TF2_SPAM_DETECTION_CNN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "TF2 SPAM DETECTION CNN.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyPAt71HKrOjysC6vyBdZQL8"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "dos2ctuOB0YH",
        "colab_type": "code",
        "outputId": "04027784-a1e5-4ab6-ff89-bd0ea3d75461",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "!pip install -q tensorflow-gpu==2.0.0"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[31mERROR: Operation cancelled by user\u001b[0m\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AJkljsJeB8pF",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "from tensorflow.keras.layers import Dense, Input, GlobalMaxPooling1D\n",
        "from tensorflow.keras.layers import  Conv1D, MaxPooling1D, Embedding\n",
        "from tensorflow.keras.models import Model\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0ApSSrDaMpku",
        "colab_type": "code",
        "outputId": "6bc931e2-fcd0-484e-fffa-4e92b013ca11",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "!wget https://lazyprogrammer.me/course_files/spam.csv"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-02-20 01:32:03--  https://lazyprogrammer.me/course_files/spam.csv\n",
            "Resolving lazyprogrammer.me (lazyprogrammer.me)... 104.31.80.48, 104.31.81.48, 2606:4700:3035::681f:5130, ...\n",
            "Connecting to lazyprogrammer.me (lazyprogrammer.me)|104.31.80.48|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 503663 (492K) [text/csv]\n",
            "Saving to: ‘spam.csv.1’\n",
            "\n",
            "\rspam.csv.1            0%[                    ]       0  --.-KB/s               \rspam.csv.1          100%[===================>] 491.86K  --.-KB/s    in 0.02s   \n",
            "\n",
            "2020-02-20 01:32:03 (28.6 MB/s) - ‘spam.csv.1’ saved [503663/503663]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FVYQgZLGNVEW",
        "colab_type": "code",
        "outputId": "ed916f3a-77c9-4297-f82b-7d1f4e23c959",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "df = pd.read_csv('spam.csv', encoding = 'ISO-8859-1')\n",
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>v1</th>\n",
              "      <th>v2</th>\n",
              "      <th>Unnamed: 2</th>\n",
              "      <th>Unnamed: 3</th>\n",
              "      <th>Unnamed: 4</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>ham</td>\n",
              "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ham</td>\n",
              "      <td>Ok lar... Joking wif u oni...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>spam</td>\n",
              "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>ham</td>\n",
              "      <td>U dun say so early hor... U c already then say...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>ham</td>\n",
              "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     v1  ... Unnamed: 4\n",
              "0   ham  ...        NaN\n",
              "1   ham  ...        NaN\n",
              "2  spam  ...        NaN\n",
              "3   ham  ...        NaN\n",
              "4   ham  ...        NaN\n",
              "\n",
              "[5 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LOIikFuIObG9",
        "colab_type": "code",
        "outputId": "9ef9a1e7-01c2-468e-fc96-f436be03c01f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "df = df.drop([\"Unnamed: 2\", \"Unnamed: 3\", \"Unnamed: 4\"], axis=1)\n",
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>v1</th>\n",
              "      <th>v2</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>ham</td>\n",
              "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ham</td>\n",
              "      <td>Ok lar... Joking wif u oni...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>spam</td>\n",
              "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>ham</td>\n",
              "      <td>U dun say so early hor... U c already then say...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>ham</td>\n",
              "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     v1                                                 v2\n",
              "0   ham  Go until jurong point, crazy.. Available only ...\n",
              "1   ham                      Ok lar... Joking wif u oni...\n",
              "2  spam  Free entry in 2 a wkly comp to win FA Cup fina...\n",
              "3   ham  U dun say so early hor... U c already then say...\n",
              "4   ham  Nah I don't think he goes to usf, he lives aro..."
