DeepLearning1.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": "code",
      "execution_count": 2,
      "metadata": {
        "id": "ndQISSFJow6B"
      },
      "outputs": [],
      "source": [
        "import keras\n",
        "import pandas as pd\n",
        "import numpy as np"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import keras.datasets\n",
        "from keras.preprocessing import image_dataset_from_directory\n",
        "from keras import Sequential, layers, models, callbacks, losses, optimizers, metrics"
      ],
      "metadata": {
        "id": "Cesi8ykEo3O0"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "(X_train, y_train),(X_test, y_test) = keras.datasets.fashion_mnist.load_data()"
      ],
      "metadata": {
        "id": "0SIxX7b6pLqA",
        "outputId": "d68cb912-8fca-4e90-cf05-863b2c9861eb",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
            "\u001b[1m29515/29515\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
            "\u001b[1m26421880/26421880\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
            "\u001b[1m5148/5148\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
            "\u001b[1m4422102/4422102\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 0us/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_train"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5XCS5Es7rHM6",
        "outputId": "a5f3c8b9-1d5d-4431-b51d-e7edb65d3bff"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([9, 0, 0, ..., 3, 0, 5], dtype=uint8)"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_train.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_nUZmN0qrcL-",
        "outputId": "6bb2e0c2-3832-4f96-cc07-2d2497d786dc"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 28, 28)"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_train[0].shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AJEBq9venFiq",
        "outputId": "7b1e9005-e6ae-4ff2-aad0-0180f0cc81fc"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(28, 28)"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#no need to flatten just normalise\n",
        "X_train = X_train.astype('float32')/255.0 #dividing by pixels max value\n",
        "X_test = X_test.astype('float32')/255.0 #dividing by pixels max"
      ],
      "metadata": {
        "id": "HTm9d5Qorjsj"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**for CNN the expected dimension is 4D as a result, we expand dimension, (batch_size, height, width, channels).\n",
        "done to understand the grey scale of image**"
      ],
      "metadata": {
        "id": "Njjduj4pWsqw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "X_train = np.expand_dims(X_train, -1)\n",
        "X_test = np.expand_dims(X_test, -1)"
      ],
      "metadata": {
        "id": "3Gs60-9crzL1"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "***flatten during before dense layers, EXPAND during before CNN***"
      ],
      "metadata": {
        "id": "3cdJ05KcYd-_"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "X_train.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lYTnS7TkW2rJ",
        "outputId": "66b35874-bb6a-4e58-e022-52e4f1b9398a"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 28, 28, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "image = X_train[0]  # Extract the first image\n",
        "\n",
        "# Plot the image\n",
        "plt.imshow(image, cmap = 'gray')  # Display in grayscale\n",
        "plt.title(\"First Image\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 452
        },
        "id": "BSw_DjrYcUNR",
        "outputId": "d072ba66-2174-4808-816c-3497ab9b3a40"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "type(X_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dHcCUktKXEEd",
        "outputId": "7b52c51b-fd64-4e0d-f82f-227dff3651fe"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "numpy.ndarray"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_train[0].shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K3X4KasSYpCf",
        "outputId": "0f15d660-372b-4715-ca07-ac14cbaba787"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(28, 28, 1)"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_train[0]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YQUcQ5ROc58W",
        "outputId": "4fbce5a6-7ac6-4955-b595-f08e6c002bd6"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.00392157],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.05098039],\n",
              "        [0.28627452],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.00392157],\n",
              "        [0.01568628],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.00392157],\n",
              "        [0.00392157],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.01176471],\n",
              "        [0.        ],\n",
              "        [0.14117648],\n",
              "        [0.53333336],\n",
              "        [0.49803922],\n",
              "        [0.24313726],\n",
              "        [0.21176471],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.00392157],\n",
              "        [0.01176471],\n",
              "        [0.01568628],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.01176471]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.02352941],\n",
              "        [0.        ],\n",
