[Captcha_Break]_Model_Captcha.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "[Captcha_Break]-Model_Captcha.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "viMQjsIQ6pZ-",
        "colab_type": "text"
      },
      "source": [
        "## Modelo de Múltipla Saída\n",
        "\n",
        "Classificação dos 5 carácteres do Captcha\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qr4z4IIPs9dm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import cv2\n",
        "from six.moves import cPickle as pickle\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import LabelBinarizer\n",
        "\n",
        "from tensorflow import keras\n",
        "from keras.models import Sequential, Model                          #the model\n",
        "from keras.layers import Dense, Dropout, Activation, Flatten, Input #fully connected layers\n",
        "from keras.layers import Conv2D, MaxPooling2D                       #the convnet\n",
        "from keras.utils import np_utils                                    #extra tools\n",
        "from keras.layers.normalization import BatchNormalization\n",
        "from keras.utils.layer_utils import count_params\n",
        "from keras.callbacks import EarlyStopping"
      ],
      "execution_count": 81,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TTX-fsFDeBbQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "6cc6e967-fdea-44ac-823c-26227b3e150d"
      },
      "source": [
        "keras.__version__"
      ],
      "execution_count": 64,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'2.4.0'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 64
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gGUMtW4GFSry",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "8b8f5514-7a47-4ad9-db6d-e33c122c7355"
      },
      "source": [
        "# Carregando Imagens e Labels\n",
        "data_dir = '/content/drive/My Drive/Colab Notebooks/Data_Colab/Captcha_Break/'\n",
        "\n",
        "with open(data_dir+'CAPTCHAS_ALL'+'_Data' +'.pickle', 'rb') as f:\n",
        "    data = pickle.load(f)\n",
        "\n",
        "with open(data_dir+'CAPTCHAS_ALL'+'_Labels'+'.pickle', 'rb') as f:\n",
        "    labels = pickle.load(f)\n",
        "\n",
        "data.shape, labels.shape"
      ],
      "execution_count": 65,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((9725, 60, 200), (9725,))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "P5eRU_qQtT_7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 118
        },
        "outputId": "f5da420e-a0a1-493a-877e-c9692f090eb2"
      },
      "source": [
        "letras_y = np.array([list(x) for x in labels])\n",
        "np.unique(letras_y,return_counts=True)"
      ],
      "execution_count": 66,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(array(['0', '4', '6', '7', 'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I',\n",
              "        'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V',\n",
              "        'W', 'X', 'Y', 'Z'], dtype='<U1'),\n",
              " array([   1,    3,    1,    4, 1970, 1948, 1899, 1913, 1951, 1834, 1950,\n",
              "        1894, 1301, 1754, 1900, 1811, 1918, 1874, 1780, 1817, 1952, 1919,\n",
              "        1938, 1834, 1929, 2010, 1868, 1992, 1785, 1875]))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 66
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "7YhWFl-ZtUAG",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 118
        },
        "outputId": "59a1faee-b4c0-45cb-fa0d-1ce7ca37a0c2"
      },
      "source": [
        "filtro = np.char.isnumeric(letras_y).sum(1)==0  # Filtro de Caracteres numericos\n",
        "data = data[filtro]\n",
        "labels = labels[filtro]\n",
        "\n",
        "letras_y = np.array([list(x) for x in labels])\n",
        "np.unique(letras_y,return_counts=True)"
      ],
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(array(['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M',\n",
              "        'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z'],\n",
              "       dtype='<U1'),\n",
              " array([1969, 1946, 1898, 1913, 1949, 1834, 1948, 1893, 1300, 1751, 1898,\n",
              "        1808, 1916, 1872, 1779, 1816, 1951, 1919, 1936, 1831, 1928, 2008,\n",
              "        1867, 1992, 1784, 1874]))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 67
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "0tv77s7YtUAc",
        "colab": {}
      },
      "source": [
        "lb = LabelBinarizer().fit(letras_y[:,1])\n",
        "lab = []\n",