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GbBBFNd6Oj_a",
        "colab_type": "code",
        "outputId": "74767edf-f1d0-4acf-b0ef-04e4885cdcd4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "df.columns = ['labels', 'data']\n",
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>labels</th>\n",
              "      <th>data</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>ham</td>\n",
              "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ham</td>\n",
              "      <td>Ok lar... Joking wif u oni...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>spam</td>\n",
              "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>ham</td>\n",
              "      <td>U dun say so early hor... U c already then say...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>ham</td>\n",
              "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  labels                                               data\n",
              "0    ham  Go until jurong point, crazy.. Available only ...\n",
              "1    ham                      Ok lar... Joking wif u oni...\n",
              "2   spam  Free entry in 2 a wkly comp to win FA Cup fina...\n",
              "3    ham  U dun say so early hor... U c already then say...\n",
              "4    ham  Nah I don't think he goes to usf, he lives aro..."
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NS0krnGAOujR",
        "colab_type": "code",
        "outputId": "ee763ff5-f79b-4e6c-ade8-42485b82388d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "df['b_labels'] = df['labels'].map({'ham': 0, 'spam': 1})\n",
        "Y = df['b_labels'].values\n",
        "df.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>labels</th>\n",
              "      <th>data</th>\n",
              "      <th>b_labels</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>ham</td>\n",
              "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ham</td>\n",
              "      <td>Ok lar... Joking wif u oni...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>spam</td>\n",
              "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>ham</td>\n",
              "      <td>U dun say so early hor... U c already then say...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>ham</td>\n",
              "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  labels                                               data  b_labels\n",
              "0    ham  Go until jurong point, crazy.. Available only ...         0\n",
              "1    ham                      Ok lar... Joking wif u oni...         0\n",
              "2   spam  Free entry in 2 a wkly comp to win FA Cup fina...         1\n",
              "3    ham  U dun say so early hor... U c already then say...         0\n",
              "4    ham  Nah I don't think he goes to usf, he lives aro...         0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SCwanWyFOx7o",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df_train, df_test, Y_train, Y_test = train_test_split(df['data'], Y, test_size = 0.33)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oRXYT4CiO55G",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "MAX_VOCAB_SIZE = 20000\n",
        "tokenizer = Tokenizer(num_words=MAX_VOCAB_SIZE)\n",
        "tokenizer.fit_on_texts(df_train)\n",
        "sequences_train = tokenizer.texts_to_sequences(df_train)\n",
        "sequences_test = tokenizer.texts_to_sequences(df_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l0J1OGoAO93e",
        "colab_type": "code",
        "outputId": "c5035d79-1c3a-4e02-ac49-5b747277bfca",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "word2idx = tokenizer.word_index\n",
        "V = len(word2idx)\n",
        "print('Se han encontrado %s tokens unicos' % V)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Se han encontrado 7222 tokens unicos\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g4HbcbFwPUjj",
        "colab_type": "code",
        "outputId": "7bd0d70d-5b4e-4bfd-c9cb-9c9582d240c2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "data_train = pad_sequences(sequences_train)\n",
        "print('Tamaño de el tensor de datos de entrenamiento:', data_train.shape)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Tamaño de el tensor de datos de entrenamiento: (3733, 189)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c3c_SXBoPVs7",
