Model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\LASYA\\Anaconda3\\lib\\site-packages\\h5py\\__init__.py:72: UserWarning: h5py is running against HDF5 1.10.2 when it was built against 1.10.3, this may cause problems\n",
      "  '{0}.{1}.{2}'.format(*version.hdf5_built_version_tuple)\n",
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from keras import layers\n",
    "from keras.layers import Input, Dense, Activation, ZeroPadding2D, BatchNormalization, Flatten, Conv2D\n",
    "from keras.layers import AveragePooling2D, MaxPooling2D, Dropout, GlobalMaxPooling2D, GlobalAveragePooling2D\n",
    "from keras.utils import np_utils, print_summary\n",
    "import pandas as pd\n",
    "from keras.models import Sequential\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "import keras.backend as K"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"data.csv\")\n",
    "dataset = np.array(data)\n",
    "np.random.shuffle(dataset)\n",
    "X = dataset\n",
    "Y = dataset\n",
    "X = X[:, 0:1024]\n",
    "Y = Y[:, 1024]\n",
    "X_train = X[0:70000, :]\n",
    "X_train = X_train / 255.\n",
    "X_test = X[70000:72001, :]\n",
    "X_test = X_test / 255.\n",
    "# Reshape\n",
    "Y = Y.reshape(Y.shape[0], 1)\n",
    "Y_train = Y[0:70000, :]\n",
    "Y_train = Y_train.T\n",
    "Y_test = Y[70000:72001, :]\n",
    "Y_test = Y_test.T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of training examples = 70000\n",
      "number of test examples = 2000\n",
      "X_train shape: (70000, 1024)\n",
      "Y_train shape: (1, 70000)\n",
      "X_test shape: (2000, 1024)\n",
      "Y_test shape: (1, 2000)\n"
     ]
    }
   ],
   "source": [
    "print(\"number of training examples = \" + str(X_train.shape[0]))\n",
    "print(\"number of test examples = \" + str(X_test.shape[0]))\n",
    "print(\"X_train shape: \" + str(X_train.shape))\n",
    "print(\"Y_train shape: \" + str(Y_train.shape))\n",
    "print(\"X_test shape: \" + str(X_test.shape))\n",
    "print(\"Y_test shape: \" + str(Y_test.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "image_x = 32\n",
    "image_y = 32\n",
    "\n",
    "train_y = np_utils.to_categorical(Y_train)\n",
    "test_y = np_utils.to_categorical(Y_test)\n",
    "train_y = train_y.reshape(train_y.shape[1], train_y.shape[2])\n",
    "test_y = test_y.reshape(test_y.shape[1], test_y.shape[2])\n",
    "X_train = X_train.reshape(X_train.shape[0], 32, 32, 1)\n",
    "X_test = X_test.reshape(X_test.shape[0], 32, 32, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train shape: (70000, 32, 32, 1)\n",
      "Y_train shape: (70000, 37)\n"
     ]
    }
   ],
   "source": [
    "print(\"X_train shape: \" + str(X_train.shape))\n",
    "print(\"Y_train shape: \" + str(train_y.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Building model\n",
    "def keras_model(image_x,image_y):\n",
    "    num_of_classes = 37\n",
    "    model = Sequential()\n",
    "    model.add(Conv2D(filters=32, kernel_size=(5, 5), input_shape=(image_x, image_y, 1), activation='relu'))\n",
    "    model.add(MaxPooling2D(pool_size=(2, 2), strides=(2, 2), padding='same'))\n",
    "    model.add(Conv2D(64, (5, 5), activation='relu'))\n",
    "    model.add(MaxPooling2D(pool_size=(5, 5), strides=(5, 5), padding='same'))\n",
    "    model.add(Flatten())\n",
    "    model.add(Dense(num_of_classes, activation='softmax'))\n",
    "    model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "    filepath = \"devanagari.h5\"\n",
    "    checkpoint1 = ModelCheckpoint(filepath, monitor='val_acc', verbose=1, save_best_only=True, mode='max')\n",
    "    callbacks_list = [checkpoint1]\n",
    "\n",
    "    return model, callbacks_list\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING: Logging before flag parsing goes to stderr.\n",
      "W0815 13:43:10.554732 10700 deprecation_wrapper.py:119] From C:\\Users\\LASYA\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:74: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.\n",
      "\n",
      "W0815 13:43:14.619377 10700 deprecation_wrapper.py:119] From C:\\Users\\LASYA\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:517: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n",
      "\n",
      "W0815 13:43:15.528964 10700 deprecation_wrapper.py:119] From C:\\Users\\LASYA\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:4138: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.\n",
