NN_using_Functional_API.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "NN using Functional API.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "mqiHdvkMzAGh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# this file is to understand the basic regression problem using Functional API\n",
        "%tensorflow_version 2.x\n",
        "from sklearn.datasets import fetch_california_housing\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "import tensorflow\n",
        "from tensorflow import keras"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "--E9sUjgzcQ4",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "d5fd23a1-10c5-4d27-94ce-771e53ebac47"
      },
      "source": [
        "#loading the dataset\n",
        "\n",
        "housing = fetch_california_housing()\n",
        "\n",
        "#splitting out the dataset\n",
        "\n",
        "#splitting the data into train and test first\n",
        "\n",
        "X_train_full, X_test, y_train_full, y_test = train_test_split(housing.data, housing.target)\n",
        "\n",
        "#now , split the full training data into train and validation set\n",
        "\n",
        "X_train, X_valid, y_train, y_valid = train_test_split(X_train_full,y_train_full)"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading Cal. housing from https://ndownloader.figshare.com/files/5976036 to /root/scikit_learn_data\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "03WlyCx7zd8l",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#data transformation (scaling the data before training it to the model\n",
        "\n",
        "scaler = StandardScaler()\n",
        "X_train = scaler.fit_transform(X_train)\n",
        "X_valid = scaler.transform(X_valid)\n",
        "X_test = scaler.transform(X_test)\n"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "f69C3SWQ3AlE",
        "colab_type": "text"
      },
      "source": [
        "## Using Single Input"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IOhzzYSCzlMl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# building model using FUnctional API\n",
        "\n",
        "input_ = keras.layers.Input(shape=X_train.shape[1:])\n",
        "hidden1 = keras.layers.Dense(30,activation='relu')(input_)\n",
        "hidden2 = keras.layers.Dense(30,activation='relu')(hidden1)\n",
        "concat = keras.layers.Concatenate()([input_,hidden2])\n",
        "output = keras.layers.Dense(1)(concat)\n",
        "\n",
        "model = keras.Model(inputs=[input_],outputs=[output])"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z3QyvEUv0r7R",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 336
        },
        "outputId": "9fca5d62-a7ce-4739-ea2d-7b5e50d28bef"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"functional_1\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            [(None, 8)]          0                                            \n",
            "__________________________________________________________________________________________________\n",
            "dense (Dense)                   (None, 30)           270         input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "dense_1 (Dense)                 (None, 30)           930         dense[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "concatenate (Concatenate)       (None, 38)           0           input_1[0][0]                    \n",
            "                                                                 dense_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "dense_2 (Dense)                 (None, 1)            39          concatenate[0][0]                \n",
            "==================================================================================================\n",
            "Total params: 1,239\n",
            "Trainable params: 1,239\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ucc5nHjl0uJe",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 533
        },
        "outputId": "19e78559-fc39-48c5-ae6c-c6effb42bf4e"
      },
      "source": [
        "# visualizing the model\n",
        "\n",
        "from keras.utils.vis_utils import plot_model\n",
        "plot_model(model, show_shapes=True, show_layer_names=True)"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R2M1YAiw1QCr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 689
        },
        "outputId": "869127ab-608f-4d62-9d36-afa36d855ac4"
      },
      "source": [
        "# compiling and fitting the model\n",
        "\n",
        "model.compile(loss='mean_squared_error', optimizer=keras.optimizers.SGD(learning_rate=0.001),metrics=['accuracy'])\n",
        "history = model.fit(X_train, y_train, epochs=20, validation_data=(X_valid, y_valid))"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "363/363 [==============================] - 1s 2ms/step - loss: 2.9160 - accuracy: 0.0026 - val_loss: 0.9167 - val_accuracy: 0.0021\n",
            "Epoch 2/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.7910 - accuracy: 0.0035 - val_loss: 0.6471 - val_accuracy: 0.0021\n",
            "Epoch 3/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.6595 - accuracy: 0.0034 - val_loss: 0.6033 - val_accuracy: 0.0021\n",
