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3317 - val_loss: 0.3613\n",
      "Epoch 9/10\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3287 - val_loss: 0.3605\n",
      "Epoch 10/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3398 - val_loss: 0.3630\n"
     ]
    }
   ],
   "source": [
    "checkpoint_cb = keras.callbacks.ModelCheckpoint(\"my_model_checkpoint.h5\", save_best_only=True)\n",
    "\n",
    "history = model.fit(X_train, y_train, epochs=10, validation_data=(X_valid, y_valid), callbacks=[checkpoint_cb])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = keras.models.load_model(\"my_model_checkpoint.h5\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Early stopping."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3285 - val_loss: 0.3611\n",
      "Epoch 2/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3263 - val_loss: 0.3602\n",
      "Epoch 3/100\n",
      "363/363 [==============================] - 1s 1ms/step - loss: 0.3258 - val_loss: 0.3650\n",
      "Epoch 4/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3253 - val_loss: 0.3612\n",
      "Epoch 5/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3239 - val_loss: 0.3560\n",
      "Epoch 6/100\n",
      "363/363 [==============================] - 0s 1ms/step - loss: 0.3233 - val_loss: 0.3588\n",
      "Epoch 7/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3236 - val_loss: 0.3556\n",
      "Epoch 8/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3219 - val_loss: 0.3598\n",
      "Epoch 9/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3328 - val_loss: 0.3604\n",
      "Epoch 10/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3221 - val_loss: 0.3579\n",
      "Epoch 11/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3216 - val_loss: 0.3575\n",
      "Epoch 12/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3226 - val_loss: 0.3512\n",
      "Epoch 13/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3217 - val_loss: 0.3550\n",
      "Epoch 14/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3205 - val_loss: 0.3545\n",
      "Epoch 15/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3198 - val_loss: 0.3514\n",
      "Epoch 16/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3185 - val_loss: 0.3604\n",
      "Epoch 17/100\n",
      "363/363 [==============================] - 0s 1ms/step - loss: 0.3182 - val_loss: 0.3503\n",
      "Epoch 18/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3167 - val_loss: 0.3645\n",
      "Epoch 19/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3166 - val_loss: 0.3561\n",
      "Epoch 20/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3163 - val_loss: 0.3524\n",
      "Epoch 21/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3192 - val_loss: 0.3525\n",
      "Epoch 22/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3190 - val_loss: 0.3571\n",
      "Epoch 23/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3192 - val_loss: 0.3474\n",
      "Epoch 24/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3141 - val_loss: 0.3548\n",
      "Epoch 25/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3141 - val_loss: 0.3430\n",
      "Epoch 26/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3122 - val_loss: 0.3497\n",
      "Epoch 27/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3119 - val_loss: 0.3445\n",
      "Epoch 28/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3150 - val_loss: 0.3503\n",
      "Epoch 29/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3123 - val_loss: 0.3489\n",
      "Epoch 30/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3113 - val_loss: 0.3502\n",
      "Epoch 31/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3130 - val_loss: 0.3457\n",
      "Epoch 32/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3093 - val_loss: 0.3502\n",
      "Epoch 33/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3097 - val_loss: 0.3449\n",
      "Epoch 34/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.3112 - val_loss: 0.3457\n",
      "Epoch 35/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3113 - val_loss: 0.3433\n"
     ]
    }
   ],
   "source": [
    "early_stopping_cb = keras.callbacks.EarlyStopping(patience=10, restore_best_weights=True)\n",
    "\n",
    "history = model.fit(X_train, y_train, epochs=100, validation_data=(X_valid, y_valid), callbacks=[checkpoint_cb, early_stopping_cb])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fine-Tuning hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_model(n_hidden=1, n_neurons=30, learning_rate=3e-3, input_shape=[8]):\n",
    "    model = keras.models.Sequential()\n",
    "    model.add(keras.layers.InputLayer(input_shape=input_shape))\n",
    "    for layer in range(n_hidden):\n",
    "        model.add(keras.layers.Dense(n_neurons, activatoin=\"relu\"))\n",
    "    model.add(keras.layers.Dense(1))\n",
    "    optimizer = keras.optimizers.SGD(learning_rate=learning_rate)\n",