        "for ltr in letras_y:\n",
        "    t = lb.transform(ltr)\n",
        "    lab.append(t)"
      ],
      "execution_count": 68,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Wr9WX6m_tUAn",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "cac5fe3b-46c3-4e25-ba51-b3e93ffc072a"
      },
      "source": [
        "labels_lb = np.stack(lab, axis=1)\n",
        "labels_lb.shape"
      ],
      "execution_count": 69,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(5, 9716, 26)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "DmKb7eaMtUBE",
        "colab": {}
      },
      "source": [
        "(train_x, val_x, train_y, val_y) = train_test_split(data, lab, test_size=0.3, random_state=123)"
      ],
      "execution_count": 70,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "SFfuwuoQtUBJ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "26f71835-e1a3-4f82-a57b-a5917bacb57f"
      },
      "source": [
        "train_y = np.stack(train_y, axis=1)\n",
        "val_y = np.stack(val_y, axis=1)\n",
        "train_y.shape, val_y.shape"
      ],
      "execution_count": 71,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((5, 6801, 26), (5, 2915, 26))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 71
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "c7YJQPDIs5KA",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "79da4e90-ffba-438b-b93a-d90252b5418c"
      },
      "source": [
        "X_train = np.expand_dims(train_x, axis = 3)\n",
        "X_valid = np.expand_dims(val_x, axis = 3)\n",
        "X_train.shape, X_valid.shape"
      ],
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((6801, 60, 200, 1), (2915, 60, 200, 1))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 72
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "ARUJ4hJ8s5KH",
        "colab": {}
      },
      "source": [
        "v_y = []\n",
        "for x in val_y:\n",
        "    v_y.append(x)\n",
        "\n",
        "t_y = []\n",
        "for x in train_y:\n",
        "    t_y.append(x)"
      ],
      "execution_count": 73,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VN4sYEfYsi6j",
        "colab_type": "text"
      },
      "source": [
        "---\n",
        "## Modelo de Classificação Múltipla"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LlGaBc5psO-z",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "img_rows = 60\n",
        "img_cols = 200\n",
        "img_channels = 1\n",
        "\n",
        "x = Input((img_rows, img_cols, img_channels))\n",
        "\n",
        "y = Conv2D(16, (5, 5))(x)\n",
        "y = Activation('relu')(y)\n",
        "y = MaxPooling2D((2,2))(y)\n",
        "y = BatchNormalization()(y)\n",
        "\n",
        "y = Conv2D(32, (5, 5))(y)\n",
        "y = Activation('relu')(y)\n",
        "y = MaxPooling2D((2,2))(y)\n",
        "y = BatchNormalization()(y)\n",
        "\n",
        "y = Conv2D(64, (3, 3))(y)\n",
        "y = Activation('relu')(y)\n",
        "y = MaxPooling2D((2,2))(y)\n",
        "y = BatchNormalization()(y)\n",
        "\n",
        "y = Flatten()(y)\n",
        "y = Dense(512, activation=\"relu\")(y)\n",
        "y = Dropout(0.6)(y)\n",
        "y = Dense(512, activation=\"relu\")(y)\n",
        "y = Dropout(0.3)(y)\n",
        "\n",
        "digit_1 = Dense(26, activation='softmax', name = 'c1')(y)\n",
        "digit_2 = Dense(26, activation='softmax', name = 'c2')(y)\n",
        "digit_3 = Dense(26, activation='softmax', name = 'c3')(y)\n",
        "digit_4 = Dense(26, activation='softmax', name = 'c4')(y)\n",
        "digit_5 = Dense(26, activation='softmax', name = 'c5')(y)\n",
        "branches = [digit_1,digit_2,digit_3,digit_4,digit_5]\n",
        "model = Model(inputs=x, outputs=branches,  name = 'Modelo_Captha')"
      ],
      "execution_count": 74,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g07Pm17xsO8d",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 928
        },
        "outputId": "42e669e0-dbe5-4746-b991-daf057380090"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"Modelo_Captha\"\n",