            "Epoch 4/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.6208 - accuracy: 0.0034 - val_loss: 0.5794 - val_accuracy: 0.0021\n",
            "Epoch 5/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5951 - accuracy: 0.0034 - val_loss: 0.5595 - val_accuracy: 0.0021\n",
            "Epoch 6/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5763 - accuracy: 0.0034 - val_loss: 0.5446 - val_accuracy: 0.0021\n",
            "Epoch 7/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5603 - accuracy: 0.0034 - val_loss: 0.5336 - val_accuracy: 0.0021\n",
            "Epoch 8/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5475 - accuracy: 0.0034 - val_loss: 0.5218 - val_accuracy: 0.0021\n",
            "Epoch 9/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5391 - accuracy: 0.0034 - val_loss: 0.5134 - val_accuracy: 0.0021\n",
            "Epoch 10/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5306 - accuracy: 0.0034 - val_loss: 0.5069 - val_accuracy: 0.0021\n",
            "Epoch 11/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5254 - accuracy: 0.0034 - val_loss: 0.5018 - val_accuracy: 0.0021\n",
            "Epoch 12/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5178 - accuracy: 0.0034 - val_loss: 0.4988 - val_accuracy: 0.0021\n",
            "Epoch 13/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5132 - accuracy: 0.0034 - val_loss: 0.4931 - val_accuracy: 0.0021\n",
            "Epoch 14/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5075 - accuracy: 0.0034 - val_loss: 0.4919 - val_accuracy: 0.0021\n",
            "Epoch 15/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5033 - accuracy: 0.0034 - val_loss: 0.4848 - val_accuracy: 0.0021\n",
            "Epoch 16/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5000 - accuracy: 0.0034 - val_loss: 0.4824 - val_accuracy: 0.0021\n",
            "Epoch 17/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.4962 - accuracy: 0.0034 - val_loss: 0.4815 - val_accuracy: 0.0021\n",
            "Epoch 18/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.4940 - accuracy: 0.0033 - val_loss: 0.4763 - val_accuracy: 0.0021\n",
            "Epoch 19/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.4882 - accuracy: 0.0033 - val_loss: 0.4724 - val_accuracy: 0.0021\n",
            "Epoch 20/20\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.4867 - accuracy: 0.0033 - val_loss: 0.4704 - val_accuracy: 0.0021\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7ZOQJ2gZ2icl",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 118
        },
        "outputId": "7f29b108-31de-4d13-89ac-eeb4414e7ac9"
      },
      "source": [
        "print(model.evaluate(X_test, y_test))\n",
        "\n",
        "# making predictions\n",
        "X_new = X_test[:3]\n",
        "predict = model.predict(X_new)\n",
        "\n",
        "print(\"Predictions made by model :\", predict)\n",
        "print(\"Actual values :\",y_test[:3])"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "162/162 [==============================] - 0s 841us/step - loss: 0.4896 - accuracy: 0.0025\n",
            "[0.4896199107170105, 0.002519379835575819]\n",
            "Predictions made by model : [[1.536622  ]\n",
            " [0.75897163]\n",
            " [2.0699995 ]]\n",
            "Actual values : [1.367 0.788 1.625]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5dFzgPTj3E-B",
        "colab_type": "text"
      },
      "source": [
        "## Using multiple Inputs"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NdLtNqQ-2u1y",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 370
        },
        "outputId": "ff86f3fa-104a-42af-dba3-4f5c1f5fb1fc"
      },
      "source": [
        "input_A = keras.layers.Input(shape=[5],name='wide_input')\n",
        "input_B = keras.layers.Input(shape=[6],name='deep_input')\n",
        "hidden1 = keras.layers.Dense(30,activation='relu')(input_B)\n",
        "hidden2 = keras.layers.Dense(30,activation='relu')(hidden1)\n",
        "concat = keras.layers.Concatenate()([input_A,hidden2])\n",
        "output = keras.layers.Dense(1,name=\"output\")(concat)\n",
        "\n",
        "model = keras.Model(inputs=[input_A,input_B],outputs=[output])\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"functional_3\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "deep_input (InputLayer)         [(None, 6)]          0                                            \n",
            "__________________________________________________________________________________________________\n",
            "dense_3 (Dense)                 (None, 30)           210         deep_input[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "wide_input (InputLayer)         [(None, 5)]          0                                            \n",
            "__________________________________________________________________________________________________\n",
            "dense_4 (Dense)                 (None, 30)           930         dense_3[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_1 (Concatenate)     (None, 35)           0           wide_input[0][0]                 \n",