    "    model.compile(loss=\"mse\", optimizer=optimizer)\n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/david/anaconda3/envs/ml37/lib/python3.7/site-packages/ipykernel_launcher.py:1: DeprecationWarning: KerasRegressor is deprecated, use Sci-Keras (https://github.com/adriangb/scikeras) instead. See https://www.adriangb.com/scikeras/stable/migration.html for help migrating.\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "keras_reg = keras.wrappers.scikit_learn.KerasRegressor(build_model) # Deprecated\n",
    "\n",
    "# scikeras should be used instead of keras.wrappers.scikit_learn.KerasRegressor\n",
    "# from scikeras.wrappers import KerasClassifier, KerasRegressor "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.stats import reciprocal\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "\n",
    "param_distribs = {\n",
    "    \"n_hidden\": [0, 1, 2, 3],\n",
    "    \"n_neurons\": np.arange(1, 100),\n",
    "    \"learning_rate\": reciprocal(3e-4, 3e-2),\n",
    "}\n",
    "\n",
    "rnd_search_cv = RandomizedSearchCV(keras_reg, param_distribs, n_iter=10, cv=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 2ms/step - loss: 2.9648 - val_loss: 1.6973\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 0s 1ms/step - loss: 3.3684 - val_loss: 3.3143\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6717 - val_loss: 2.5054\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 7.0723 - val_loss: 7.7960\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 0s 1ms/step - loss: 2.8391 - val_loss: 13.7060\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 39.8642 - val_loss: 56.7629\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 3.9170 - val_loss: 214.8645\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 721.4573 - val_loss: 892.3691\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 0s 1ms/step - loss: 69.4606 - val_loss: 4686.2407\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 18079.4082 - val_loss: 23259.3945\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 2ms/step - loss: 5544.4243 - val_loss: 112807.6172\n",
      "121/121 [==============================] - 0s 439us/step - loss: 965.9213\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 0s 997us/step - loss: 1.1539 - val_loss: 0.7159\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 0.5708 - val_loss: 0.6229\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 0.5245 - val_loss: 0.8756\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 0s 974us/step - loss: 0.5077 - val_loss: 1.1654\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 0s 1ms/step - loss: 0.5011 - val_loss: 1.4426\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 0s 846us/step - loss: 0.4982 - val_loss: 1.6789\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 0s 1ms/step - loss: 0.4963 - val_loss: 1.8394\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 2ms/step - loss: 0.4966 - val_loss: 2.0286\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 0.4963 - val_loss: 2.1821\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 0s 1ms/step - loss: 0.4959 - val_loss: 2.3020\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 0s 1ms/step - loss: 0.4956 - val_loss: 2.3598\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.4961 - val_loss: 2.3481\n",
      "121/121 [==============================] - 0s 460us/step - loss: 1.4116\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 0s 1ms/step - loss: 1.3084 - val_loss: 10.3049\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 0s 993us/step - loss: 4.6466 - val_loss: 37.3560\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 0s 1ms/step - loss: 2.9530 - val_loss: 154.8656\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 11.7185 - val_loss: 599.3502\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 1812.3209 - val_loss: 2047.7659\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 0s 984us/step - loss: 5446.8398 - val_loss: 8047.9546\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 0s 975us/step - loss: 30861.5625 - val_loss: 35265.8125\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 0s 2ms/step - loss: 1790.6537 - val_loss: 158096.9844\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 0s 1ms/step - loss: 451068.0000 - val_loss: 645480.0000\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 0s 836us/step - loss: 45379.1797 - val_loss: 2796506.0000\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 0s 942us/step - loss: 133812.2031 - val_loss: 10230671.0000\n",
      "121/121 [==============================] - 0s 497us/step - loss: 43313.5625\n",
      "Epoch 1/100\n",
      " 42/363 [==>...........................] - ETA: 0s - loss: 2.5175 "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/david/anaconda3/envs/ml37/lib/python3.7/site-packages/sklearn/model_selection/_validation.py:372: FitFailedWarning: \n",
      "27 fits failed out of a total of 30.\n",
      "The score on these train-test partitions for these parameters will be set to nan.\n",
      "If these failures are not expected, you can try to debug them by setting error_score='raise'.\n",
      "\n",
      "Below are more details about the failures:\n",