              "        [0.4       ],\n",
              "        [0.8       ],\n",
              "        [0.6901961 ],\n",
              "        [0.5254902 ],\n",
              "        [0.5647059 ],\n",
              "        [0.48235294],\n",
              "        [0.09019608],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.04705882],\n",
              "        [0.03921569],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.60784316],\n",
              "        [0.9254902 ],\n",
              "        [0.8117647 ],\n",
              "        [0.69803923],\n",
              "        [0.41960785],\n",
              "        [0.6117647 ],\n",
              "        [0.6313726 ],\n",
              "        [0.42745098],\n",
              "        [0.2509804 ],\n",
              "        [0.09019608],\n",
              "        [0.3019608 ],\n",
              "        [0.50980395],\n",
              "        [0.28235295],\n",
              "        [0.05882353]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.00392157],\n",
              "        [0.        ],\n",
              "        [0.27058825],\n",
              "        [0.8117647 ],\n",
              "        [0.8745098 ],\n",
              "        [0.85490197],\n",
              "        [0.84705883],\n",
              "        [0.84705883],\n",
              "        [0.6392157 ],\n",
              "        [0.49803922],\n",
              "        [0.4745098 ],\n",
              "        [0.47843137],\n",
              "        [0.57254905],\n",
              "        [0.5529412 ],\n",
              "        [0.34509805],\n",
              "        [0.6745098 ],\n",
              "        [0.25882354]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.00392157],\n",
              "        [0.00392157],\n",
              "        [0.00392157],\n",
              "        [0.        ],\n",
              "        [0.78431374],\n",
              "        [0.9098039 ],\n",
              "        [0.9098039 ],\n",
              "        [0.9137255 ],\n",
              "        [0.8980392 ],\n",
              "        [0.8745098 ],\n",
              "        [0.8745098 ],\n",
              "        [0.84313726],\n",
              "        [0.8352941 ],\n",
              "        [0.6431373 ],\n",
              "        [0.49803922],\n",
              "        [0.48235294],\n",
              "        [0.76862746],\n",
              "        [0.8980392 ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.7176471 ],\n",
              "        [0.88235295],\n",
              "        [0.84705883],\n",
              "        [0.8745098 ],\n",
              "        [0.89411765],\n",
              "        [0.92156863],\n",
              "        [0.8901961 ],\n",
              "        [0.8784314 ],\n",
              "        [0.87058824],\n",
              "        [0.8784314 ],\n",
              "        [0.8666667 ],\n",
              "        [0.8745098 ],\n",
              "        [0.9607843 ],\n",
              "        [0.6784314 ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.75686276],\n",
              "        [0.89411765],\n",
              "        [0.85490197],\n",
              "        [0.8352941 ],\n",
              "        [0.7764706 ],\n",
              "        [0.7058824 ],\n",
              "        [0.83137256],\n",
              "        [0.8235294 ],\n",
              "        [0.827451  ],\n",
              "        [0.8352941 ],\n",
              "        [0.8745098 ],\n",
              "        [0.8627451 ],\n",
              "        [0.9529412 ],\n",
              "        [0.7921569 ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.00392157],\n",
              "        [0.01176471],\n",
              "        [0.        ],\n",
              "        [0.04705882],\n",
              "        [0.85882354],\n",
              "        [0.8627451 ],\n",
              "        [0.83137256],\n",
              "        [0.85490197],\n",
              "        [0.7529412 ],\n",
              "        [0.6627451 ],\n",
              "        [0.8901961 ],\n",
              "        [0.8156863 ],\n",
              "        [0.85490197],\n",
              "        [0.8784314 ],\n",
              "        [0.83137256],\n",
              "        [0.8862745 ],\n",
              "        [0.77254903],\n",
              "        [0.81960785],\n",
              "        [0.20392157]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.02352941],\n",
              "        [0.        ],\n",
              "        [0.3882353 ],\n",
              "        [0.95686275],\n",
              "        [0.87058824],\n",
              "        [0.8627451 ],\n",
              "        [0.85490197],\n",
              "        [0.79607844],\n",
              "        [0.7764706 ],\n",
              "        [0.8666667 ],\n",
              "        [0.84313726],\n",
              "        [0.8352941 ],\n",
              "        [0.87058824],\n",
              "        [0.8627451 ],\n",
              "        [0.9607843 ],\n",
              "        [0.46666667],\n",
              "        [0.654902  ],\n",
              "        [0.21960784]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.01568628],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.21568628],\n",