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_4 (InputLayer)            [(None, 60, 200, 1)] 0                                            \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_9 (Conv2D)               (None, 56, 196, 16)  416         input_4[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "activation_9 (Activation)       (None, 56, 196, 16)  0           conv2d_9[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_9 (MaxPooling2D)  (None, 28, 98, 16)   0           activation_9[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_9 (BatchNor (None, 28, 98, 16)   64          max_pooling2d_9[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_10 (Conv2D)              (None, 24, 94, 32)   12832       batch_normalization_9[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "activation_10 (Activation)      (None, 24, 94, 32)   0           conv2d_10[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_10 (MaxPooling2D) (None, 12, 47, 32)   0           activation_10[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_10 (BatchNo (None, 12, 47, 32)   128         max_pooling2d_10[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_11 (Conv2D)              (None, 10, 45, 64)   18496       batch_normalization_10[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "activation_11 (Activation)      (None, 10, 45, 64)   0           conv2d_11[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling2d_11 (MaxPooling2D) (None, 5, 22, 64)    0           activation_11[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_11 (BatchNo (None, 5, 22, 64)    256         max_pooling2d_11[0][0]           \n",
            "__________________________________________________________________________________________________\n",
            "flatten_3 (Flatten)             (None, 7040)         0           batch_normalization_11[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "dense_6 (Dense)                 (None, 512)          3604992     flatten_3[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "dropout_6 (Dropout)             (None, 512)          0           dense_6[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "dense_7 (Dense)                 (None, 512)          262656      dropout_6[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "dropout_7 (Dropout)             (None, 512)          0           dense_7[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "c1 (Dense)                      (None, 26)           13338       dropout_7[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "c2 (Dense)                      (None, 26)           13338       dropout_7[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "c3 (Dense)                      (None, 26)           13338       dropout_7[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "c4 (Dense)                      (None, 26)           13338       dropout_7[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "c5 (Dense)                      (None, 26)           13338       dropout_7[0][0]                  \n",
            "==================================================================================================\n",
            "Total params: 3,966,530\n",
            "Trainable params: 3,966,306\n",
            "Non-trainable params: 224\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pR7bgXnZsO2i",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "c893157a-a25d-4c75-c366-01f736f3e360"
      },
      "source": [
        "keras.utils.plot_model(model, \"shape_info.png\", show_shapes=True, show_layer_names=False)"
      ],
      "execution_count": 76,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 76
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "W6I4szhOs5J6",
        "colab": {}
      },
      "source": [
        "model.compile(loss=\"categorical_crossentropy\", optimizer=\"adam\", metrics=[\"accuracy\"])"
      ],
      "execution_count": 77,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "3LXaBePds5KY",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "9695dfd8-3ef6-4267-ef75-5ec36b3b5394"
      },
      "source": [
        "estop = EarlyStopping(patience=50, mode='min', min_delta=0.001, monitor='val_loss')\n",
        "\n",
        "history = model.fit(X_train, t_y,\n",
        "                    validation_data=(X_valid, v_y), \n",
        "                    batch_size=100, epochs=500, verbose=1, \n",
        "                    callbacks=[estop])"
      ],
      "execution_count": 78,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/500\n",
            "69/69 [==============================] - 3s 38ms/step - loss: 16.6020 - c1_loss: 3.3207 - c2_loss: 3.3238 - c3_loss: 3.3186 - c4_loss: 3.3254 - c5_loss: 3.3135 - c1_accuracy: 0.0390 - c2_accuracy: 0.0432 - c3_accuracy: 0.0396 - c4_accuracy: 0.0401 - c5_accuracy: 0.0469 - val_loss: 16.2838 - val_c1_loss: 3.2566 - val_c2_loss: 3.2562 - val_c3_loss: 3.2561 - val_c4_loss: 3.2572 - val_c5_loss: 3.2576 - val_c1_accuracy: 0.0456 - val_c2_accuracy: 0.0401 - val_c3_accuracy: 0.0408 - val_c4_accuracy: 0.0377 - val_c5_accuracy: 0.0436\n",
