Custom_loss_function_keras.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "authorship_tag": "ABX9TyMNdOx1lZkCPJ5DjsIv2s8Y",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Anant-mishra1729/Machine-Learning-Notebooks/blob/main/Custom_loss_function_keras.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras.datasets import mnist\n",
        "from tensorflow.keras.losses import Loss\n",
        "from tensorflow.keras import layers\n",
        "import matplotlib.pyplot as plt"
      ],
      "metadata": {
        "id": "Al1O4I1J7WtW"
      },
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "(X_train,y_train),(X_test,y_test) = mnist.load_data()\n",
        "X_train.shape, y_train.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ttaoys0e9hWL",
        "outputId": "c8bbff68-4520-4fe0-b702-d7b2ba04a001"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11490434/11490434 [==============================] - 0s 0us/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((60000, 28, 28), (60000,))"
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize = (10,4))\n",
        "for i in range(10):\n",
        "  ax = plt.subplot(2,5,i+1)\n",
        "  ax.imshow(X_train[i])\n",
        "  ax.set_axis_off()\n",
        "  ax.set_title(y_train[i])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 360
        },
        "id": "UOx8_Ti59om7",
        "outputId": "123b59b3-e93b-4465-9b88-0155426dcdf6"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x400 with 10 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_train = X_train / 255.0\n",
        "X_test = X_test / 255.0"
      ],
      "metadata": {
        "id": "pAe5Zbr094Jt"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# One hot encoding\n",
        "y_train = tf.keras.utils.to_categorical(y_train)\n",
        "y_test = tf.keras.utils.to_categorical(y_test)"
      ],
      "metadata": {
        "id": "tyxep3Q6-jnv"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "y_train"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "e1K_RoJB_syU",
        "outputId": "d66a915f-5b6d-4ca2-aa69-1faa3bc41d27"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0., 0., 0., ..., 0., 0., 0.],\n",
              "       [1., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       ...,\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 1., 0.]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_test"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VY3N4mO7_u71",
        "outputId": "b233f422-51f3-45dd-804f-98a77e7f6051"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0., 0., 0., ..., 1., 0., 0.],\n",
              "       [0., 0., 1., ..., 0., 0., 0.],\n",
              "       [0., 1., 0., ..., 0., 0., 0.],\n",
              "       ...,\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.],\n",
              "       [0., 0., 0., ..., 0., 0., 0.]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Creating custom loss"
      ],
      "metadata": {
        "id": "HhzgqsMwDKit"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the custom categorical cross-entropy loss function\n",
        "class CustomCategoricalCrossEntropy(Loss):\n",
        "    def __init__(self, name='custom_categorical_crossentropy'):\n",
        "        super().__init__(name=name)\n",
        "\n",
        "    def call(self, y_true, y_pred):\n",
        "        y_pred = tf.clip_by_value(y_pred, tf.keras.backend.epsilon(), 1 - tf.keras.backend.epsilon())\n",
        "        loss = -tf.reduce_sum(y_true * tf.math.log(y_pred), axis=-1)\n",
        "        return loss"
      ],
      "metadata": {
        "id": "InXIpoGiDKD0"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Creating keras model"
      ],
      "metadata": {
        "id": "hZsGROxe92V5"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def getModel():\n",
        "  model = tf.keras.Sequential([\n",
        "      layers.Conv2D(32,(3,3),activation = 'relu', input_shape = (28,28,1)),\n",
        "      layers.MaxPooling2D((2,2)),\n",
        "      layers.Conv2D(64,(3,3),activation = 'relu', input_shape = (28,28,1)),\n",
        "      layers.Conv2D(64,(3,3),activation = 'relu', input_shape = (28,28,1)),\n",
        "      layers.MaxPooling2D((2,2)),\n",
        "      layers.Flatten(),\n",
        "      layers.Dense(100,activation = 'relu'),\n",
        "      layers.Dense(10,activation = 'softmax')\n",
        "  ])\n",
        "  customloss = CustomCategoricalCrossEntropy()\n",
        "  model.compile(\n",
        "      optimizer = 'adam',\n",
        "      loss = customloss,\n",
        "      metrics = ['accuracy']\n",
        "  )\n",
        "  return model"
      ],
      "metadata": {