              "        [0.9254902 ],\n",
              "        [0.89411765],\n",
              "        [0.9019608 ],\n",
              "        [0.89411765],\n",
              "        [0.9411765 ],\n",
              "        [0.9098039 ],\n",
              "        [0.8352941 ],\n",
              "        [0.85490197],\n",
              "        [0.8745098 ],\n",
              "        [0.91764706],\n",
              "        [0.8509804 ],\n",
              "        [0.8509804 ],\n",
              "        [0.81960785],\n",
              "        [0.36078432],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.00392157],\n",
              "        [0.01568628],\n",
              "        [0.02352941],\n",
              "        [0.02745098],\n",
              "        [0.00784314],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.92941177],\n",
              "        [0.8862745 ],\n",
              "        [0.8509804 ],\n",
              "        [0.8745098 ],\n",
              "        [0.87058824],\n",
              "        [0.85882354],\n",
              "        [0.87058824],\n",
              "        [0.8666667 ],\n",
              "        [0.84705883],\n",
              "        [0.8745098 ],\n",
              "        [0.8980392 ],\n",
              "        [0.84313726],\n",
              "        [0.85490197],\n",
              "        [1.        ],\n",
              "        [0.3019608 ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.01176471],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.24313726],\n",
              "        [0.5686275 ],\n",
              "        [0.8       ],\n",
              "        [0.89411765],\n",
              "        [0.8117647 ],\n",
              "        [0.8352941 ],\n",
              "        [0.8666667 ],\n",
              "        [0.85490197],\n",
              "        [0.8156863 ],\n",
              "        [0.827451  ],\n",
              "        [0.85490197],\n",
              "        [0.8784314 ],\n",
              "        [0.8745098 ],\n",
              "        [0.85882354],\n",
              "        [0.84313726],\n",
              "        [0.8784314 ],\n",
              "        [0.95686275],\n",
              "        [0.62352943],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.07058824],\n",
              "        [0.17254902],\n",
              "        [0.32156864],\n",
              "        [0.41960785],\n",
              "        [0.7411765 ],\n",
              "        [0.89411765],\n",
              "        [0.8627451 ],\n",
              "        [0.87058824],\n",
              "        [0.8509804 ],\n",
              "        [0.8862745 ],\n",
              "        [0.78431374],\n",
              "        [0.8039216 ],\n",
              "        [0.827451  ],\n",
              "        [0.9019608 ],\n",
              "        [0.8784314 ],\n",
              "        [0.91764706],\n",
              "        [0.6901961 ],\n",
              "        [0.7372549 ],\n",
              "        [0.98039216],\n",
              "        [0.972549  ],\n",
              "        [0.9137255 ],\n",
              "        [0.93333334],\n",
              "        [0.84313726],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.22352941],\n",
              "        [0.73333335],\n",
              "        [0.8156863 ],\n",
              "        [0.8784314 ],\n",
              "        [0.8666667 ],\n",
              "        [0.8784314 ],\n",
              "        [0.8156863 ],\n",
              "        [0.8       ],\n",
              "        [0.8392157 ],\n",
              "        [0.8156863 ],\n",
              "        [0.81960785],\n",
              "        [0.78431374],\n",
              "        [0.62352943],\n",
              "        [0.9607843 ],\n",
              "        [0.75686276],\n",
              "        [0.80784315],\n",
              "        [0.8745098 ],\n",
              "        [1.        ],\n",
              "        [1.        ],\n",
              "        [0.8666667 ],\n",
              "        [0.91764706],\n",
              "        [0.8666667 ],\n",
              "        [0.827451  ],\n",
              "        [0.8627451 ],\n",
              "        [0.9098039 ],\n",
              "        [0.9647059 ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.01176471],\n",
              "        [0.7921569 ],\n",
              "        [0.89411765],\n",
              "        [0.8784314 ],\n",
              "        [0.8666667 ],\n",
              "        [0.827451  ],\n",
              "        [0.827451  ],\n",
              "        [0.8392157 ],\n",
              "        [0.8039216 ],\n",
              "        [0.8039216 ],\n",
              "        [0.8039216 ],\n",
              "        [0.8627451 ],\n",
              "        [0.9411765 ],\n",
              "        [0.3137255 ],\n",
              "        [0.5882353 ],\n",
              "        [1.        ],\n",
              "        [0.8980392 ],\n",
              "        [0.8666667 ],\n",
              "        [0.7372549 ],\n",
              "        [0.6039216 ],\n",
              "        [0.7490196 ],\n",
              "        [0.8235294 ],\n",
              "        [0.8       ],\n",
              "        [0.81960785],\n",
              "        [0.87058824],\n",