            "                                                                 dense_4[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "output (Dense)                  (None, 1)            36          concatenate_1[0][0]              \n",
            "==================================================================================================\n",
            "Total params: 1,176\n",
            "Trainable params: 1,176\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mihk0GMN4pbT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 533
        },
        "outputId": "e482f0d7-c6d8-4f73-d893-ecf13534b5c8"
      },
      "source": [
        "# visualizing the model\n",
        "plot_model(model, show_shapes=True, show_layer_names=True)"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ndsPGEtg4x0L",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(loss='mse',optimizer=keras.optimizers.SGD(learning_rate=1e-3))"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nYepvi6l5TVQ",
        "colab_type": "text"
      },
      "source": [
        "#### Whenever we are using  multiple inputs , we follow the normal process of creating the model. BUt for fitting the model , instead of passing in a single input matrix,we need to pass a pair of matrices , one per input"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "umX9kTHp5JXr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#splitting the dataset into pair of inputs to be passed into the model having multiple inputs\n",
        "\n",
        "X_train_A , X_train_B = X_train[:,:5],X_train[:,2:]\n",
        "X_valid_A , X_valid_B = X_valid[:,:5],X_valid[:,2:]\n",
        "X_test_A , X_test_B = X_test[:,:5],X_test[:,2:]\n",
        "X_new_A , X_new_B = X_test_A[:5],X_test_B[:5]"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FFqQ_4E06P-n",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 857
        },
        "outputId": "2cb9296a-3b22-4a48-932c-e21da42e8d2d"
      },
      "source": [
        "#fitting the model\n",
        "\n",
        "history = model.fit((X_train_A,X_train_B),y_train,epochs=25,validation_data=((X_valid_A,X_valid_B),y_valid))"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/25\n",
            "363/363 [==============================] - 1s 2ms/step - loss: 1.9558 - val_loss: 0.7437\n",
            "Epoch 2/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.7630 - val_loss: 0.6643\n",
            "Epoch 3/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.6948 - val_loss: 0.6303\n",
            "Epoch 4/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.6561 - val_loss: 0.6067\n",
            "Epoch 5/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.6296 - val_loss: 0.5875\n",
            "Epoch 6/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.6095 - val_loss: 0.5732\n",
            "Epoch 7/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5919 - val_loss: 0.5619\n",
            "Epoch 8/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5788 - val_loss: 0.5522\n",
            "Epoch 9/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5651 - val_loss: 0.5410\n",
            "Epoch 10/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5580 - val_loss: 0.5327\n",
            "Epoch 11/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5467 - val_loss: 0.5253\n",
            "Epoch 12/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5385 - val_loss: 0.5207\n",
            "Epoch 13/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5303 - val_loss: 0.5144\n",
            "Epoch 14/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5251 - val_loss: 0.5116\n",
            "Epoch 15/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5194 - val_loss: 0.5078\n",
            "Epoch 16/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5147 - val_loss: 0.5029\n",
            "Epoch 17/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5102 - val_loss: 0.4992\n",
            "Epoch 18/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5066 - val_loss: 0.4963\n",
            "Epoch 19/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.5028 - val_loss: 0.4937\n",
            "Epoch 20/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.4995 - val_loss: 0.4917\n",
            "Epoch 21/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.4967 - val_loss: 0.4879\n",
            "Epoch 22/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.4939 - val_loss: 0.4854\n",
            "Epoch 23/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.4915 - val_loss: 0.4833\n",
            "Epoch 24/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.4888 - val_loss: 0.4807\n",
            "Epoch 25/25\n",
            "363/363 [==============================] - 0s 1ms/step - loss: 0.4865 - val_loss: 0.4800\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pgRAz82F6jUh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "d340db6a-8db9-4967-ed46-708eb532f72b"
      },
      "source": [
        "mse_test = model.evaluate((X_test_A,X_test_B),y_test)\n"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "162/162 [==============================] - 0s 912us/step - loss: 0.4909\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n6WGKloE6zfc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 101
        },
