Keras_Tensorboard_CNN (1).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this Tutorial we are going to load the summaries of keras with Tensorboard for better visualization. We are goin to train a CNN architecture/model with cifar dataset which is already preloaded in Keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from time import time\n",
    "import matplotlib.pyplot as plt\n",
    "from tensorflow.python.keras.callbacks import TensorBoard\n",
    "from tensorflow.python.keras import datasets\n",
    "keras = tf.keras"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Collecting all necessary tools to construct a CNN layer with the help of Keras in Tensorflow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "sequential = keras.Sequential\n",
    "dense = keras.layers.Dense\n",
    "dropout = keras.layers.Dropout\n",
    "cnn = keras.layers.Conv2D\n",
    "max_pool = keras.layers.MaxPool2D\n",
    "flat = keras.layers.Flatten\n",
    "batch_size = 10\n",
    "epochs = 10\n",
    "nclasses = 10\n",
    "model_name = 'keras_model.h5'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As the dataset is very large and due to lack of computation power (from next time we will see better results as we are going to train on a GPU) we are now only taking first 1000 examples in training set and 100 in test set\n",
    "\n",
    "We are preprocessing it with dividing with 255 as the input pixels maximum size is 255 and to get the values in the range of 0 to 1 we are fividing with 255 (max value of pixel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(50000, 32, 32, 3) (50000, 1)\n"
     ]
    }
   ],
   "source": [
    "(x_train,y_train),(x_test,y_test) = datasets.cifar10.load_data()\n",
    "print(x_train.shape,y_train.shape)\n",
    "x_train,y_train = x_train[:1000],y_train[:1000]\n",
    "x_test,y_test = x_test[:100],y_test[:100]\n",
    "x_train = x_train/255\n",
    "x_test = x_test/255\n",
    "#converting to categorical outputs\n",
    "y_train = keras.utils.to_categorical(y_train,num_classes=nclasses)\n",
    "y_test = keras.utils.to_categorical(y_test,num_classes=nclasses)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Creating a CNN layer of Three layer followed by a Flattening and softmax to predict the outcomes of the each image given out of the 10 categories available to us. One layer is Conv2d+conv2d+maxpool+dropout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#creating the model of cnn\n",
    "model = sequential()\n",
    "model.add(cnn(32,(3,3),padding='same',activation='relu',input_shape=x_train.shape[1:]))\n",
    "model.add(cnn(32,(3,3),activation='relu'))\n",
    "model.add(max_pool(pool_size=(2,2)))\n",
    "model.add(dropout(0.2))\n",
    "\n",
    "model.add(cnn(64,(3,3),padding='same',activation='relu'))\n",
    "model.add(cnn(64,(3,3),activation='relu'))\n",
    "model.add(max_pool(pool_size=(2,2)))\n",
    "model.add(dropout(0.4))\n",
    "\n",
    "model.add(cnn(64,(3,3),padding='same',activation='relu'))\n",
    "model.add(cnn(64,(3,3),activation='relu'))\n",
    "model.add(max_pool(pool_size=(2,2)))\n",
    "model.add(dropout(0.5))\n",
    "\n",
    "model.add(flat())\n",
    "model.add(dense(512,activation='relu'))\n",
    "model.add(dropout(0.5))\n",
    "model.add(dense(nclasses,activation='softmax'))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compiling the model to build the layers we can see the model summary here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 32, 32, 32)        896       \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 30, 30, 32)        9248      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 15, 15, 32)        0         \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 15, 15, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 15, 15, 64)        18496     \n",
      "_________________________________________________________________\n",
      "conv2d_4 (Conv2D)            (None, 13, 13, 64)        36928     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 6, 6, 64)          0         \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 6, 6, 64)          0         \n",
      "_________________________________________________________________\n",
      "conv2d_5 (Conv2D)            (None, 6, 6, 64)          36928     \n",
      "_________________________________________________________________\n",