              "        [0.89411765],\n",
              "        [0.88235295],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.38431373],\n",
              "        [0.9137255 ],\n",
              "        [0.7764706 ],\n",
              "        [0.8235294 ],\n",
              "        [0.87058824],\n",
              "        [0.8980392 ],\n",
              "        [0.8980392 ],\n",
              "        [0.91764706],\n",
              "        [0.9764706 ],\n",
              "        [0.8627451 ],\n",
              "        [0.7607843 ],\n",
              "        [0.84313726],\n",
              "        [0.8509804 ],\n",
              "        [0.94509804],\n",
              "        [0.25490198],\n",
              "        [0.28627452],\n",
              "        [0.41568628],\n",
              "        [0.45882353],\n",
              "        [0.65882355],\n",
              "        [0.85882354],\n",
              "        [0.8666667 ],\n",
              "        [0.84313726],\n",
              "        [0.8509804 ],\n",
              "        [0.8745098 ],\n",
              "        [0.8745098 ],\n",
              "        [0.8784314 ],\n",
              "        [0.8980392 ],\n",
              "        [0.11372549]],\n",
              "\n",
              "       [[0.29411766],\n",
              "        [0.8       ],\n",
              "        [0.83137256],\n",
              "        [0.8       ],\n",
              "        [0.75686276],\n",
              "        [0.8039216 ],\n",
              "        [0.827451  ],\n",
              "        [0.88235295],\n",
              "        [0.84705883],\n",
              "        [0.7254902 ],\n",
              "        [0.77254903],\n",
              "        [0.80784315],\n",
              "        [0.7764706 ],\n",
              "        [0.8352941 ],\n",
              "        [0.9411765 ],\n",
              "        [0.7647059 ],\n",
              "        [0.8901961 ],\n",
              "        [0.9607843 ],\n",
              "        [0.9372549 ],\n",
              "        [0.8745098 ],\n",
              "        [0.85490197],\n",
              "        [0.83137256],\n",
              "        [0.81960785],\n",
              "        [0.87058824],\n",
              "        [0.8627451 ],\n",
              "        [0.8666667 ],\n",
              "        [0.9019608 ],\n",
              "        [0.2627451 ]],\n",
              "\n",
              "       [[0.1882353 ],\n",
              "        [0.79607844],\n",
              "        [0.7176471 ],\n",
              "        [0.7607843 ],\n",
              "        [0.8352941 ],\n",
              "        [0.77254903],\n",
              "        [0.7254902 ],\n",
              "        [0.74509805],\n",
              "        [0.7607843 ],\n",
              "        [0.7529412 ],\n",
              "        [0.7921569 ],\n",
              "        [0.8392157 ],\n",
              "        [0.85882354],\n",
              "        [0.8666667 ],\n",
              "        [0.8627451 ],\n",
              "        [0.9254902 ],\n",
              "        [0.88235295],\n",
              "        [0.84705883],\n",
              "        [0.78039217],\n",
              "        [0.80784315],\n",
              "        [0.7294118 ],\n",
              "        [0.70980394],\n",
              "        [0.69411767],\n",
              "        [0.6745098 ],\n",
              "        [0.70980394],\n",
              "        [0.8039216 ],\n",
              "        [0.80784315],\n",
              "        [0.4509804 ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.47843137],\n",
              "        [0.85882354],\n",
              "        [0.75686276],\n",
              "        [0.7019608 ],\n",
              "        [0.67058825],\n",
              "        [0.7176471 ],\n",
              "        [0.76862746],\n",
              "        [0.8       ],\n",
              "        [0.8235294 ],\n",
              "        [0.8352941 ],\n",
              "        [0.8117647 ],\n",
              "        [0.827451  ],\n",
              "        [0.8235294 ],\n",
              "        [0.78431374],\n",
              "        [0.76862746],\n",
              "        [0.7607843 ],\n",
              "        [0.7490196 ],\n",
              "        [0.7647059 ],\n",
              "        [0.7490196 ],\n",
              "        [0.7764706 ],\n",
              "        [0.7529412 ],\n",
              "        [0.6901961 ],\n",
              "        [0.6117647 ],\n",
              "        [0.654902  ],\n",
              "        [0.69411767],\n",
              "        [0.8235294 ],\n",
              "        [0.36078432]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.2901961 ],\n",
              "        [0.7411765 ],\n",
              "        [0.83137256],\n",
              "        [0.7490196 ],\n",
              "        [0.6862745 ],\n",
              "        [0.6745098 ],\n",
              "        [0.6862745 ],\n",
              "        [0.70980394],\n",
              "        [0.7254902 ],\n",
              "        [0.7372549 ],\n",
              "        [0.7411765 ],\n",
              "        [0.7372549 ],\n",
              "        [0.75686276],\n",
              "        [0.7764706 ],\n",
              "        [0.8       ],\n",
              "        [0.81960785],\n",