      "--------------------------------------------------------------------------------\n",
      "27 fits failed with the following error:\n",
      "Traceback (most recent call last):\n",
      "  File \"/home/david/anaconda3/envs/ml37/lib/python3.7/site-packages/sklearn/model_selection/_validation.py\", line 680, in _fit_and_score\n",
      "    estimator.fit(X_train, y_train, **fit_params)\n",
      "  File \"/home/david/anaconda3/envs/ml37/lib/python3.7/site-packages/keras/wrappers/scikit_learn.py\", line 164, in fit\n",
      "    self.model = self.build_fn(**self.filter_sk_params(self.build_fn))\n",
      "  File \"/tmp/ipykernel_5285/1070398292.py\", line 5, in build_model\n",
      "    model.add(keras.layers.Dense(n_neurons, activatoin=\"relu\"))\n",
      "  File \"/home/david/anaconda3/envs/ml37/lib/python3.7/site-packages/keras/dtensor/utils.py\", line 96, in _wrap_function\n",
      "    init_method(layer_instance, *args, **kwargs)\n",
      "  File \"/home/david/anaconda3/envs/ml37/lib/python3.7/site-packages/keras/layers/core/dense.py\", line 117, in __init__\n",
      "    super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n",
      "  File \"/home/david/anaconda3/envs/ml37/lib/python3.7/site-packages/tensorflow/python/trackable/base.py\", line 205, in _method_wrapper\n",
      "    result = method(self, *args, **kwargs)\n",
      "  File \"/home/david/anaconda3/envs/ml37/lib/python3.7/site-packages/keras/engine/base_layer.py\", line 340, in __init__\n",
      "    generic_utils.validate_kwargs(kwargs, allowed_kwargs)\n",
      "  File \"/home/david/anaconda3/envs/ml37/lib/python3.7/site-packages/keras/utils/generic_utils.py\", line 515, in validate_kwargs\n",
      "    raise TypeError(error_message, kwarg)\n",
      "TypeError: ('Keyword argument not understood:', 'activatoin')\n",
      "\n",
      "  warnings.warn(some_fits_failed_message, FitFailedWarning)\n",
      "/home/david/anaconda3/envs/ml37/lib/python3.7/site-packages/sklearn/model_selection/_search.py:972: UserWarning: One or more of the test scores are non-finite: [            nan             nan             nan             nan\n",
      "             nan             nan             nan -14760.29848027\n",
      "             nan             nan]\n",
      "  category=UserWarning,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "363/363 [==============================] - 1s 2ms/step - loss: 3.5557 - val_loss: 5.0567\n",
      "Epoch 2/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 0.7033 - val_loss: 8.6828\n",
      "Epoch 3/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 20.6477 - val_loss: 24.9571\n",
      "Epoch 4/100\n",
      "363/363 [==============================] - 0s 1ms/step - loss: 54.2020 - val_loss: 74.3954\n",
      "Epoch 5/100\n",
      "363/363 [==============================] - 0s 1ms/step - loss: 177.1187 - val_loss: 238.6773\n",
      "Epoch 6/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 374.8270 - val_loss: 638.4384\n",
      "Epoch 7/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 1021.9598 - val_loss: 72023.4844\n",
      "Epoch 8/100\n",
      "363/363 [==============================] - 1s 2ms/step - loss: 11311.6260 - val_loss: 234811.5156\n",
      "Epoch 9/100\n",
      "363/363 [==============================] - 1s 1ms/step - loss: 9850.9229 - val_loss: 801950.5625\n",
      "Epoch 10/100\n",
      "363/363 [==============================] - 1s 1ms/step - loss: 90140.3828 - val_loss: 2472144.7500\n",
      "Epoch 11/100\n",
      "363/363 [==============================] - 0s 1ms/step - loss: 346295.1562 - val_loss: 8265741.0000\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "RandomizedSearchCV(cv=3,\n",
       "                   estimator=<keras.wrappers.scikit_learn.KerasRegressor object at 0x7f5c5c6ae2d0>,\n",
       "                   param_distributions={'learning_rate': <scipy.stats._distn_infrastructure.rv_frozen object at 0x7f5c5c6aedd0>,\n",
       "                                        'n_hidden': [0, 1, 2, 3],\n",
       "                                        'n_neurons': array([ 1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17,\n",
       "       18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34,\n",
       "       35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51,\n",
       "       52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68,\n",
       "       69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85,\n",
       "       86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99])})"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rnd_search_cv.fit(X_train, y_train, epochs=100, validation_data=(X_valid, y_valid), callbacks=[keras.callbacks.EarlyStopping(patience=10)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'learning_rate': 0.014556882187920332, 'n_hidden': 0, 'n_neurons': 17}"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rnd_search_cv.best_params_"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "rnd_search_cv.best_score_"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.7.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}