        "colab_type": "code",
        "outputId": "e8cfc680-cd8f-4cdd-d182-dc9841a02c1c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "T = data_train.shape[1]\n",
        "print(T)\n",
        "data_test = pad_sequences(sequences_test, maxlen = T)\n",
        "print('Tamaño de el tensor de datos de prueba:', data_test.shape)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "189\n",
            "Tamaño de el tensor de datos de prueba: (1839, 189)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oB-MCqDyPcc5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "D = 20\n",
        "\n",
        "i = Input(shape=(T,))\n",
        "x = Embedding(V + 1, D)(i)\n",
        "x = Conv1D(32, 3, activation='relu')(x)\n",
        "x = MaxPooling1D(3)(x)\n",
        "x = Conv1D(64, 3, activation='relu')(x)\n",
        "x = MaxPooling1D(3)(x)\n",
        "x = Conv1D(128, 3, activation='relu')(x)\n",
        "x = GlobalMaxPooling1D()(x)\n",
        "x = Dense(1, activation='sigmoid')(x)\n",
        "\n",
        "model = Model(i, x)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jfoaH7AdQgPE",
        "colab_type": "code",
        "outputId": "5a7cd795-b3d1-41e4-c407-5317332e99d6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "model.compile(\n",
        "    loss= 'binary_crossentropy',\n",
        "    optimizer = 'adam',\n",
        "    metrics = ['accuracy']\n",
        ")\n",
        "\n",
        "print('A entrenar!!...')\n",
        "\n",
        "r = model.fit(\n",
        "    data_train,\n",
        "    Y_train,\n",
        "    epochs = 30,\n",
        "    validation_data = (data_test, Y_test)\n",
        ")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "A entrenar!!...\n",
            "Train on 3733 samples, validate on 1839 samples\n",
            "Epoch 1/30\n",
            "3733/3733 [==============================] - 9s 2ms/sample - loss: 0.3825 - accuracy: 0.8634 - val_loss: 0.2647 - val_accuracy: 0.8684\n",
            "Epoch 2/30\n",
            "3733/3733 [==============================] - 1s 264us/sample - loss: 0.1086 - accuracy: 0.9628 - val_loss: 0.1578 - val_accuracy: 0.9652\n",
            "Epoch 3/30\n",
            "3733/3733 [==============================] - 1s 271us/sample - loss: 0.0212 - accuracy: 0.9938 - val_loss: 0.0719 - val_accuracy: 0.9826\n",
            "Epoch 4/30\n",
            "3733/3733 [==============================] - 1s 277us/sample - loss: 0.0042 - accuracy: 0.9997 - val_loss: 0.0924 - val_accuracy: 0.9815\n",
            "Epoch 5/30\n",
            "3733/3733 [==============================] - 1s 288us/sample - loss: 0.0021 - accuracy: 0.9997 - val_loss: 0.0829 - val_accuracy: 0.9831\n",
            "Epoch 6/30\n",
            "3733/3733 [==============================] - 1s 295us/sample - loss: 0.0015 - accuracy: 0.9997 - val_loss: 0.0922 - val_accuracy: 0.9826\n",
            "Epoch 7/30\n",
            "3733/3733 [==============================] - 1s 275us/sample - loss: 0.0014 - accuracy: 0.9997 - val_loss: 0.0953 - val_accuracy: 0.9831\n",
            "Epoch 8/30\n",
            "3733/3733 [==============================] - 1s 272us/sample - loss: 0.0011 - accuracy: 0.9997 - val_loss: 0.1017 - val_accuracy: 0.9821\n",
            "Epoch 9/30\n",
            "3733/3733 [==============================] - 1s 296us/sample - loss: 9.7153e-04 - accuracy: 0.9997 - val_loss: 0.1055 - val_accuracy: 0.9821\n",
            "Epoch 10/30\n",
            "3733/3733 [==============================] - 1s 275us/sample - loss: 7.1397e-04 - accuracy: 0.9997 - val_loss: 0.0955 - val_accuracy: 0.9853\n",
            "Epoch 11/30\n",
            "3733/3733 [==============================] - 1s 289us/sample - loss: 4.4624e-04 - accuracy: 1.0000 - val_loss: 0.1078 - val_accuracy: 0.9842\n",
            "Epoch 12/30\n",
            "3733/3733 [==============================] - 1s 296us/sample - loss: 3.2232e-04 - accuracy: 0.9997 - val_loss: 0.0937 - val_accuracy: 0.9864\n",
            "Epoch 13/30\n",
            "3733/3733 [==============================] - 1s 286us/sample - loss: 1.6779e-04 - accuracy: 1.0000 - val_loss: 0.1095 - val_accuracy: 0.9848\n",
            "Epoch 14/30\n",