              "        [0.8235294 ],\n",
              "        [0.8235294 ],\n",
              "        [0.827451  ],\n",
              "        [0.7372549 ],\n",
              "        [0.7372549 ],\n",
              "        [0.7607843 ],\n",
              "        [0.7529412 ],\n",
              "        [0.84705883],\n",
              "        [0.6666667 ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.00784314],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.25882354],\n",
              "        [0.78431374],\n",
              "        [0.87058824],\n",
              "        [0.92941177],\n",
              "        [0.9372549 ],\n",
              "        [0.9490196 ],\n",
              "        [0.9647059 ],\n",
              "        [0.9529412 ],\n",
              "        [0.95686275],\n",
              "        [0.8666667 ],\n",
              "        [0.8627451 ],\n",
              "        [0.75686276],\n",
              "        [0.7490196 ],\n",
              "        [0.7019608 ],\n",
              "        [0.7137255 ],\n",
              "        [0.7137255 ],\n",
              "        [0.70980394],\n",
              "        [0.6901961 ],\n",
              "        [0.6509804 ],\n",
              "        [0.65882355],\n",
              "        [0.3882353 ],\n",
              "        [0.22745098],\n",
              "        [0.        ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.15686275],\n",
              "        [0.23921569],\n",
              "        [0.17254902],\n",
              "        [0.28235295],\n",
              "        [0.16078432],\n",
              "        [0.13725491],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ]],\n",
              "\n",
              "       [[0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ],\n",
              "        [0.        ]]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "num_classes = 10\n",
        "#here num_classes is number of output classes\n",
        "input_shape = (28,28,1)"
      ],
      "metadata": {
        "id": "pSGT72o9YwXW"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = keras.Sequential(\n",
        "    [\n",
        "     #defining 1st input layer\n",
        "     keras.layers.Input(shape = input_shape),\n",
        "     #defining CNN with relu 2 layers 64 filters\n",
        "     keras.layers.Conv2D(64, kernel_size = (3,3), activation = 'relu'),\n",
        "     keras.layers.Conv2D(64, kernel_size = (3,3), activation = 'relu'),\n",
        "     #Max pooling\n",
        "     keras.layers.MaxPooling2D(pool_size = (2,2)),\n",
        "     #defining CNN with relu 2 layers 128 filters\n",
        "     keras.layers.Conv2D(128, kernel_size = (3,3), activation = 'relu'),\n",
        "     keras.layers.Conv2D(128, kernel_size = (3,3), activation = 'relu'),\n",
        "     #average pooling\n",
        "     keras.layers.GlobalAveragePooling2D(),\n",
        "     #dropout\n",
        "     keras.layers.Dropout(0.5),\n",
        "     #full CNN connected to last dense layer with softmax\n",
        "     keras.layers.Dense(num_classes, activation = 'softmax'),\n",
        "     ]\n",
        ")"
      ],
      "metadata": {
        "id": "91ROiPlTZftA"
      },
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 442
        },
        "id": "mHlzQCeergCI",
        "outputId": "d88da302-fd43-4ab3-d087-af51427f1b9b"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1mModel: \"sequential\"\u001b[0m\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential\"</span>\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n",
              "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                        \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape               \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m        Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n",
              "│ conv2d (\u001b[38;5;33mConv2D\u001b[0m)                      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m26\u001b[0m, \u001b[38;5;34m26\u001b[0m, \u001b[38;5;34m64\u001b[0m)          │             \u001b[38;5;34m640\u001b[0m │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ conv2d_1 (\u001b[38;5;33mConv2D\u001b[0m)                    │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m24\u001b[0m, \u001b[38;5;34m24\u001b[0m, \u001b[38;5;34m64\u001b[0m)          │          \u001b[38;5;34m36,928\u001b[0m │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ max_pooling2d (\u001b[38;5;33mMaxPooling2D\u001b[0m)         │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m12\u001b[0m, \u001b[38;5;34m12\u001b[0m, \u001b[38;5;34m64\u001b[0m)          │               \u001b[38;5;34m0\u001b[0m │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ conv2d_2 (\u001b[38;5;33mConv2D\u001b[0m)                    │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m, \u001b[38;5;34m10\u001b[0m, \u001b[38;5;34m128\u001b[0m)         │          \u001b[38;5;34m73,856\u001b[0m │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ conv2d_3 (\u001b[38;5;33mConv2D\u001b[0m)                    │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m8\u001b[0m, \u001b[38;5;34m8\u001b[0m, \u001b[38;5;34m128\u001b[0m)           │         \u001b[38;5;34m147,584\u001b[0m │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ global_average_pooling2d             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m)                 │               \u001b[38;5;34m0\u001b[0m │\n",