            "Epoch 2/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 16.2800 - c1_loss: 3.2541 - c2_loss: 3.2554 - c3_loss: 3.2567 - c4_loss: 3.2579 - c5_loss: 3.2560 - c1_accuracy: 0.0463 - c2_accuracy: 0.0438 - c3_accuracy: 0.0437 - c4_accuracy: 0.0423 - c5_accuracy: 0.0447 - val_loss: 16.2795 - val_c1_loss: 3.2527 - val_c2_loss: 3.2548 - val_c3_loss: 3.2561 - val_c4_loss: 3.2576 - val_c5_loss: 3.2583 - val_c1_accuracy: 0.0532 - val_c2_accuracy: 0.0453 - val_c3_accuracy: 0.0429 - val_c4_accuracy: 0.0384 - val_c5_accuracy: 0.0453\n",
            "Epoch 3/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 16.2624 - c1_loss: 3.2497 - c2_loss: 3.2512 - c3_loss: 3.2547 - c4_loss: 3.2525 - c5_loss: 3.2543 - c1_accuracy: 0.0469 - c2_accuracy: 0.0429 - c3_accuracy: 0.0446 - c4_accuracy: 0.0479 - c5_accuracy: 0.0469 - val_loss: 16.2763 - val_c1_loss: 3.2526 - val_c2_loss: 3.2529 - val_c3_loss: 3.2549 - val_c4_loss: 3.2549 - val_c5_loss: 3.2609 - val_c1_accuracy: 0.0542 - val_c2_accuracy: 0.0456 - val_c3_accuracy: 0.0432 - val_c4_accuracy: 0.0425 - val_c5_accuracy: 0.0343\n",
            "Epoch 4/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 16.2482 - c1_loss: 3.2473 - c2_loss: 3.2462 - c3_loss: 3.2520 - c4_loss: 3.2467 - c5_loss: 3.2560 - c1_accuracy: 0.0437 - c2_accuracy: 0.0485 - c3_accuracy: 0.0451 - c4_accuracy: 0.0513 - c5_accuracy: 0.0431 - val_loss: 16.2758 - val_c1_loss: 3.2550 - val_c2_loss: 3.2530 - val_c3_loss: 3.2526 - val_c4_loss: 3.2589 - val_c5_loss: 3.2563 - val_c1_accuracy: 0.0408 - val_c2_accuracy: 0.0432 - val_c3_accuracy: 0.0401 - val_c4_accuracy: 0.0422 - val_c5_accuracy: 0.0429\n",
            "Epoch 5/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 16.2585 - c1_loss: 3.2392 - c2_loss: 3.2444 - c3_loss: 3.2616 - c4_loss: 3.2588 - c5_loss: 3.2545 - c1_accuracy: 0.0571 - c2_accuracy: 0.0510 - c3_accuracy: 0.0419 - c4_accuracy: 0.0493 - c5_accuracy: 0.0496 - val_loss: 16.1777 - val_c1_loss: 3.2032 - val_c2_loss: 3.2317 - val_c3_loss: 3.2463 - val_c4_loss: 3.2514 - val_c5_loss: 3.2450 - val_c1_accuracy: 0.0655 - val_c2_accuracy: 0.0628 - val_c3_accuracy: 0.0501 - val_c4_accuracy: 0.0525 - val_c5_accuracy: 0.0549\n",
            "Epoch 6/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 16.1650 - c1_loss: 3.2070 - c2_loss: 3.2246 - c3_loss: 3.2484 - c4_loss: 3.2406 - c5_loss: 3.2444 - c1_accuracy: 0.0584 - c2_accuracy: 0.0544 - c3_accuracy: 0.0446 - c4_accuracy: 0.0544 - c5_accuracy: 0.0494 - val_loss: 16.1662 - val_c1_loss: 3.2082 - val_c2_loss: 3.2281 - val_c3_loss: 3.2507 - val_c4_loss: 3.2360 - val_c5_loss: 3.2431 - val_c1_accuracy: 0.0734 - val_c2_accuracy: 0.0532 - val_c3_accuracy: 0.0473 - val_c4_accuracy: 0.0556 - val_c5_accuracy: 0.0494\n",
            "Epoch 7/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 16.1163 - c1_loss: 3.1866 - c2_loss: 3.2034 - c3_loss: 3.2424 - c4_loss: 3.2425 - c5_loss: 3.2413 - c1_accuracy: 0.0575 - c2_accuracy: 0.0641 - c3_accuracy: 0.0544 - c4_accuracy: 0.0556 - c5_accuracy: 0.0566 - val_loss: 16.0518 - val_c1_loss: 3.1518 - val_c2_loss: 3.1901 - val_c3_loss: 3.2265 - val_c4_loss: 3.2420 - val_c5_loss: 3.2414 - val_c1_accuracy: 0.0662 - val_c2_accuracy: 0.0758 - val_c3_accuracy: 0.0645 - val_c4_accuracy: 0.0583 - val_c5_accuracy: 0.0539\n",
            "Epoch 8/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 16.0846 - c1_loss: 3.1606 - c2_loss: 3.1978 - c3_loss: 3.2484 - c4_loss: 3.2340 - c5_loss: 3.2438 - c1_accuracy: 0.0629 - c2_accuracy: 0.0678 - c3_accuracy: 0.0522 - c4_accuracy: 0.0585 - c5_accuracy: 0.0494 - val_loss: 16.0223 - val_c1_loss: 3.1145 - val_c2_loss: 3.1883 - val_c3_loss: 3.2465 - val_c4_loss: 3.2377 - val_c5_loss: 3.2352 - val_c1_accuracy: 0.0871 - val_c2_accuracy: 0.0635 - val_c3_accuracy: 0.0508 - val_c4_accuracy: 0.0587 - val_c5_accuracy: 0.0549\n",
            "Epoch 9/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 16.0518 - c1_loss: 3.1506 - c2_loss: 3.1964 - c3_loss: 3.2390 - c4_loss: 3.2269 - c5_loss: 3.2388 - c1_accuracy: 0.0604 - c2_accuracy: 0.0616 - c3_accuracy: 0.0494 - c4_accuracy: 0.0568 - c5_accuracy: 0.0573 - val_loss: 15.9507 - val_c1_loss: 3.0752 - val_c2_loss: 3.1794 - val_c3_loss: 3.2398 - val_c4_loss: 3.2308 - val_c5_loss: 3.2255 - val_c1_accuracy: 0.0789 - val_c2_accuracy: 0.0765 - val_c3_accuracy: 0.0563 - val_c4_accuracy: 0.0607 - val_c5_accuracy: 0.0645\n",
            "Epoch 10/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 15.9693 - c1_loss: 3.1371 - c2_loss: 3.1650 - c3_loss: 3.2281 - c4_loss: 3.2127 - c5_loss: 3.2263 - c1_accuracy: 0.0632 - c2_accuracy: 0.0743 - c3_accuracy: 0.0557 - c4_accuracy: 0.0638 - c5_accuracy: 0.0626 - val_loss: 15.8029 - val_c1_loss: 3.0399 - val_c2_loss: 3.1026 - val_c3_loss: 3.2263 - val_c4_loss: 3.2140 - val_c5_loss: 3.2200 - val_c1_accuracy: 0.1173 - val_c2_accuracy: 0.1026 - val_c3_accuracy: 0.0545 - val_c4_accuracy: 0.0727 - val_c5_accuracy: 0.0666\n",