        "id": "FaeRyX2R-Ixc"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = getModel()\n",
        "model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "65tvnpBL_En0",
        "outputId": "08d6da5f-65a9-43c2-b752-2231d6de9b7b"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d (Conv2D)             (None, 26, 26, 32)        320       \n",
            "                                                                 \n",
            " max_pooling2d (MaxPooling2  (None, 13, 13, 32)        0         \n",
            " D)                                                              \n",
            "                                                                 \n",
            " conv2d_1 (Conv2D)           (None, 11, 11, 64)        18496     \n",
            "                                                                 \n",
            " conv2d_2 (Conv2D)           (None, 9, 9, 64)          36928     \n",
            "                                                                 \n",
            " max_pooling2d_1 (MaxPoolin  (None, 4, 4, 64)          0         \n",
            " g2D)                                                            \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 1024)              0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 100)               102500    \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 10)                1010      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 159254 (622.09 KB)\n",
            "Trainable params: 159254 (622.09 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "history = model.fit(X_train,y_train, epochs = 6, validation_data = (X_test, y_test))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "41jLMyRQ_HCn",
        "outputId": "17060c73-0def-4af8-df2d-4c12251064db"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/6\n",
            "1875/1875 [==============================] - 27s 5ms/step - loss: 0.1212 - accuracy: 0.9632 - val_loss: 0.0436 - val_accuracy: 0.9847\n",
            "Epoch 2/6\n",
            "1875/1875 [==============================] - 7s 4ms/step - loss: 0.0411 - accuracy: 0.9880 - val_loss: 0.0348 - val_accuracy: 0.9889\n",
            "Epoch 3/6\n",
            "1875/1875 [==============================] - 8s 4ms/step - loss: 0.0292 - accuracy: 0.9907 - val_loss: 0.0445 - val_accuracy: 0.9868\n",
            "Epoch 4/6\n",
            "1875/1875 [==============================] - 7s 4ms/step - loss: 0.0212 - accuracy: 0.9936 - val_loss: 0.0277 - val_accuracy: 0.9912\n",
            "Epoch 5/6\n",
            "1875/1875 [==============================] - 8s 4ms/step - loss: 0.0160 - accuracy: 0.9949 - val_loss: 0.0269 - val_accuracy: 0.9917\n",
            "Epoch 6/6\n",
            "1875/1875 [==============================] - 8s 4ms/step - loss: 0.0144 - accuracy: 0.9952 - val_loss: 0.0257 - val_accuracy: 0.9930\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Training history\n",
        "def showHistory(epochs, history):\n",
        "  # Accuracy\n",
        "  plt.subplot(1,2,1)\n",
        "  plt.plot(epochs,history['accuracy'], 'g'),\n",
        "  plt.plot(epochs,history['val_accuracy'], 'r'),\n",
        "  plt.title('Training vs Validation accuracy')\n",
        "  plt.legend(['Training accuracy','Validation accuracy'])\n",
        "  plt.xlabel(\"Epochs\")\n",
        "  plt.ylabel(\"Accuracy\")\n",
        "  # Loss\n",
        "  plt.subplot(1,2,2)\n",
        "  plt.plot(epochs, history['loss'],'g')\n",
        "  plt.plot(epochs, history['val_loss'],'r')\n",
        "  plt.title('Training vs Validation loss')\n",
        "  plt.legend(['Training loss','Validation loss'])\n",
        "  plt.xlabel(\"Epochs\")\n",
        "  plt.ylabel(\"Loss\")\n",
        "  plt.tight_layout()\n",
        "  plt.show()\n",
        "showHistory(range(1,7),history.history)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 487
        },
        "id": "WcIWC8xR_LW0",
        "outputId": "594fbde0-30da-44c5-c1b5-a70c4565fc99"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"Results:\")\n",
        "model.evaluate(X_test,y_test)"
      ],
      "metadata": {
        "id": "4HWHsaQ1BRsI",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "13e7a753-aeb3-450e-badd-f9142221a917"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Results:\n",
            "313/313 [==============================] - 1s 3ms/step - loss: 0.0257 - accuracy: 0.9930\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.025675052776932716, 0.9929999709129333]"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "4Zok18PM6jY6"
      },
      "execution_count": 13,
      "outputs": []
    }
  ]
}