              "│ (\u001b[38;5;33mGlobalAveragePooling2D\u001b[0m)             │                             │                 │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ dropout (\u001b[38;5;33mDropout\u001b[0m)                    │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m)                 │               \u001b[38;5;34m0\u001b[0m │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ dense (\u001b[38;5;33mDense\u001b[0m)                        │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m10\u001b[0m)                  │           \u001b[38;5;34m1,290\u001b[0m │\n",
              "└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n",
              "┃<span style=\"font-weight: bold\"> Layer (type)                         </span>┃<span style=\"font-weight: bold\"> Output Shape                </span>┃<span style=\"font-weight: bold\">         Param # </span>┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n",
              "│ conv2d (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)                      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">26</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">26</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)          │             <span style=\"color: #00af00; text-decoration-color: #00af00\">640</span> │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ conv2d_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)                    │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">24</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">24</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)          │          <span style=\"color: #00af00; text-decoration-color: #00af00\">36,928</span> │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ max_pooling2d (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">MaxPooling2D</span>)         │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">12</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">12</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)          │               <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ conv2d_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)                    │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)         │          <span style=\"color: #00af00; text-decoration-color: #00af00\">73,856</span> │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ conv2d_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)                    │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">8</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">8</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)           │         <span style=\"color: #00af00; text-decoration-color: #00af00\">147,584</span> │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ global_average_pooling2d             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)                 │               <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">GlobalAveragePooling2D</span>)             │                             │                 │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ dropout (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)                    │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)                 │               <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ dense (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                        │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>)                  │           <span style=\"color: #00af00; text-decoration-color: #00af00\">1,290</span> │\n",
              "└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m260,298\u001b[0m (1016.79 KB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">260,298</span> (1016.79 KB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m260,298\u001b[0m (1016.79 KB)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">260,298</span> (1016.79 KB)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
            ],
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n",
              "</pre>\n"
            ]
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(\n",
        "              loss = keras.losses.SparseCategoricalCrossentropy(),\n",
        "              optimizer = keras.optimizers.Adam(learning_rate = 0.001),\n",
        "              metrics = [\n",
        "                  keras.metrics.SparseCategoricalAccuracy(name = 'categ'), 'accuracy'\n",
        "              ])"
      ],
      "metadata": {
        "id": "KYLQM10SsKyN"
      },
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**fitting model**"
      ],
      "metadata": {
        "id": "UbqSEnbAz13R"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "callbacks = [\n",
        "    keras.callbacks.ModelCheckpoint(\"save_at_{epoch}.keras\"),\n",
        "    keras.callbacks.EarlyStopping(patience = 2, monitor = 'val_loss'),\n",
        "]\n",
        "history = model.fit(X_train, y_train, batch_size = 128, epochs = 20, validation_split = 0.15, callbacks = callbacks)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5A1vfaNSnAr8",