            "Epoch 11/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 15.8681 - c1_loss: 3.0920 - c2_loss: 3.1433 - c3_loss: 3.2213 - c4_loss: 3.1952 - c5_loss: 3.2163 - c1_accuracy: 0.0773 - c2_accuracy: 0.0818 - c3_accuracy: 0.0616 - c4_accuracy: 0.0682 - c5_accuracy: 0.0645 - val_loss: 15.6950 - val_c1_loss: 3.0152 - val_c2_loss: 3.1079 - val_c3_loss: 3.2000 - val_c4_loss: 3.1873 - val_c5_loss: 3.1846 - val_c1_accuracy: 0.1249 - val_c2_accuracy: 0.1211 - val_c3_accuracy: 0.0731 - val_c4_accuracy: 0.0813 - val_c5_accuracy: 0.0806\n",
            "Epoch 12/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 15.7462 - c1_loss: 3.0665 - c2_loss: 3.1156 - c3_loss: 3.2166 - c4_loss: 3.1674 - c5_loss: 3.1802 - c1_accuracy: 0.0857 - c2_accuracy: 0.0851 - c3_accuracy: 0.0648 - c4_accuracy: 0.0731 - c5_accuracy: 0.0720 - val_loss: 15.4474 - val_c1_loss: 2.9658 - val_c2_loss: 3.0035 - val_c3_loss: 3.2050 - val_c4_loss: 3.1214 - val_c5_loss: 3.1518 - val_c1_accuracy: 0.1211 - val_c2_accuracy: 0.1269 - val_c3_accuracy: 0.0573 - val_c4_accuracy: 0.0895 - val_c5_accuracy: 0.1022\n",
            "Epoch 13/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 15.5912 - c1_loss: 3.0401 - c2_loss: 3.0707 - c3_loss: 3.2033 - c4_loss: 3.1431 - c5_loss: 3.1340 - c1_accuracy: 0.0872 - c2_accuracy: 0.0947 - c3_accuracy: 0.0681 - c4_accuracy: 0.0782 - c5_accuracy: 0.0870 - val_loss: 15.3242 - val_c1_loss: 2.9544 - val_c2_loss: 2.9964 - val_c3_loss: 3.1648 - val_c4_loss: 3.1114 - val_c5_loss: 3.0972 - val_c1_accuracy: 0.1509 - val_c2_accuracy: 0.1455 - val_c3_accuracy: 0.0823 - val_c4_accuracy: 0.1043 - val_c5_accuracy: 0.1132\n",
            "Epoch 14/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 15.4191 - c1_loss: 2.9820 - c2_loss: 3.0360 - c3_loss: 3.1894 - c4_loss: 3.0948 - c5_loss: 3.1170 - c1_accuracy: 0.1091 - c2_accuracy: 0.1067 - c3_accuracy: 0.0738 - c4_accuracy: 0.0926 - c5_accuracy: 0.0951 - val_loss: 15.1205 - val_c1_loss: 2.8979 - val_c2_loss: 2.9796 - val_c3_loss: 3.1491 - val_c4_loss: 3.0222 - val_c5_loss: 3.0716 - val_c1_accuracy: 0.1736 - val_c2_accuracy: 0.1575 - val_c3_accuracy: 0.0851 - val_c4_accuracy: 0.1290 - val_c5_accuracy: 0.1341\n",
            "Epoch 15/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 15.2571 - c1_loss: 2.9331 - c2_loss: 3.0037 - c3_loss: 3.1757 - c4_loss: 3.0553 - c5_loss: 3.0892 - c1_accuracy: 0.1222 - c2_accuracy: 0.1151 - c3_accuracy: 0.0765 - c4_accuracy: 0.1032 - c5_accuracy: 0.1028 - val_loss: 14.7967 - val_c1_loss: 2.7744 - val_c2_loss: 2.9159 - val_c3_loss: 3.1163 - val_c4_loss: 2.9758 - val_c5_loss: 3.0143 - val_c1_accuracy: 0.1931 - val_c2_accuracy: 0.1750 - val_c3_accuracy: 0.1050 - val_c4_accuracy: 0.1417 - val_c5_accuracy: 0.1513\n",
            "Epoch 16/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 14.9635 - c1_loss: 2.8732 - c2_loss: 2.9421 - c3_loss: 3.1202 - c4_loss: 3.0175 - c5_loss: 3.0105 - c1_accuracy: 0.1376 - c2_accuracy: 0.1295 - c3_accuracy: 0.0943 - c4_accuracy: 0.1085 - c5_accuracy: 0.1259 - val_loss: 14.3622 - val_c1_loss: 2.7050 - val_c2_loss: 2.7571 - val_c3_loss: 3.0711 - val_c4_loss: 2.9182 - val_c5_loss: 2.9108 - val_c1_accuracy: 0.2137 - val_c2_accuracy: 0.2093 - val_c3_accuracy: 0.1156 - val_c4_accuracy: 0.1437 - val_c5_accuracy: 0.1866\n",
            "Epoch 17/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 14.6786 - c1_loss: 2.8183 - c2_loss: 2.8812 - c3_loss: 3.0578 - c4_loss: 2.9770 - c5_loss: 2.9442 - c1_accuracy: 0.1457 - c2_accuracy: 0.1366 - c3_accuracy: 0.1051 - c4_accuracy: 0.1237 - c5_accuracy: 0.1332 - val_loss: 13.8023 - val_c1_loss: 2.5671 - val_c2_loss: 2.7000 - val_c3_loss: 2.9470 - val_c4_loss: 2.7934 - val_c5_loss: 2.7949 - val_c1_accuracy: 0.2446 - val_c2_accuracy: 0.2230 - val_c3_accuracy: 0.1557 - val_c4_accuracy: 0.1798 - val_c5_accuracy: 0.2134\n",
            "Epoch 18/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 14.0346 - c1_loss: 2.6507 - c2_loss: 2.7834 - c3_loss: 2.9235 - c4_loss: 2.8426 - c5_loss: 2.8343 - c1_accuracy: 0.1809 - c2_accuracy: 0.1598 - c3_accuracy: 0.1350 - c4_accuracy: 0.1407 - c5_accuracy: 0.1553 - val_loss: 13.0667 - val_c1_loss: 2.4190 - val_c2_loss: 2.5501 - val_c3_loss: 2.7849 - val_c4_loss: 2.6669 - val_c5_loss: 2.6458 - val_c1_accuracy: 0.2957 - val_c2_accuracy: 0.2587 - val_c3_accuracy: 0.2003 - val_c4_accuracy: 0.2377 - val_c5_accuracy: 0.2559\n",
