fashion_mnist_cnn.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "Basic CNN for fashion MNIST problem\n",
        "\n",
        "Code by Ayush Agarwal , IIT BHU ECE24, Infoedge DS"
      ],
      "metadata": {
        "id": "Ee1XXu_nWuIS"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9M05sO7-NS1G",
        "outputId": "50e10bae-348d-454c-dd86-a0b24bb9a4c6"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Tensorflow version : 2.15.0\n"
          ]
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "print(\"Tensorflow version : \",end=\"\")\n",
        "print(tf.__version__)\n",
        "\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# These are the datasets availaible by default in tensorflow keras - https://www.tensorflow.org/api_docs/python/tf/keras/datasets\n",
        "\n",
        "# I am going to use Fasion MNIST to make a clothes classifier\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/datasets/fashion_mnist/load_data\n",
        "\n",
        "fashion_mnist = tf.keras.datasets.fashion_mnist\n",
        "(train_images, train_labels), (test_images, test_labels) = fashion_mnist.load_data()\n",
        "(X_train,y_train),(X_valid,y_valid) = (train_images, train_labels), (test_images, test_labels)\n",
        "\n",
        "class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',\n",
        "               'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']\n",
        "\n",
        "print(X_train.shape)\n",
        "print(y_train.shape)\n",
        "print(X_valid.shape)\n",
        "print(y_valid.shape)\n",
        "\n",
        "# In Python, a single element within parentheses without a comma (e.g., (60000)) is interpreted as a regular integer, not a tuple. To create a tuple with a single element, you need the trailing comma. This distinction helps the language correctly interpret the shape of the array.\n",
        "# The (60000,) notation is a tuple indicating a one-dimensional array with 60,000 elements."
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NCwwUWSbPW11",
        "outputId": "8e40045a-b76b-4b9a-8b51-b96d623c2efe"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(60000, 28, 28)\n",
            "(60000,)\n",
            "(10000, 28, 28)\n",
            "(10000,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Label\tDescription\n",
        "\n",
        "0\tT-shirt/top\n",
        "\n",
        "1\tTrouser\n",
        "\n",
        "2\tPullover\n",
        "\n",
        "3\tDress\n",
        "\n",
        "4\tCoat\n",
        "\n",
        "5\tSandal\n",
        "\n",
        "6\tShirt\n",
        "\n",
        "7\tSneaker\n",
        "\n",
        "8\tBag\n",
        "\n",
        "9\tAnkle boot"
      ],
      "metadata": {
        "id": "n_SpTzXvRimF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# visualization\n",
        "# print(train_images[0])\n",
        "# an image is just a 28x28 array with values from 0 to 255 in it"
      ],
      "metadata": {
        "id": "dNjNzS-gQ0bP"
      },
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Normalize pixel values to be between 0 and 1\n",
        "train_images, test_images = train_images / 255.0, test_images / 255.0\n",
        "\n",
        "# python is useful for double assignents like this , remember a,b=b,a for swapping numbers, just like that"
      ],
      "metadata": {
        "id": "w20oEbS9SxB7"
      },
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# This subplot way is actually a very common way to visualize data in tensorflow guide , saw it in multiple docs\n",
        "\n",
        "plt.figure(figsize=(14,4))\n",
        "for i in range(20):\n",
        "    # https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.subplot.html\n",
        "    plt.subplot(2,10,i+1)\n",
        "    # https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.xticks.html\n",
        "    # we can set x and y labels in a plot\n",
        "    plt.xticks([])\n",
        "    plt.yticks([])\n",
        "    plt.grid(False)\n",
        "    plt.imshow(train_images[i], cmap=plt.cm.binary)\n",
        "    # here we used the array which we made above to replace numbers with names\n",
        "    plt.xlabel(class_names[train_labels[i]])\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 305
        },
        "id": "SRb0t3EsTdcj",
        "outputId": "a73e5100-1e88-4d5c-8d8b-5c90140d64ff"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1400x400 with 20 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers, models\n",
        "\n",
        "# number of output classes\n",