        "outputId": "ccf98205-e02d-4b7c-bd0e-abe3bee5b994"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 17ms/step - accuracy: 0.8491 - categ: 0.8491 - loss: 0.4199 - val_accuracy: 0.8749 - val_categ: 0.8749 - val_loss: 0.3426\n",
            "Epoch 2/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 11ms/step - accuracy: 0.8597 - categ: 0.8597 - loss: 0.3866 - val_accuracy: 0.8841 - val_categ: 0.8841 - val_loss: 0.3240\n",
            "Epoch 3/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 12ms/step - accuracy: 0.8688 - categ: 0.8688 - loss: 0.3737 - val_accuracy: 0.8770 - val_categ: 0.8770 - val_loss: 0.3315\n",
            "Epoch 4/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 11ms/step - accuracy: 0.8730 - categ: 0.8730 - loss: 0.3581 - val_accuracy: 0.8924 - val_categ: 0.8924 - val_loss: 0.2923\n",
            "Epoch 5/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 12ms/step - accuracy: 0.8807 - categ: 0.8807 - loss: 0.3419 - val_accuracy: 0.8920 - val_categ: 0.8920 - val_loss: 0.2900\n",
            "Epoch 6/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 11ms/step - accuracy: 0.8824 - categ: 0.8824 - loss: 0.3316 - val_accuracy: 0.8883 - val_categ: 0.8883 - val_loss: 0.2993\n",
            "Epoch 7/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 13ms/step - accuracy: 0.8912 - categ: 0.8912 - loss: 0.3083 - val_accuracy: 0.9030 - val_categ: 0.9030 - val_loss: 0.2750\n",
            "Epoch 8/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 12ms/step - accuracy: 0.8892 - categ: 0.8892 - loss: 0.3108 - val_accuracy: 0.9011 - val_categ: 0.9011 - val_loss: 0.2675\n",
            "Epoch 9/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 11ms/step - accuracy: 0.8943 - categ: 0.8943 - loss: 0.2939 - val_accuracy: 0.9054 - val_categ: 0.9054 - val_loss: 0.2618\n",
            "Epoch 10/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 12ms/step - accuracy: 0.8959 - categ: 0.8959 - loss: 0.2880 - val_accuracy: 0.9061 - val_categ: 0.9061 - val_loss: 0.2561\n",
            "Epoch 11/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 11ms/step - accuracy: 0.9008 - categ: 0.9008 - loss: 0.2805 - val_accuracy: 0.9064 - val_categ: 0.9064 - val_loss: 0.2495\n",
            "Epoch 12/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 11ms/step - accuracy: 0.9047 - categ: 0.9047 - loss: 0.2717 - val_accuracy: 0.9112 - val_categ: 0.9112 - val_loss: 0.2440\n",
            "Epoch 13/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 12ms/step - accuracy: 0.9048 - categ: 0.9048 - loss: 0.2660 - val_accuracy: 0.9109 - val_categ: 0.9109 - val_loss: 0.2513\n",
            "Epoch 14/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 11ms/step - accuracy: 0.9087 - categ: 0.9087 - loss: 0.2588 - val_accuracy: 0.9142 - val_categ: 0.9142 - val_loss: 0.2373\n",
            "Epoch 15/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 12ms/step - accuracy: 0.9071 - categ: 0.9071 - loss: 0.2600 - val_accuracy: 0.9128 - val_categ: 0.9128 - val_loss: 0.2445\n",
            "Epoch 16/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 13ms/step - accuracy: 0.9101 - categ: 0.9101 - loss: 0.2491 - val_accuracy: 0.9133 - val_categ: 0.9133 - val_loss: 0.2352\n",
            "Epoch 17/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 11ms/step - accuracy: 0.9132 - categ: 0.9132 - loss: 0.2432 - val_accuracy: 0.9157 - val_categ: 0.9157 - val_loss: 0.2276\n",
            "Epoch 18/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 12ms/step - accuracy: 0.9148 - categ: 0.9148 - loss: 0.2380 - val_accuracy: 0.9149 - val_categ: 0.9149 - val_loss: 0.2372\n",
            "Epoch 19/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 11ms/step - accuracy: 0.9138 - categ: 0.9138 - loss: 0.2382 - val_accuracy: 0.9173 - val_categ: 0.9173 - val_loss: 0.2248\n",
            "Epoch 20/20\n",
            "\u001b[1m399/399\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 12ms/step - accuracy: 0.9186 - categ: 0.9186 - loss: 0.2236 - val_accuracy: 0.9186 - val_categ: 0.9186 - val_loss: 0.2226\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.save('final_mnist_model.keras')"
      ],
      "metadata": {
        "id": "XetiTLgP060e"
      },
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['accuracy'])\n",
        "plt.plot(history.history['val_accuracy'])\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['Train','Test'])\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "uPNlCpIG2lHY",
        "outputId": "31b244a9-a5b2-4555-f1b2-0a18471b6df3"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['Train','Test'])\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "id": "KGqXJI5f3QPO",
        "outputId": "14518a57-b043-477d-d430-0c4f9fc9fa3a"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.save('trained_model_mnist.h5')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3rvYwsl93nh2",
        "outputId": "9fb6aab9-9069-4379-9258-fc875f1af94c"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:absl:You are saving your model as an HDF5 file via `model.save()` or `keras.saving.save_model(model)`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')` or `keras.saving.save_model(model, 'my_model.keras')`. \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "8eDGaMve35Uu"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}