TP_batchnorm.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Ramyrahmeni/DeepLearning/blob/main/TP_batchnorm.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "B9mjkyMAEn_N"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras import datasets, layers, models\n",
        "# Load data set\n",
        "mnist = datasets.mnist"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "tCYgNBeyEn_S",
        "outputId": "7ecaa216-1ecc-4f26-92dc-7e89ff04e242"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "\u001b[1m11490434/11490434\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 0us/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "(train_images, train_labels), (test_images, test_labels) = datasets.mnist.load_data()\n",
        "\n",
        "plt.imshow(train_images[0], cmap='gray')\n",
        "plt.title(f'Label: {train_labels[0]}')\n",
        "plt.show()\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "mnist = datasets.mnist\n",
        "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
        "x_train, x_test = x_train / 255.0, x_test / 255.0\n",
        "model = models.Sequential()\n",
        "model.add(layers.Flatten(input_shape=(28, 28)))\n",
        "model.add(layers.Dense(512, activation=tf.nn.relu))\n",
        "model.add(layers.Dense(10, activation=tf.nn.softmax))\n",
        "model.compile(optimizer=\"adam\",\n",
        " loss=\"sparse_categorical_crossentropy\",\n",
        " metrics=[\"accuracy\"])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fdN3EAdmFBfv",
        "outputId": "668a98eb-df9b-4a4d-a701-063d2aa5d646"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/layers/reshaping/flatten.py:37: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
            "  super().__init__(**kwargs)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.fit(x_train, y_train, epochs=5)\n",
        "model.evaluate(x_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UK1GmYN1FyKf",
        "outputId": "03dbcb10-a207-4a6c-9a3f-0d4a28792795"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "\u001b[1m1875/1875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m15s\u001b[0m 7ms/step - accuracy: 0.9007 - loss: 0.3397\n",
            "Epoch 2/5\n",
            "\u001b[1m1875/1875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m14s\u001b[0m 7ms/step - accuracy: 0.9753 - loss: 0.0818\n",
            "Epoch 3/5\n",
            "\u001b[1m1875/1875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m21s\u001b[0m 7ms/step - accuracy: 0.9843 - loss: 0.0492\n",
            "Epoch 4/5\n",
            "\u001b[1m1875/1875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m19s\u001b[0m 6ms/step - accuracy: 0.9881 - loss: 0.0358\n",
            "Epoch 5/5\n",
            "\u001b[1m1875/1875\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m20s\u001b[0m 6ms/step - accuracy: 0.9928 - loss: 0.0231\n",
            "\u001b[1m313/313\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - accuracy: 0.9750 - loss: 0.0819\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.06795815378427505, 0.9793000221252441]"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "def add_noise(images, noise_factor=0.5):\n",
        "    noise = noise_factor * np.random.randn(*images.shape)\n",
        "    noisy_images = images + noise\n",
        "    noisy_images = np.clip(noisy_images, 0., 1.)\n",
        "    return noisy_images\n",
        "\n",
        "noisy_test_images = add_noise(test_images, noise_factor=0.5)\n",
        "\n",
        "n = 20\n",
        "plt.figure(figsize=(40, 2))\n",
        "for i in range(n):\n",
        "    plt.subplot(1, n, i + 1)\n",
        "    plt.imshow(noisy_test_images[i], cmap='gray')\n",
        "    plt.title(f\"Label: {test_labels[i]}\")\n",
        "    plt.axis('off')\n",
        "plt.show()\n",
        "\n",
        "test_loss, test_acc = model.evaluate(noisy_test_images, test_labels, verbose=2)\n",
        "print(f'\\nTest accuracy on noisy images: {test_acc:.2f}')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 145
        },
        "id": "LYvvO6IJGATz",
        "outputId": "67f286e1-7c6c-40a6-949a-fa11e5e097c7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 4000x200 with 20 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 - 1s - 2ms/step - accuracy: 0.6320 - loss: 1.6054\n",
            "\n",
            "Test accuracy on noisy images: 0.63\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "azb9tydCHkYr"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.11.4"
    },
    "colab": {
      "provenance": [],
      "include_colab_link": true
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}