        "num_classes = len(class_names)\n",
        "\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/Sequential\n",
        "# very good explanation of all classes of tf keras models\n",
        "\n",
        "# Docs -\n",
        "# tf.keras.Sequential(\n",
        "#     layers=None, trainable=True, name=None\n",
        "# )\n",
        "\n",
        "# Different types of layers\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/layers\n",
        "\n",
        "cnn_model_1 = models.Sequential(\n",
        "    layers = [\n",
        "        # layers.Rescaling(1./255, input_shape=(img_height, img_width, 3)),\n",
        "        # I won't need rescaling layer as I already have done /255 above\n",
        "        layers.Conv2D(16, 3, padding='same', activation='relu'),\n",
        "        layers.MaxPooling2D(),\n",
        "        # https://www.tensorflow.org/api_docs/python/tf/keras/layers/ELU\n",
        "        layers.Conv2D(32, 3, padding='same', activation='elu'),\n",
        "        layers.MaxPooling2D(),\n",
        "        layers.Conv2D(64, 3, padding='same', activation='elu'),\n",
        "        layers.MaxPooling2D(),\n",
        "        # dropout layer to reduce overfitting\n",
        "        layers.Dropout(0.2),\n",
        "        layers.Flatten(),\n",
        "        # layers.Dense(128, activation='relu'),\n",
        "        # layers.Dense(num_classes)\n",
        "    ]\n",
        ")\n",
        "\n",
        "# Docs -\n",
        "# add(\n",
        "#     layer, rebuild=True\n",
        "# )\n",
        "\n",
        "# cnn_model_1.add(layers.Flatten())\n",
        "cnn_model_1.add(layers.Dense(64, activation='relu'))\n",
        "cnn_model_1.add(layers.Dense(10))\n",
        "\n",
        "# We can add layers by directly adding them while declaring the model\n",
        "# as well as using the add method like this\n",
        "\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/optimizers/Adam\n",
        "cnn_model_1.compile(optimizer=keras.optimizers.Adam(learning_rate=1e-3),       #'adam', default lr is 0.001 for adam\n",
        "              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "# agar end of last epoch tak lag raha hai ki accuracy badh raha tha to learning rate badha do number of epochs same rakh ke\n",
        "# ya fir number of epochs bhi badha sakte ho (but that would take more time )\n",
        "\n",
        "# Docs -\n",
        "# compile(\n",
        "#     optimizer='rmsprop',\n",
        "#     loss=None,\n",
        "#     loss_weights=None,\n",
        "#     metrics=None,\n",
        "#     weighted_metrics=None,\n",
        "#     run_eagerly=False,\n",
        "#     steps_per_execution=1,\n",
        "#     jit_compile='auto',\n",
        "#     auto_scale_loss=True\n",
        "# )\n",
        "\n",
        "# When using the delayed-build pattern (no input shape specified), you can choose to manually build your model by calling\n",
        "# `build(batch_input_shape)`\n",
        "\n",
        "# BATCH SIZE\n",
        "# Higher the batch size, the more memory would be required to train the model\n",
        "# If we take batch_size = number of samples (60000 here) then it will move perfectly but converge very slowly , would require lots of epochs\n",
        "# if we take batch_size = 1 then its stochatic gradient descent, it will move values fast but won't converge to correct value\n",
        "# so we use mini batch gradient descent\n",
        "# batch_size = 300\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/layers/Conv2D\n",
        "# Input shape = A 4D tensor with shape: (batch_size, height, width, channels)\n",
        "cnn_model_1.build(input_shape=(1,28,28,1))\n",
        "cnn_model_1.summary()\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xg_nGJBJTgGB",
        "outputId": "7abea8ce-4b91-4ff7-9654-f566653b16b8"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_7\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d_6 (Conv2D)           (1, 28, 28, 16)           160       \n",
            "                                                                 \n",
            " max_pooling2d_6 (MaxPoolin  (1, 14, 14, 16)           0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_7 (Conv2D)           (1, 14, 14, 32)           4640      \n",
            "                                                                 \n",
            " max_pooling2d_7 (MaxPoolin  (1, 7, 7, 32)             0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_8 (Conv2D)           (1, 7, 7, 64)             18496     \n",