      "\n",
      "W0815 13:43:16.692497 10700 deprecation_wrapper.py:119] From C:\\Users\\LASYA\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:3976: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.\n",
      "\n",
      "W0815 13:43:17.080791 10700 deprecation_wrapper.py:119] From C:\\Users\\LASYA\\Anaconda3\\lib\\site-packages\\keras\\optimizers.py:790: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n",
      "\n",
      "W0815 13:43:17.288521 10700 deprecation_wrapper.py:119] From C:\\Users\\LASYA\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:3295: The name tf.log is deprecated. Please use tf.math.log instead.\n",
      "\n",
      "W0815 13:43:21.696010 10700 deprecation.py:323] From C:\\Users\\LASYA\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\ops\\math_grad.py:1250: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n",
      "W0815 13:43:21.905169 10700 deprecation_wrapper.py:119] From C:\\Users\\LASYA\\Anaconda3\\lib\\site-packages\\keras\\backend\\tensorflow_backend.py:986: The name tf.assign_add is deprecated. Please use tf.compat.v1.assign_add instead.\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 70000 samples, validate on 2000 samples\n",
      "Epoch 1/8\n",
      "70000/70000 [==============================] - 107s 2ms/step - loss: 0.7642 - acc: 0.7845 - val_loss: 0.2939 - val_acc: 0.9040\n",
      "\n",
      "Epoch 00001: val_acc improved from -inf to 0.90400, saving model to devanagari.h5\n",
      "Epoch 2/8\n",
      "70000/70000 [==============================] - 96s 1ms/step - loss: 0.2268 - acc: 0.9315 - val_loss: 0.2175 - val_acc: 0.9255\n",
      "\n",
      "Epoch 00002: val_acc improved from 0.90400 to 0.92550, saving model to devanagari.h5\n",
      "Epoch 3/8\n",
      "70000/70000 [==============================] - 92s 1ms/step - loss: 0.1485 - acc: 0.9551 - val_loss: 0.1373 - val_acc: 0.9535\n",
      "\n",
      "Epoch 00003: val_acc improved from 0.92550 to 0.95350, saving model to devanagari.h5\n",
      "Epoch 4/8\n",
      "70000/70000 [==============================] - 93s 1ms/step - loss: 0.1099 - acc: 0.9661 - val_loss: 0.1270 - val_acc: 0.9650\n",
      "\n",
      "Epoch 00004: val_acc improved from 0.95350 to 0.96500, saving model to devanagari.h5\n",
      "Epoch 5/8\n",
      "70000/70000 [==============================] - 92s 1ms/step - loss: 0.0875 - acc: 0.9739 - val_loss: 0.1350 - val_acc: 0.9570\n",
      "\n",
      "Epoch 00005: val_acc did not improve from 0.96500\n",
      "Epoch 6/8\n",
      "70000/70000 [==============================] - 92s 1ms/step - loss: 0.0723 - acc: 0.9776 - val_loss: 0.0969 - val_acc: 0.9690\n",
      "\n",
      "Epoch 00006: val_acc improved from 0.96500 to 0.96900, saving model to devanagari.h5\n",
      "Epoch 7/8\n",
      "70000/70000 [==============================] - 93s 1ms/step - loss: 0.0598 - acc: 0.9811 - val_loss: 0.1032 - val_acc: 0.9690\n",
      "\n",
      "Epoch 00007: val_acc did not improve from 0.96900\n",
      "Epoch 8/8\n",
      "70000/70000 [==============================] - 94s 1ms/step - loss: 0.0492 - acc: 0.9840 - val_loss: 0.0998 - val_acc: 0.9720\n",
      "\n",
      "Epoch 00008: val_acc improved from 0.96900 to 0.97200, saving model to devanagari.h5\n",
      "CNN Error: 2.80%\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 28, 28, 32)        832       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 14, 14, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 10, 10, 64)        51264     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 2, 2, 64)          0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 37)                9509      \n",
      "=================================================================\n",
      "Total params: 61,605\n",
      "Trainable params: 61,605\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model, callbacks_list = keras_model(image_x, image_y)\n",
    "model.fit(X_train, train_y, validation_data=(X_test, test_y), epochs=8, batch_size=64,callbacks=callbacks_list)\n",
    "scores = model.evaluate(X_test, test_y, verbose=0)\n",
    "print(\"CNN Error: %.2f%%\" % (100 - scores[1] * 100))\n",
    "print_summary(model)\n",
    "model.save('devanagari.h5')"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}