fashion_mnist_train.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "fc3ef337",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow import keras\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "92759f7f",
   "metadata": {},
   "outputs": [],
   "source": [
    "(x_train , y_train) , (x_test , y_test) = keras.datasets.fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "9a754e5b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((60000, 28, 28), (10000, 28, 28))"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_train.shape , x_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "7cbb9e45",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 16 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "classes = ['Dress','Coat' ,'T-shirt_top','Trouser','Pullover','Sandal','Shirt','Sneaker','Bag','Ankle boot']\n",
    "fig, axes = plt.subplots(nrows=4, ncols=4, figsize=(10,10))\n",
    "\n",
    "for _ , axis in enumerate(axes.flat):\n",
    "    rand = np.random.randint(60000)\n",
    "    axis.imshow(x_train[rand] , cmap = 'gray')\n",
    "    axis.set_title(classes[y_train[rand]])\n",
    "    axis.axis('off')\n",
    "    \n",
    "plt.subplots_adjust(wspace=0.4, hspace=0.4)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "09aefbcc",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = keras.Sequential([\n",
    "    keras.layers.Rescaling(1/255.0 , input_shape = ((28,28,1))),\n",
    "    \n",
    "    keras.layers.Conv2D(32, kernel_size=(3, 3) , padding = 'same' , activation = 'relu'),\n",
    "    keras.layers.BatchNormalization(),\n",
    "    keras.layers.MaxPooling2D(pool_size=(2, 2)),\n",
    "    keras.layers.Conv2D(64, kernel_size=(3, 3) , padding = 'same' , activation = 'relu'),\n",
    "    keras.layers.BatchNormalization(),\n",
    "    keras.layers.MaxPooling2D(pool_size=(2, 2)),\n",
    "    \n",
    "    keras.layers.Flatten(),\n",
    "    keras.layers.Dense(256 , activation = 'relu'),\n",
    "    keras.layers.Dense(64 , activation = 'relu'),\n",
    "    keras.layers.Dense(10, activation=\"softmax\")]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b038671d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " rescaling (Rescaling)       (None, 28, 28, 1)         0         \n",
      "                                                                 \n",
      " conv2d (Conv2D)             (None, 28, 28, 32)        320       \n",
      "                                                                 \n",
      " batch_normalization (BatchN  (None, 28, 28, 32)       128       \n",
      " ormalization)                                                   \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 14, 14, 32)       0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 14, 14, 64)        18496     \n",
      "                                                                 \n",
      " batch_normalization_1 (Batc  (None, 14, 14, 64)       256       \n",
      " hNormalization)                                                 \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 7, 7, 64)         0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 3136)              0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 256)               803072    \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 64)                16448     \n",
      "                                                                 \n",
      " dense_2 (Dense)             (None, 10)                650       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 839,370\n",
      "Trainable params: 839,178\n",
      "Non-trainable params: 192\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "00cef6e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer='adam' , loss = 'sparse_categorical_crossentropy' , metrics = ['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "65076e10",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "422/422 [==============================] - 31s 71ms/step - loss: 0.3941 - accuracy: 0.8584 - val_loss: 0.6648 - val_accuracy: 0.7817\n",
      "Epoch 2/10\n",
      "422/422 [==============================] - 34s 80ms/step - loss: 0.2498 - accuracy: 0.9086 - val_loss: 0.2527 - val_accuracy: 0.9102\n",
      "Epoch 3/10\n",
      "422/422 [==============================] - 38s 90ms/step - loss: 0.2002 - accuracy: 0.9260 - val_loss: 0.2408 - val_accuracy: 0.9123\n",
      "Epoch 4/10\n",
      "422/422 [==============================] - 34s 82ms/step - loss: 0.1701 - accuracy: 0.9370 - val_loss: 0.2258 - val_accuracy: 0.9205\n",
      "Epoch 5/10\n",
      "422/422 [==============================] - 34s 80ms/step - loss: 0.1436 - accuracy: 0.9468 - val_loss: 0.2548 - val_accuracy: 0.9133\n",
      "Epoch 6/10\n",
      "422/422 [==============================] - 34s 80ms/step - loss: 0.1220 - accuracy: 0.9541 - val_loss: 0.2487 - val_accuracy: 0.9143\n",
      "Epoch 7/10\n",
      "422/422 [==============================] - 34s 81ms/step - loss: 0.0988 - accuracy: 0.9630 - val_loss: 0.2475 - val_accuracy: 0.9257\n",
      "Epoch 8/10\n",
      "422/422 [==============================] - 34s 80ms/step - loss: 0.0840 - accuracy: 0.9687 - val_loss: 0.2929 - val_accuracy: 0.9123\n",
      "Epoch 9/10\n",
      "422/422 [==============================] - 34s 80ms/step - loss: 0.0732 - accuracy: 0.9727 - val_loss: 0.2817 - val_accuracy: 0.9263\n",
      "Epoch 10/10\n",
      "422/422 [==============================] - 34s 81ms/step - loss: 0.0626 - accuracy: 0.9761 - val_loss: 0.3510 - val_accuracy: 0.9015\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1bcdc7d8a30>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(x_train , y_train , batch_size = 128 , validation_split = 0.1 , epochs=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "7e477e61",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 3s 9ms/step - loss: 0.3919 - accuracy: 0.9022\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.39189982414245605, 0.9021999835968018]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(x_test , y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d0bb20ee",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save('fashion_mnist.h5')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "keras",
   "language": "python",
   "name": "keras"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}