            "Epoch 19/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 13.5507 - c1_loss: 2.5307 - c2_loss: 2.6843 - c3_loss: 2.8446 - c4_loss: 2.7354 - c5_loss: 2.7556 - c1_accuracy: 0.2161 - c2_accuracy: 0.1744 - c3_accuracy: 0.1534 - c4_accuracy: 0.1719 - c5_accuracy: 0.1776 - val_loss: 12.0886 - val_c1_loss: 2.2078 - val_c2_loss: 2.3848 - val_c3_loss: 2.5696 - val_c4_loss: 2.4476 - val_c5_loss: 2.4787 - val_c1_accuracy: 0.3420 - val_c2_accuracy: 0.2837 - val_c3_accuracy: 0.2659 - val_c4_accuracy: 0.2738 - val_c5_accuracy: 0.2858\n",
            "Epoch 20/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 12.6971 - c1_loss: 2.3898 - c2_loss: 2.5212 - c3_loss: 2.6340 - c4_loss: 2.5588 - c5_loss: 2.5932 - c1_accuracy: 0.2448 - c2_accuracy: 0.2261 - c3_accuracy: 0.2041 - c4_accuracy: 0.2186 - c5_accuracy: 0.2182 - val_loss: 11.1127 - val_c1_loss: 1.9957 - val_c2_loss: 2.2052 - val_c3_loss: 2.3874 - val_c4_loss: 2.2617 - val_c5_loss: 2.2627 - val_c1_accuracy: 0.4141 - val_c2_accuracy: 0.3372 - val_c3_accuracy: 0.2988 - val_c4_accuracy: 0.3455 - val_c5_accuracy: 0.3605\n",
            "Epoch 21/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 11.6241 - c1_loss: 2.1747 - c2_loss: 2.2996 - c3_loss: 2.4414 - c4_loss: 2.3392 - c5_loss: 2.3692 - c1_accuracy: 0.3120 - c2_accuracy: 0.2744 - c3_accuracy: 0.2470 - c4_accuracy: 0.2735 - c5_accuracy: 0.2678 - val_loss: 9.8403 - val_c1_loss: 1.7792 - val_c2_loss: 1.9684 - val_c3_loss: 2.1106 - val_c4_loss: 1.9740 - val_c5_loss: 2.0081 - val_c1_accuracy: 0.4765 - val_c2_accuracy: 0.4196 - val_c3_accuracy: 0.3887 - val_c4_accuracy: 0.4274 - val_c5_accuracy: 0.4093\n",
            "Epoch 22/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 10.6161 - c1_loss: 1.9735 - c2_loss: 2.1468 - c3_loss: 2.2170 - c4_loss: 2.1253 - c5_loss: 2.1535 - c1_accuracy: 0.3549 - c2_accuracy: 0.3233 - c3_accuracy: 0.2979 - c4_accuracy: 0.3211 - c5_accuracy: 0.3177 - val_loss: 8.5490 - val_c1_loss: 1.5568 - val_c2_loss: 1.6987 - val_c3_loss: 1.7868 - val_c4_loss: 1.6943 - val_c5_loss: 1.8124 - val_c1_accuracy: 0.5383 - val_c2_accuracy: 0.4590 - val_c3_accuracy: 0.4772 - val_c4_accuracy: 0.5108 - val_c5_accuracy: 0.4484\n",
            "Epoch 23/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 9.5555 - c1_loss: 1.7867 - c2_loss: 1.8915 - c3_loss: 1.9793 - c4_loss: 1.9175 - c5_loss: 1.9804 - c1_accuracy: 0.4155 - c2_accuracy: 0.3835 - c3_accuracy: 0.3664 - c4_accuracy: 0.3720 - c5_accuracy: 0.3639 - val_loss: 7.4356 - val_c1_loss: 1.3575 - val_c2_loss: 1.5372 - val_c3_loss: 1.5553 - val_c4_loss: 1.4426 - val_c5_loss: 1.5429 - val_c1_accuracy: 0.5561 - val_c2_accuracy: 0.5005 - val_c3_accuracy: 0.5094 - val_c4_accuracy: 0.5441 - val_c5_accuracy: 0.5132\n",
            "Epoch 24/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 8.8260 - c1_loss: 1.6797 - c2_loss: 1.7781 - c3_loss: 1.8154 - c4_loss: 1.7673 - c5_loss: 1.7855 - c1_accuracy: 0.4532 - c2_accuracy: 0.4252 - c3_accuracy: 0.4039 - c4_accuracy: 0.4245 - c5_accuracy: 0.4188 - val_loss: 6.4376 - val_c1_loss: 1.1845 - val_c2_loss: 1.2817 - val_c3_loss: 1.3183 - val_c4_loss: 1.3023 - val_c5_loss: 1.3508 - val_c1_accuracy: 0.6494 - val_c2_accuracy: 0.6161 - val_c3_accuracy: 0.5986 - val_c4_accuracy: 0.6048 - val_c5_accuracy: 0.5688\n",
            "Epoch 25/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 7.7631 - c1_loss: 1.4788 - c2_loss: 1.5768 - c3_loss: 1.5618 - c4_loss: 1.5372 - c5_loss: 1.6086 - c1_accuracy: 0.5079 - c2_accuracy: 0.4817 - c3_accuracy: 0.4905 - c4_accuracy: 0.4864 - c5_accuracy: 0.4708 - val_loss: 5.7776 - val_c1_loss: 1.0912 - val_c2_loss: 1.1422 - val_c3_loss: 1.1537 - val_c4_loss: 1.1793 - val_c5_loss: 1.2112 - val_c1_accuracy: 0.6617 - val_c2_accuracy: 0.6604 - val_c3_accuracy: 0.6703 - val_c4_accuracy: 0.6477 - val_c5_accuracy: 0.6370\n",
            "Epoch 26/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 7.0791 - c1_loss: 1.3773 - c2_loss: 1.4278 - c3_loss: 1.4296 - c4_loss: 1.4084 - c5_loss: 1.4361 - c1_accuracy: 0.5482 - c2_accuracy: 0.5352 - c3_accuracy: 0.5192 - c4_accuracy: 0.5342 - c5_accuracy: 0.5235 - val_loss: 5.3253 - val_c1_loss: 1.0045 - val_c2_loss: 1.0746 - val_c3_loss: 1.0640 - val_c4_loss: 1.0596 - val_c5_loss: 1.1225 - val_c1_accuracy: 0.6998 - val_c2_accuracy: 0.6854 - val_c3_accuracy: 0.6913 - val_c4_accuracy: 0.6909 - val_c5_accuracy: 0.6710\n",