      "conv2d_6 (Conv2D)            (None, 4, 4, 64)          36928     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_3 (MaxPooling2 (None, 2, 2, 64)          0         \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 2, 2, 64)          0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 512)               131584    \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 10)                5130      \n",
      "=================================================================\n",
      "Total params: 276,138\n",
      "Trainable params: 276,138\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "#Here loss is the difference between the y-y(i) (predicted to actual values)\n",
    "#optimizer is the one which is way it need to be optimize the loss ex: Gradientdescentoptimizer,rms,adam\n",
    "model.compile(optimizer='rmsprop',loss='categorical_crossentropy',metrics=['accuracy'])\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Getting the Tensoboard ready for saving the summaries and creating the directory with name logs in the parent folder "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "tensorboard = TensorBoard(log_dir=\"logs/\",histogram_freq=0,write_graph=True,write_images=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To run the tensorboard we need to run\n",
    "\n",
    "tensorboard --logdir==training:path_to_directory --host=127.0.0.1\n",
    "\n",
    "for example:\n",
    "tensorboard --logdir==training:C:\\Users\\jatoth.kumar\\PycharmProjects\\robot_project_hyd\\logs --host=127.0.0.1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are going to train the model with train_x and train_y and going to validate on test_x and text_y datasets. Here we need to do the callbacks to tensorboard for visualisation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 1000 samples, validate on 100 samples\n",
      "Epoch 1/10\n",
      "1000/1000 [==============================] - 8s 8ms/step - loss: 2.3091 - acc: 0.0930 - val_loss: 2.2974 - val_acc: 0.2000\n",
      "Epoch 2/10\n",
      "1000/1000 [==============================] - 7s 7ms/step - loss: 2.2664 - acc: 0.1260 - val_loss: 2.1847 - val_acc: 0.2400\n",
      "Epoch 3/10\n",
      "1000/1000 [==============================] - 7s 7ms/step - loss: 2.1645 - acc: 0.1830 - val_loss: 2.0448 - val_acc: 0.2100\n",
      "Epoch 4/10\n",
      "1000/1000 [==============================] - 7s 7ms/step - loss: 2.0751 - acc: 0.2140 - val_loss: 1.9817 - val_acc: 0.3000\n",
      "Epoch 5/10\n",
      "1000/1000 [==============================] - 7s 7ms/step - loss: 2.0207 - acc: 0.2530 - val_loss: 2.0023 - val_acc: 0.2600\n",
      "Epoch 6/10\n",
      "1000/1000 [==============================] - 7s 7ms/step - loss: 1.9689 - acc: 0.2540 - val_loss: 1.8677 - val_acc: 0.3300\n",
      "Epoch 7/10\n",
      "1000/1000 [==============================] - 7s 7ms/step - loss: 1.9026 - acc: 0.2830 - val_loss: 1.8145 - val_acc: 0.3800\n",
      "Epoch 8/10\n",
      "1000/1000 [==============================] - 7s 7ms/step - loss: 1.8748 - acc: 0.2820 - val_loss: 1.9620 - val_acc: 0.1900\n",
      "Epoch 9/10\n",
      "1000/1000 [==============================] - 7s 7ms/step - loss: 1.8598 - acc: 0.3090 - val_loss: 1.8684 - val_acc: 0.2200\n",
      "Epoch 10/10\n",
      "1000/1000 [==============================] - 7s 7ms/step - loss: 1.7968 - acc: 0.3200 - val_loss: 1.8404 - val_acc: 0.3500\n"
     ]
    }
   ],
   "source": [
    "his = model.fit(x_train,y_train,batch_size=batch_size,epochs=epochs,validation_data=(x_test,y_test),shuffle=True,callbacks=[tensorboard])\n",
    "model.save(model_name)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With much less data we have achieved the accuracy to be 0.35 which is a great acheivement. Now we are going to calculate the scores on x_test and y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100/100 [==============================] - 0s 2ms/step\n",
      "1.840434603691101 0.35\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xd6cfb70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "scores = model.evaluate(x_test,y_test,verbose=1)\n",
    "print(scores[0],scores[1])\n",
    "plt.plot(his.history['acc'])\n",
    "plt.xlabel('epochs')\n",
    "plt.ylabel('accuracy')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The graph which we plot with the matplotlib and the Tensorboard should be the same. After running the above command in the Terminal (Tensorboard command) just open any browser and go to 127.0.0.1:6006. This will show all the log visualizations which will give a better understanding about the model you have built"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}