            "                                                                 \n",
            " max_pooling2d_8 (MaxPoolin  (1, 3, 3, 64)             0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_2 (Dropout)         (1, 3, 3, 64)             0         \n",
            "                                                                 \n",
            " flatten_2 (Flatten)         (1, 576)                  0         \n",
            "                                                                 \n",
            " dense_4 (Dense)             (1, 64)                   36928     \n",
            "                                                                 \n",
            " dense_5 (Dense)             (1, 10)                   650       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 60874 (237.79 KB)\n",
            "Trainable params: 60874 (237.79 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "history = cnn_model_1.fit(train_images, train_labels, epochs=10, batch_size=30,\n",
        "                    validation_data=(test_images, test_labels))\n",
        "\n",
        "# My observation - Batch size can be changed both here as well as above in the input layer of model\n",
        "# would be better to set the batch size here only\n",
        "\n",
        "# fit(\n",
        "#     x=None,\n",
        "#     y=None,\n",
        "#     batch_size=None,\n",
        "#     epochs=1,\n",
        "#     verbose='auto',\n",
        "#     callbacks=None,\n",
        "#     validation_split=0.0,\n",
        "#     validation_data=None,\n",
        "#     shuffle=True,\n",
        "#     class_weight=None,\n",
        "#     sample_weight=None,\n",
        "#     initial_epoch=0,\n",
        "#     steps_per_epoch=None,\n",
        "#     validation_steps=None,\n",
        "#     validation_batch_size=None,\n",
        "#     validation_freq=1\n",
        "# )"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QWBPHocNV6Zf",
        "outputId": "1ca148eb-e9d4-4732-dfb4-d82ead3fc9fa"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "2000/2000 [==============================] - 10s 4ms/step - loss: 0.4849 - accuracy: 0.8213 - val_loss: 0.3355 - val_accuracy: 0.8762\n",
            "Epoch 2/10\n",
            "2000/2000 [==============================] - 9s 5ms/step - loss: 0.3143 - accuracy: 0.8845 - val_loss: 0.2971 - val_accuracy: 0.8903\n",
            "Epoch 3/10\n",
            "2000/2000 [==============================] - 8s 4ms/step - loss: 0.2704 - accuracy: 0.8997 - val_loss: 0.2760 - val_accuracy: 0.8960\n",
            "Epoch 4/10\n",
            "2000/2000 [==============================] - 9s 4ms/step - loss: 0.2462 - accuracy: 0.9078 - val_loss: 0.2601 - val_accuracy: 0.9063\n",
            "Epoch 5/10\n",
            "2000/2000 [==============================] - 9s 4ms/step - loss: 0.2247 - accuracy: 0.9165 - val_loss: 0.2547 - val_accuracy: 0.9060\n",
            "Epoch 6/10\n",
            "2000/2000 [==============================] - 8s 4ms/step - loss: 0.2070 - accuracy: 0.9227 - val_loss: 0.2490 - val_accuracy: 0.9093\n",
            "Epoch 7/10\n",
            "2000/2000 [==============================] - 9s 4ms/step - loss: 0.1930 - accuracy: 0.9281 - val_loss: 0.2446 - val_accuracy: 0.9123\n",
            "Epoch 8/10\n",
            "2000/2000 [==============================] - 9s 4ms/step - loss: 0.1819 - accuracy: 0.9319 - val_loss: 0.2461 - val_accuracy: 0.9125\n",
            "Epoch 9/10\n",
            "2000/2000 [==============================] - 9s 4ms/step - loss: 0.1712 - accuracy: 0.9354 - val_loss: 0.2448 - val_accuracy: 0.9154\n",
            "Epoch 10/10\n",
            "2000/2000 [==============================] - 10s 5ms/step - loss: 0.1605 - accuracy: 0.9394 - val_loss: 0.2450 - val_accuracy: 0.9168\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Visualizing the results for the model\n",
        "\n",
        "# evaluate the model at the end\n",
        "test_loss, test_acc = cnn_model_1.evaluate(test_images,  test_labels, verbose=2)\n",
        "print('\\nTest accuracy:', test_acc)\n",
        "\n",
        "# Returns the loss value & metrics values for the model in test mode.\n",
        "# evaluate(\n",
        "#     x=None,\n",
        "#     y=None,\n",
        "#     batch_size=None,\n",
        "#     verbose='auto',\n",
        "#     sample_weight=None,\n",
        "#     steps=None,\n",
        "#     callbacks=None,\n",
        "#     return_dict=False,\n",
        "#     **kwargs\n",
        "# )\n",
        "\n",
        "\n",
        "# way 1 to visualize\n",