        "outputId": "e1579261-8311-41f2-b7ed-d5c0ac0d5414"
      },
      "source": [
        "y_pred = model.predict((X_new_A,X_new_B))\n",
        "y_pred"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[1.6151942 ],\n",
              "       [0.74973434],\n",
              "       [1.9560661 ],\n",
              "       [0.9059689 ],\n",
              "       [2.4021225 ]], dtype=float32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "28Govyan7zbq",
        "colab_type": "text"
      },
      "source": [
        "## Handling Multiple Outputs"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Xe_ATz4068nK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 403
        },
        "outputId": "7c2ffd92-cc28-4fd0-e686-cd01c2a4f833"
      },
      "source": [
        "input_A = keras.layers.Input(shape=[5],name='wide_input') #first input\n",
        "input_B = keras.layers.Input(shape=[6],name='deep_input') # second input\n",
        "hidden1 = keras.layers.Dense(30,activation='relu')(input_B) \n",
        "hidden2 = keras.layers.Dense(30,activation='relu')(hidden1)\n",
        "concat = keras.layers.Concatenate()([input_A,hidden2])\n",
        "output = keras.layers.Dense(1,name=\"main_output\")(concat)  #first output\n",
        "aux_output = keras.layers.Dense(1,name=\"aux_output\")(hidden2)  #first output\n",
        "\n",
        "model = keras.Model(inputs=[input_A,input_B],outputs=[output,aux_output])\n",
        "\n",
        "model.summary()"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"functional_5\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "deep_input (InputLayer)         [(None, 6)]          0                                            \n",
            "__________________________________________________________________________________________________\n",
            "dense_5 (Dense)                 (None, 30)           210         deep_input[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "wide_input (InputLayer)         [(None, 5)]          0                                            \n",
            "__________________________________________________________________________________________________\n",
            "dense_6 (Dense)                 (None, 30)           930         dense_5[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_2 (Concatenate)     (None, 35)           0           wide_input[0][0]                 \n",
            "                                                                 dense_6[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "main_output (Dense)             (None, 1)            36          concatenate_2[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "aux_output (Dense)              (None, 1)            31          dense_6[0][0]                    \n",
            "==================================================================================================\n",
            "Total params: 1,207\n",
            "Trainable params: 1,207\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FsUp4gvV7DcW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 533
        },
        "outputId": "fd5f0b50-2089-4834-f6c4-00e88853ad8c"
      },
      "source": [
        "# visualizing the model\n",
        "plot_model(model, show_shapes=True, show_layer_names=True)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NwEiAVow8hYd",
        "colab_type": "text"
      },
      "source": [
        "#### When we try to compile a model , we need to pass a list of all the losses. Also , we can determine which output to give more weights or priority in other words. \n",
        "\n",
        "#### here , we will assign main output more weights than that of the auxiliary output\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ifSMH0YV8YG2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(optimizer='sgd',loss=['mse','mse'],loss_weights=[0.9,0.1])"
      ],
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZHicX41m9bhR",
        "colab_type": "text"
      },
      "source": [
        "#### when we train the model , we need to pass the labels or target variables for each output"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_k0ga4U99ZPR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 709
        },
        "outputId": "b88c01b5-a0bc-4cf5-a7c1-b7634df4d69a"
      },
      "source": [
        "history = model.fit((X_train_A,X_train_B),(y_train,y_train),epochs=20,validation_data=((X_valid_A,X_valid_B),(y_valid,y_valid)))"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "363/363 [==============================] - 1s 2ms/step - loss: 1.1032 - main_output_loss: 0.9876 - aux_output_loss: 2.1436 - val_loss: 0.6003 - val_main_output_loss: 0.5369 - val_aux_output_loss: 1.1705\n",
            "Epoch 2/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.5911 - main_output_loss: 0.5315 - aux_output_loss: 1.1276 - val_loss: 0.5432 - val_main_output_loss: 0.4938 - val_aux_output_loss: 0.9875\n",
            "Epoch 3/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.5554 - main_output_loss: 0.5081 - aux_output_loss: 0.9814 - val_loss: 0.5058 - val_main_output_loss: 0.4656 - val_aux_output_loss: 0.8677\n",
