Untitled4.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "bpUkfO7GnU2R",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "95d9e275-910c-484d-8001-ccd5ad34eedd"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting visualkeras\n",
            "  Downloading visualkeras-0.0.2-py3-none-any.whl (12 kB)\n",
            "Requirement already satisfied: numpy>=1.18.1 in /usr/local/lib/python3.8/dist-packages (from visualkeras) (1.22.4)\n",
            "Collecting aggdraw>=1.3.11\n",
            "  Downloading aggdraw-1.3.16-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (992 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m992.0/992.0 KB\u001b[0m \u001b[31m50.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.8/dist-packages (from visualkeras) (7.1.2)\n",
            "Installing collected packages: aggdraw, visualkeras\n",
            "Successfully installed aggdraw-1.3.16 visualkeras-0.0.2\n"
          ]
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import tensorflow as tf\n",
        "!pip install visualkeras\n",
        "import visualkeras\n",
        "from PIL import ImageFont\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "from tensorflow.keras import layers\n",
        "\n",
        "\n",
        "model = tf.keras.Sequential([\n",
        "      tf.keras.layers.Flatten(input_shape=(28, 28)),\n",
        "      tf.keras.layers.Dense(128, activation='relu', kernel_regularizer=tf.keras.regularizers.l2(0.001)),\n",
        "      tf.keras.layers.Dense(128, activation='relu'),\n",
        "      tf.keras.layers.Dense(128, activation='relu'),\n",
        "      tf.keras.layers.Dense(128, activation='relu'),\n",
        "      tf.keras.layers.Dense(128, activation='relu'),\n",
        "      tf.keras.layers.Dropout(0.5),\n",
        "      tf.keras.layers.Dense(10, activation='softmax')])\n",
        "  \n"
      ],
      "metadata": {
        "id": "wL1DksO8z3BE"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# tf.keras.utils.plot_model(model, to_file='model_plot.png', show_shapes=True, show_layer_names=True)"
      ],
      "metadata": {
        "id": "v1UPosXr0C5b"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model2 = tf.keras.Sequential([\n",
        "      tf.keras.layers.Flatten(input_shape=(28, 28)),\n",
        "      tf.keras.layers.Dense(512, activation='relu', kernel_regularizer=tf.keras.regularizers.l2(0.001) ),\n",
        "      tf.keras.layers.Dropout(0.3),\n",
        "      tf.keras.layers.Dense(256, activation='relu'),\n",
        "      tf.keras.layers.Dropout(0.3),\n",
        "      tf.keras.layers.Dense(128, activation='relu'),\n",
        "      tf.keras.layers.Dropout(0.3),\n",
        "      tf.keras.layers.Dense(64, activation='relu'),\n",
        "      tf.keras.layers.Dropout(0.3),\n",
        "      tf.keras.layers.Dense(10, activation='softmax')\n",
        "  ])"
      ],
      "metadata": {
        "id": "DjoDuqg206Yq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import visualkeras\n",
        "visualkeras.layered_view(model2, legend=True)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 80
        },
        "id": "aTpEujr_2oeN",
        "outputId": "6f3f3680-0a66-4155-b767-8366e6bb6978"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<PIL.Image.Image image mode=RGBA size=449x63 at 0x7F77743C57F0>"
            ],
            "image/png": "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\n"
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tf.keras.utils.plot_model(model2, to_file='model_plot.png', show_shapes=True, show_layer_names=True)"
      ],
      "metadata": {
        "id": "yc-DeZyj1GNq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "0BLE2CUeg7rw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# import matplotlib.pyplot as plt  # to visualize data and draw plots\n",
        "# import numpy as np  # to use numpy arrays\n",
        "# import tensorflow as tf  # to specify and run computation graphs\n",
        "# from sklearn.metrics import confusion_matrix\n",
        "# from sklearn.model_selection import train_test_split\n",
        "# from tensorflow.keras.callbacks import EarlyStopping\n",
        "# from tensorflow.keras.datasets import cifar100\n",
        "# from tensorflow.keras.layers import *\n",
        "# from tensorflow.keras.models import Model, Sequential\n",
        "# from tensorflow.keras.regularizers import l1, l2\n",
        "# from tensorflow.keras.utils import to_categorical\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers, models\n",
        "# def Avhi_model():\n",
        "    # With Just Batch Normalization and L2 Regularization\n",
        "model = models.Sequential()\n",
        "\n",
        "model.add(Conv2D(input_shape=(32, 32, 3),kernel_size=(2, 2),padding=\"same\",strides=(2, 2),filters=32,kernel_regularizer=l2(0.001)))#,kernel_regularizer=l2(0.001)\n",
        "model.add(BatchNormalization())\n",
        "model.add(Activation(\"relu\"))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2), strides=(1, 1), padding=\"same\"))\n",
        "# model.add(Dropout(0.1))\n",
        "\n",
        "model.add(Conv2D(kernel_size=(2, 2), padding=\"same\", strides=(2, 2), filters=64))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Activation(\"relu\"))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2), strides=(1, 1), padding=\"same\"))\n",
        "\n",
        "model.add(Conv2D(kernel_size=(2, 2), padding=\"same\", strides=(2, 2), filters=128))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Activation(\"relu\"))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2), strides=(1, 1), padding=\"same\"))\n",
        "\n",
        "\n",
        "model.add(Conv2D(kernel_size=(2, 2), padding=\"same\", strides=(2, 2), filters=256))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Activation(\"relu\"))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2), strides=(1, 1), padding=\"same\"))\n",
        "# model.add(Dropout(0.1))\n",
        "\n",
        "model.add(Conv2D(kernel_size=(2, 2), padding=\"same\", strides=(2, 2), filters=512))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Activation(\"relu\"))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2), strides=(1, 1), padding=\"same\"))\n",
        "\n",
        "model.add(Conv2D(kernel_size=(2, 2), padding=\"same\", strides=(2, 2), filters=1024))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Activation(\"relu\"))\n",
        "model.add(MaxPooling2D(pool_size=(2, 2), strides=(1, 1), padding=\"same\"))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Dense(256))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Activation(\"relu\"))\n",
        "\n",
        "model.add(Dense(128))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Activation(\"relu\"))\n",
        "\n",
        "model.add(Dense(100, activation=\"softmax\"))\n",
        "\n",
        "model.summary()\n",
        "\n",
        "    # return model"
      ],
      "metadata": {
        "id": "QLSSgv9sg4lM"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "visualkeras.layered_view(model, legend=True)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 128
        },
        "id": "LKdc98wwg9WM",
        "outputId": "46195bca-87bf-4d74-fdb3-961f25241643"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<PIL.Image.Image image mode=RGBA size=1561x122 at 0x7F776682AD90>"
            ],
            "image/png": "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\n"
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    }
  ]
}