        "# plt.plot(history.history['accuracy'], label='accuracy')\n",
        "# plt.plot(history.history['val_accuracy'], label = 'val_accuracy')\n",
        "# plt.xlabel('Epoch')\n",
        "# plt.ylabel('Accuracy')\n",
        "# plt.ylim([0.5, 1])\n",
        "# plt.legend(loc='lower right')\n",
        "\n",
        "\n",
        "# way 2 to visualize -\n",
        "\n",
        "acc = history.history['accuracy']\n",
        "val_acc = history.history['val_accuracy']\n",
        "\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "\n",
        "epochs = 10 # was meant to be defined above but we directly wrote 10 there so defining here\n",
        "epochs_range = range(epochs)\n",
        "\n",
        "plt.figure(figsize=(12, 5))\n",
        "plt.subplot(1, 2, 1)\n",
        "plt.plot(epochs_range, acc, label='Training Accuracy')\n",
        "plt.plot(epochs_range, val_acc, label='Validation Accuracy')\n",
        "plt.legend(loc='lower right')\n",
        "plt.title('Training and Validation Accuracy')\n",
        "\n",
        "plt.subplot(1, 2, 2)\n",
        "plt.plot(epochs_range, loss, label='Training Loss')\n",
        "plt.plot(epochs_range, val_loss, label='Validation Loss')\n",
        "plt.legend(loc='upper right')\n",
        "plt.title('Training and Validation Loss')\n",
        "plt.show()\n",
        "\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 522
        },
        "id": "H_NZO1rkwNmE",
        "outputId": "48b127fb-415b-48d0-cb11-df10a43d513e"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 - 1s - loss: 0.2450 - accuracy: 0.9168 - 613ms/epoch - 2ms/step\n",
            "\n",
            "Test accuracy: 0.9168000221252441\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1200x500 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# https://github.com/tensorflow/docs/blob/master/site/en/tutorials/keras/classification.ipynb\n",
        "# To see the model predictions on few cases\n",
        "\n",
        "probability_model = tf.keras.Sequential([cnn_model_1,\n",
        "                                         tf.keras.layers.Softmax()])\n",
        "predictions = probability_model.predict(test_images)\n",
        "\n",
        "plt.figure(figsize=(10,12))\n",
        "for i in range(20):\n",
        "    # https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.subplot.html\n",
        "    plt.subplot(5,5,i+1)\n",
        "    # https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.xticks.html\n",
        "    # we can set x and y labels in a plot\n",
        "    plt.xticks([])\n",
        "    plt.yticks([])\n",
        "    plt.grid(False)\n",
        "    plt.imshow(test_images[i], cmap=plt.cm.binary)\n",
        "    # here we used the array which we made above to replace numbers with names\n",
        "    plt.xlabel(\"Actual : \" + str(class_names[test_labels[i]]) + \"\\n Predicted : \" + str(class_names[np.argmax(predictions[i])]))\n",
        "    # plt.ylabel(class_names[np.argmax(predictions[i])])\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 797
        },
        "id": "mPQg-nOR2W4W",
        "outputId": "c8b0b1ce-0d19-4a5a-e985-67c938c0e11b"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 1s 2ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x1200 with 20 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Saving ML model as .keras format\n",
        "\n",
        "cnn_model_1.save('cnn_fashion_mnist_model.keras')\n",
        "# https://www.tensorflow.org/tutorials/keras/save_and_load\n",
        "saved_model = tf.keras.models.load_model('cnn_fashion_mnist_model.keras')\n",
        "\n",
        "# Show the model architecture\n",
        "saved_model.summary()\n",
        "\n",
        "loss, acc = saved_model.evaluate(test_images, test_labels, verbose=2)\n",
        "print('Restored model, accuracy: {:5.2f}%'.format(100 * acc))\n",
        "\n",
        "print(saved_model.predict(test_images).shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bPwLgquk6pIa",
        "outputId": "9c7f8dd3-1f26-4e31-bb49-cc1d436bbcfa"
      },
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_7\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d_6 (Conv2D)           (1, 28, 28, 16)           160       \n",
            "                                                                 \n",
            " max_pooling2d_6 (MaxPoolin  (1, 14, 14, 16)           0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_7 (Conv2D)           (1, 14, 14, 32)           4640      \n",
            "                                                                 \n",