            "3733/3733 [==============================] - 1s 323us/sample - loss: 1.0738e-04 - accuracy: 1.0000 - val_loss: 0.1124 - val_accuracy: 0.9848\n",
            "Epoch 15/30\n",
            "3733/3733 [==============================] - 1s 288us/sample - loss: 7.8394e-05 - accuracy: 1.0000 - val_loss: 0.1151 - val_accuracy: 0.9848\n",
            "Epoch 16/30\n",
            "3733/3733 [==============================] - 1s 297us/sample - loss: 6.5629e-05 - accuracy: 1.0000 - val_loss: 0.1139 - val_accuracy: 0.9848\n",
            "Epoch 17/30\n",
            "3733/3733 [==============================] - 1s 278us/sample - loss: 4.7245e-05 - accuracy: 1.0000 - val_loss: 0.1147 - val_accuracy: 0.9848\n",
            "Epoch 18/30\n",
            "3733/3733 [==============================] - 1s 282us/sample - loss: 3.7949e-05 - accuracy: 1.0000 - val_loss: 0.1179 - val_accuracy: 0.9848\n",
            "Epoch 19/30\n",
            "3733/3733 [==============================] - 1s 287us/sample - loss: 3.0844e-05 - accuracy: 1.0000 - val_loss: 0.1201 - val_accuracy: 0.9848\n",
            "Epoch 20/30\n",
            "3733/3733 [==============================] - 1s 292us/sample - loss: 2.5834e-05 - accuracy: 1.0000 - val_loss: 0.1210 - val_accuracy: 0.9848\n",
            "Epoch 21/30\n",
            "3733/3733 [==============================] - 1s 275us/sample - loss: 2.1967e-05 - accuracy: 1.0000 - val_loss: 0.1221 - val_accuracy: 0.9848\n",
            "Epoch 22/30\n",
            "3733/3733 [==============================] - 1s 278us/sample - loss: 1.8679e-05 - accuracy: 1.0000 - val_loss: 0.1238 - val_accuracy: 0.9848\n",
            "Epoch 23/30\n",
            "3733/3733 [==============================] - 1s 284us/sample - loss: 1.5906e-05 - accuracy: 1.0000 - val_loss: 0.1242 - val_accuracy: 0.9848\n",
            "Epoch 24/30\n",
            "3733/3733 [==============================] - 1s 273us/sample - loss: 1.3719e-05 - accuracy: 1.0000 - val_loss: 0.1265 - val_accuracy: 0.9848\n",
            "Epoch 25/30\n",
            "3733/3733 [==============================] - 1s 334us/sample - loss: 1.2091e-05 - accuracy: 1.0000 - val_loss: 0.1271 - val_accuracy: 0.9848\n",
            "Epoch 26/30\n",
            "3733/3733 [==============================] - 1s 287us/sample - loss: 1.0708e-05 - accuracy: 1.0000 - val_loss: 0.1266 - val_accuracy: 0.9848\n",
            "Epoch 27/30\n",
            "3733/3733 [==============================] - 1s 285us/sample - loss: 9.4162e-06 - accuracy: 1.0000 - val_loss: 0.1279 - val_accuracy: 0.9848\n",
            "Epoch 28/30\n",
            "3733/3733 [==============================] - 1s 272us/sample - loss: 8.2830e-06 - accuracy: 1.0000 - val_loss: 0.1304 - val_accuracy: 0.9848\n",
            "Epoch 29/30\n",
            "3733/3733 [==============================] - 1s 276us/sample - loss: 7.3699e-06 - accuracy: 1.0000 - val_loss: 0.1316 - val_accuracy: 0.9848\n",
            "Epoch 30/30\n",
            "3733/3733 [==============================] - 1s 284us/sample - loss: 6.5844e-06 - accuracy: 1.0000 - val_loss: 0.1324 - val_accuracy: 0.9848\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1TwNgrAoQxu9",
        "colab_type": "code",
        "outputId": "c851d924-ebca-465e-dfe4-acc48f479e5a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        }
      },
      "source": [
        "plt.plot(r.history['loss'], label='loss')\n",
        "plt.plot(r.history['val_loss'], label='val_loss')\n",
        "plt.legend()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f983bff6048>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAD8CAYAAACW/ATfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxqSA/w+F+dJ9T725xwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wqiPYRT_Q0nl",
        "colab_type": "code",
        "outputId": "31634b8b-b989-43ae-b0c6-74810d5dbed1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        }
      },
      "source": [
        "plt.plot(r.history['accuracy'], label='accuracy')\n",
        "plt.plot(r.history['val_accuracy'], label='val_accuracy')\n",
        "plt.legend()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f983bac6b70>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAD4CAYAAAANbUbJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "G9npKOASQ9cs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}