            "Epoch 27/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 6.6595 - c1_loss: 1.2979 - c2_loss: 1.3585 - c3_loss: 1.3114 - c4_loss: 1.3073 - c5_loss: 1.3844 - c1_accuracy: 0.5702 - c2_accuracy: 0.5567 - c3_accuracy: 0.5596 - c4_accuracy: 0.5687 - c5_accuracy: 0.5454 - val_loss: 4.5541 - val_c1_loss: 0.8865 - val_c2_loss: 0.9110 - val_c3_loss: 0.8978 - val_c4_loss: 0.8924 - val_c5_loss: 0.9664 - val_c1_accuracy: 0.7242 - val_c2_accuracy: 0.7290 - val_c3_accuracy: 0.7437 - val_c4_accuracy: 0.7400 - val_c5_accuracy: 0.7067\n",
            "Epoch 28/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 6.0776 - c1_loss: 1.2028 - c2_loss: 1.2438 - c3_loss: 1.1798 - c4_loss: 1.1896 - c5_loss: 1.2614 - c1_accuracy: 0.6026 - c2_accuracy: 0.5876 - c3_accuracy: 0.6029 - c4_accuracy: 0.5980 - c5_accuracy: 0.5834 - val_loss: 4.3275 - val_c1_loss: 0.8363 - val_c2_loss: 0.8615 - val_c3_loss: 0.8288 - val_c4_loss: 0.8806 - val_c5_loss: 0.9202 - val_c1_accuracy: 0.7503 - val_c2_accuracy: 0.7571 - val_c3_accuracy: 0.7794 - val_c4_accuracy: 0.7509 - val_c5_accuracy: 0.7290\n",
            "Epoch 29/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 5.5905 - c1_loss: 1.1132 - c2_loss: 1.1358 - c3_loss: 1.0594 - c4_loss: 1.1152 - c5_loss: 1.1668 - c1_accuracy: 0.6293 - c2_accuracy: 0.6248 - c3_accuracy: 0.6446 - c4_accuracy: 0.6311 - c5_accuracy: 0.6205 - val_loss: 3.8343 - val_c1_loss: 0.7578 - val_c2_loss: 0.7931 - val_c3_loss: 0.7080 - val_c4_loss: 0.7538 - val_c5_loss: 0.8216 - val_c1_accuracy: 0.7763 - val_c2_accuracy: 0.7654 - val_c3_accuracy: 0.8062 - val_c4_accuracy: 0.7993 - val_c5_accuracy: 0.7571\n",
            "Epoch 30/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 5.0190 - c1_loss: 1.0164 - c2_loss: 1.0323 - c3_loss: 0.9392 - c4_loss: 0.9720 - c5_loss: 1.0591 - c1_accuracy: 0.6628 - c2_accuracy: 0.6592 - c3_accuracy: 0.6827 - c4_accuracy: 0.6665 - c5_accuracy: 0.6495 - val_loss: 3.3371 - val_c1_loss: 0.6724 - val_c2_loss: 0.6857 - val_c3_loss: 0.6089 - val_c4_loss: 0.6271 - val_c5_loss: 0.7430 - val_c1_accuracy: 0.8034 - val_c2_accuracy: 0.8072 - val_c3_accuracy: 0.8439 - val_c4_accuracy: 0.8326 - val_c5_accuracy: 0.7746\n",
            "Epoch 31/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 4.7810 - c1_loss: 0.9714 - c2_loss: 0.9858 - c3_loss: 0.8868 - c4_loss: 0.9375 - c5_loss: 0.9995 - c1_accuracy: 0.6762 - c2_accuracy: 0.6721 - c3_accuracy: 0.6943 - c4_accuracy: 0.6865 - c5_accuracy: 0.6705 - val_loss: 2.8179 - val_c1_loss: 0.5729 - val_c2_loss: 0.5818 - val_c3_loss: 0.4987 - val_c4_loss: 0.5302 - val_c5_loss: 0.6344 - val_c1_accuracy: 0.8346 - val_c2_accuracy: 0.8333 - val_c3_accuracy: 0.8669 - val_c4_accuracy: 0.8532 - val_c5_accuracy: 0.8069\n",
            "Epoch 32/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 4.1808 - c1_loss: 0.8471 - c2_loss: 0.8712 - c3_loss: 0.7714 - c4_loss: 0.7863 - c5_loss: 0.9048 - c1_accuracy: 0.7190 - c2_accuracy: 0.7018 - c3_accuracy: 0.7311 - c4_accuracy: 0.7312 - c5_accuracy: 0.6927 - val_loss: 2.5370 - val_c1_loss: 0.5170 - val_c2_loss: 0.5133 - val_c3_loss: 0.4432 - val_c4_loss: 0.4930 - val_c5_loss: 0.5705 - val_c1_accuracy: 0.8501 - val_c2_accuracy: 0.8528 - val_c3_accuracy: 0.8775 - val_c4_accuracy: 0.8714 - val_c5_accuracy: 0.8254\n",
            "Epoch 33/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 3.8859 - c1_loss: 0.7813 - c2_loss: 0.7949 - c3_loss: 0.7119 - c4_loss: 0.7628 - c5_loss: 0.8349 - c1_accuracy: 0.7283 - c2_accuracy: 0.7296 - c3_accuracy: 0.7571 - c4_accuracy: 0.7399 - c5_accuracy: 0.7075 - val_loss: 2.3832 - val_c1_loss: 0.4822 - val_c2_loss: 0.4964 - val_c3_loss: 0.4139 - val_c4_loss: 0.4619 - val_c5_loss: 0.5288 - val_c1_accuracy: 0.8525 - val_c2_accuracy: 0.8521 - val_c3_accuracy: 0.8816 - val_c4_accuracy: 0.8744 - val_c5_accuracy: 0.8305\n",
            "Epoch 34/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 3.6924 - c1_loss: 0.7328 - c2_loss: 0.7977 - c3_loss: 0.6756 - c4_loss: 0.6985 - c5_loss: 0.7877 - c1_accuracy: 0.7447 - c2_accuracy: 0.7343 - c3_accuracy: 0.7678 - c4_accuracy: 0.7592 - c5_accuracy: 0.7308 - val_loss: 2.0543 - val_c1_loss: 0.4155 - val_c2_loss: 0.4160 - val_c3_loss: 0.3588 - val_c4_loss: 0.3964 - val_c5_loss: 0.4675 - val_c1_accuracy: 0.8827 - val_c2_accuracy: 0.8840 - val_c3_accuracy: 0.9005 - val_c4_accuracy: 0.8967 - val_c5_accuracy: 0.8552\n",
            "Epoch 35/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 3.3256 - c1_loss: 0.6589 - c2_loss: 0.6794 - c3_loss: 0.6129 - c4_loss: 0.6550 - c5_loss: 0.7193 - c1_accuracy: 0.7739 - c2_accuracy: 0.7683 - c3_accuracy: 0.7900 - c4_accuracy: 0.7717 - c5_accuracy: 0.7546 - val_loss: 1.8704 - val_c1_loss: 0.3772 - val_c2_loss: 0.3820 - val_c3_loss: 0.3329 - val_c4_loss: 0.3714 - val_c5_loss: 0.4068 - val_c1_accuracy: 0.8902 - val_c2_accuracy: 0.8875 - val_c3_accuracy: 0.9077 - val_c4_accuracy: 0.8998 - val_c5_accuracy: 0.8799\n",