            "Epoch 4/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.5912 - main_output_loss: 0.5534 - aux_output_loss: 0.9314 - val_loss: 0.4956 - val_main_output_loss: 0.4582 - val_aux_output_loss: 0.8323\n",
            "Epoch 5/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.5317 - main_output_loss: 0.5003 - aux_output_loss: 0.8140 - val_loss: 0.4933 - val_main_output_loss: 0.4669 - val_aux_output_loss: 0.7307\n",
            "Epoch 6/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4916 - main_output_loss: 0.4617 - aux_output_loss: 0.7602 - val_loss: 0.4666 - val_main_output_loss: 0.4430 - val_aux_output_loss: 0.6791\n",
            "Epoch 7/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4696 - main_output_loss: 0.4447 - aux_output_loss: 0.6942 - val_loss: 0.4593 - val_main_output_loss: 0.4387 - val_aux_output_loss: 0.6443\n",
            "Epoch 8/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4619 - main_output_loss: 0.4401 - aux_output_loss: 0.6582 - val_loss: 0.4444 - val_main_output_loss: 0.4248 - val_aux_output_loss: 0.6212\n",
            "Epoch 9/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4541 - main_output_loss: 0.4343 - aux_output_loss: 0.6325 - val_loss: 0.6465 - val_main_output_loss: 0.6460 - val_aux_output_loss: 0.6507\n",
            "Epoch 10/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4602 - main_output_loss: 0.4410 - aux_output_loss: 0.6330 - val_loss: 0.4245 - val_main_output_loss: 0.4075 - val_aux_output_loss: 0.5776\n",
            "Epoch 11/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4299 - main_output_loss: 0.4119 - aux_output_loss: 0.5918 - val_loss: 0.4470 - val_main_output_loss: 0.4322 - val_aux_output_loss: 0.5802\n",
            "Epoch 12/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4216 - main_output_loss: 0.4046 - aux_output_loss: 0.5747 - val_loss: 0.4070 - val_main_output_loss: 0.3920 - val_aux_output_loss: 0.5419\n",
            "Epoch 13/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4424 - main_output_loss: 0.4285 - aux_output_loss: 0.5681 - val_loss: 0.3997 - val_main_output_loss: 0.3852 - val_aux_output_loss: 0.5300\n",
            "Epoch 14/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4247 - main_output_loss: 0.4114 - aux_output_loss: 0.5443 - val_loss: 0.3890 - val_main_output_loss: 0.3756 - val_aux_output_loss: 0.5094\n",
            "Epoch 15/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4097 - main_output_loss: 0.3969 - aux_output_loss: 0.5253 - val_loss: 0.3811 - val_main_output_loss: 0.3687 - val_aux_output_loss: 0.4920\n",
            "Epoch 16/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.4275 - main_output_loss: 0.4173 - aux_output_loss: 0.5198 - val_loss: 0.3845 - val_main_output_loss: 0.3720 - val_aux_output_loss: 0.4972\n",
            "Epoch 17/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.3849 - main_output_loss: 0.3706 - aux_output_loss: 0.5135 - val_loss: 0.3681 - val_main_output_loss: 0.3560 - val_aux_output_loss: 0.4774\n",
            "Epoch 18/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.3778 - main_output_loss: 0.3646 - aux_output_loss: 0.4973 - val_loss: 0.3624 - val_main_output_loss: 0.3506 - val_aux_output_loss: 0.4686\n",
            "Epoch 19/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.3766 - main_output_loss: 0.3638 - aux_output_loss: 0.4922 - val_loss: 0.3626 - val_main_output_loss: 0.3514 - val_aux_output_loss: 0.4639\n",
            "Epoch 20/20\n",
            "363/363 [==============================] - 1s 1ms/step - loss: 0.3706 - main_output_loss: 0.3582 - aux_output_loss: 0.4817 - val_loss: 0.3573 - val_main_output_loss: 0.3462 - val_aux_output_loss: 0.4575\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Zbs5DviI99uG",
        "colab_type": "text"
      },
      "source": [
        "#### even for testing the model, we need to provide labels for each output"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5zjtq-Fa9xz1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "e7ee12a8-6d45-4d1d-dac0-1bafa56ad54a"
      },
      "source": [
        "model.evaluate([X_test_A,X_test_B],[y_test,y_test])"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "162/162 [==============================] - 0s 966us/step - loss: 0.3706 - main_output_loss: 0.3578 - aux_output_loss: 0.4860\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.3705834150314331, 0.357757568359375, 0.4860157072544098]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zOmTUA8L-Tyy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 168
        },
        "outputId": "dd20d428-0a37-458a-b865-7642a53b50dc"
      },
      "source": [
        "# for prediction\n",
        "model.predict([X_new_A,X_new_B])"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[array([[1.2670124],\n",
              "        [0.8314538],\n",
              "        [1.6868083],\n",
              "        [1.0453032],\n",
              "        [2.0926318]], dtype=float32), array([[1.3346807],\n",
              "        [1.0147464],\n",
              "        [1.2783837],\n",
              "        [1.0791674],\n",
              "        [1.6341279]], dtype=float32)]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zhswF9DM-eF-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}