            " max_pooling2d_7 (MaxPoolin  (1, 7, 7, 32)             0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_8 (Conv2D)           (1, 7, 7, 64)             18496     \n",
            "                                                                 \n",
            " max_pooling2d_8 (MaxPoolin  (1, 3, 3, 64)             0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_2 (Dropout)         (1, 3, 3, 64)             0         \n",
            "                                                                 \n",
            " flatten_2 (Flatten)         (1, 576)                  0         \n",
            "                                                                 \n",
            " dense_4 (Dense)             (1, 64)                   36928     \n",
            "                                                                 \n",
            " dense_5 (Dense)             (1, 10)                   650       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 60874 (237.79 KB)\n",
            "Trainable params: 60874 (237.79 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n",
            "313/313 - 1s - loss: 0.2450 - accuracy: 0.9168 - 806ms/epoch - 3ms/step\n",
            "Restored model, accuracy: 91.68%\n",
            "313/313 [==============================] - 1s 2ms/step\n",
            "(10000, 10)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Saving ML model as .h5 format\n",
        "\n",
        "cnn_model_1.save('cnn_fashion_mnist_model.h5')\n",
        "# Recreate the exact same model, including its weights and the optimizer\n",
        "new_model = tf.keras.models.load_model('cnn_fashion_mnist_model.h5')\n",
        "\n",
        "# Show the model architecture\n",
        "new_model.summary()\n",
        "\n",
        "loss, acc = new_model.evaluate(test_images, test_labels, verbose=2)\n",
        "print('Restored model, accuracy: {:5.2f}%'.format(100 * acc))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AQSLcrJg8E3x",
        "outputId": "af854fcc-9fc4-4dbe-bb1a-a0eb90ae9454"
      },
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_7\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d_6 (Conv2D)           (1, 28, 28, 16)           160       \n",
            "                                                                 \n",
            " max_pooling2d_6 (MaxPoolin  (1, 14, 14, 16)           0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_7 (Conv2D)           (1, 14, 14, 32)           4640      \n",
            "                                                                 \n",
            " max_pooling2d_7 (MaxPoolin  (1, 7, 7, 32)             0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " conv2d_8 (Conv2D)           (1, 7, 7, 64)             18496     \n",
            "                                                                 \n",
            " max_pooling2d_8 (MaxPoolin  (1, 3, 3, 64)             0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " dropout_2 (Dropout)         (1, 3, 3, 64)             0         \n",
            "                                                                 \n",
            " flatten_2 (Flatten)         (1, 576)                  0         \n",
            "                                                                 \n",
            " dense_4 (Dense)             (1, 64)                   36928     \n",
            "                                                                 \n",
            " dense_5 (Dense)             (1, 10)                   650       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 60874 (237.79 KB)\n",
            "Trainable params: 60874 (237.79 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n",
            "313/313 - 1s - loss: 0.2450 - accuracy: 0.9168 - 1s/epoch - 3ms/step\n",
            "Restored model, accuracy: 91.68%\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Source from links from which I learnt to code this -\n",
        "\n",
        "# https://github.com/tensorflow/docs/blob/master/site/en/tutorials/images/classification.ipynb\n",
        "# fashion mnist loading details - https://github.com/tensorflow/docs/blob/master/site/en/tutorials/keras/classification.ipynb\n",
        "# https://www.tensorflow.org/tutorials/images/cnn"
      ],
      "metadata": {
        "id": "TzGPeRJt0w19"
      },
      "execution_count": 34,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Code by Ayush Agarwal , IIT BHU ECE'24, Infoedge DS"
      ],
      "metadata": {
        "id": "H3rp3idK5-Lt"
      }
    }
  ]
}