            "Epoch 36/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 3.2318 - c1_loss: 0.6610 - c2_loss: 0.6791 - c3_loss: 0.6015 - c4_loss: 0.6229 - c5_loss: 0.6672 - c1_accuracy: 0.7768 - c2_accuracy: 0.7672 - c3_accuracy: 0.7934 - c4_accuracy: 0.7881 - c5_accuracy: 0.7717 - val_loss: 1.7287 - val_c1_loss: 0.3488 - val_c2_loss: 0.3486 - val_c3_loss: 0.3039 - val_c4_loss: 0.3374 - val_c5_loss: 0.3900 - val_c1_accuracy: 0.8978 - val_c2_accuracy: 0.9050 - val_c3_accuracy: 0.9177 - val_c4_accuracy: 0.9129 - val_c5_accuracy: 0.8834\n",
            "Epoch 37/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 2.9938 - c1_loss: 0.6165 - c2_loss: 0.6147 - c3_loss: 0.5569 - c4_loss: 0.5729 - c5_loss: 0.6328 - c1_accuracy: 0.7900 - c2_accuracy: 0.7916 - c3_accuracy: 0.8097 - c4_accuracy: 0.8058 - c5_accuracy: 0.7855 - val_loss: 1.6174 - val_c1_loss: 0.3190 - val_c2_loss: 0.3371 - val_c3_loss: 0.2826 - val_c4_loss: 0.3130 - val_c5_loss: 0.3657 - val_c1_accuracy: 0.9153 - val_c2_accuracy: 0.8991 - val_c3_accuracy: 0.9232 - val_c4_accuracy: 0.9208 - val_c5_accuracy: 0.8906\n",
            "Epoch 38/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 2.8546 - c1_loss: 0.5869 - c2_loss: 0.5778 - c3_loss: 0.5389 - c4_loss: 0.5445 - c5_loss: 0.6065 - c1_accuracy: 0.7986 - c2_accuracy: 0.8009 - c3_accuracy: 0.8099 - c4_accuracy: 0.8136 - c5_accuracy: 0.7906 - val_loss: 1.4955 - val_c1_loss: 0.3022 - val_c2_loss: 0.3064 - val_c3_loss: 0.2641 - val_c4_loss: 0.3064 - val_c5_loss: 0.3164 - val_c1_accuracy: 0.9146 - val_c2_accuracy: 0.9070 - val_c3_accuracy: 0.9293 - val_c4_accuracy: 0.9125 - val_c5_accuracy: 0.9105\n",
            "Epoch 39/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 2.8408 - c1_loss: 0.5702 - c2_loss: 0.5718 - c3_loss: 0.5590 - c4_loss: 0.5469 - c5_loss: 0.5930 - c1_accuracy: 0.8071 - c2_accuracy: 0.8044 - c3_accuracy: 0.8058 - c4_accuracy: 0.8122 - c5_accuracy: 0.7949 - val_loss: 1.5475 - val_c1_loss: 0.3093 - val_c2_loss: 0.3214 - val_c3_loss: 0.2743 - val_c4_loss: 0.3189 - val_c5_loss: 0.3236 - val_c1_accuracy: 0.9187 - val_c2_accuracy: 0.9105 - val_c3_accuracy: 0.9249 - val_c4_accuracy: 0.9211 - val_c5_accuracy: 0.9136\n",
            "Epoch 40/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 2.4813 - c1_loss: 0.4869 - c2_loss: 0.5007 - c3_loss: 0.4686 - c4_loss: 0.5041 - c5_loss: 0.5210 - c1_accuracy: 0.8297 - c2_accuracy: 0.8238 - c3_accuracy: 0.8372 - c4_accuracy: 0.8262 - c5_accuracy: 0.8208 - val_loss: 1.3728 - val_c1_loss: 0.2749 - val_c2_loss: 0.2908 - val_c3_loss: 0.2369 - val_c4_loss: 0.2749 - val_c5_loss: 0.2952 - val_c1_accuracy: 0.9197 - val_c2_accuracy: 0.9208 - val_c3_accuracy: 0.9331 - val_c4_accuracy: 0.9338 - val_c5_accuracy: 0.9136\n",
            "Epoch 41/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 2.4650 - c1_loss: 0.4867 - c2_loss: 0.4984 - c3_loss: 0.4571 - c4_loss: 0.4976 - c5_loss: 0.5251 - c1_accuracy: 0.8324 - c2_accuracy: 0.8288 - c3_accuracy: 0.8465 - c4_accuracy: 0.8334 - c5_accuracy: 0.8152 - val_loss: 1.3606 - val_c1_loss: 0.2642 - val_c2_loss: 0.2864 - val_c3_loss: 0.2497 - val_c4_loss: 0.2698 - val_c5_loss: 0.2906 - val_c1_accuracy: 0.9334 - val_c2_accuracy: 0.9204 - val_c3_accuracy: 0.9321 - val_c4_accuracy: 0.9317 - val_c5_accuracy: 0.9132\n",
            "Epoch 42/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 2.3383 - c1_loss: 0.4674 - c2_loss: 0.4836 - c3_loss: 0.4646 - c4_loss: 0.4373 - c5_loss: 0.4854 - c1_accuracy: 0.8321 - c2_accuracy: 0.8338 - c3_accuracy: 0.8477 - c4_accuracy: 0.8472 - c5_accuracy: 0.8338 - val_loss: 1.2122 - val_c1_loss: 0.2313 - val_c2_loss: 0.2551 - val_c3_loss: 0.2177 - val_c4_loss: 0.2422 - val_c5_loss: 0.2658 - val_c1_accuracy: 0.9379 - val_c2_accuracy: 0.9283 - val_c3_accuracy: 0.9417 - val_c4_accuracy: 0.9386 - val_c5_accuracy: 0.9208\n",
            "Epoch 43/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 2.0955 - c1_loss: 0.4150 - c2_loss: 0.4295 - c3_loss: 0.4140 - c4_loss: 0.4030 - c5_loss: 0.4341 - c1_accuracy: 0.8566 - c2_accuracy: 0.8488 - c3_accuracy: 0.8603 - c4_accuracy: 0.8681 - c5_accuracy: 0.8488 - val_loss: 1.2541 - val_c1_loss: 0.2387 - val_c2_loss: 0.2629 - val_c3_loss: 0.2394 - val_c4_loss: 0.2475 - val_c5_loss: 0.2656 - val_c1_accuracy: 0.9386 - val_c2_accuracy: 0.9310 - val_c3_accuracy: 0.9324 - val_c4_accuracy: 0.9424 - val_c5_accuracy: 0.9286\n",
            "Epoch 44/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 2.3291 - c1_loss: 0.4673 - c2_loss: 0.4710 - c3_loss: 0.4826 - c4_loss: 0.4381 - c5_loss: 0.4701 - c1_accuracy: 0.8438 - c2_accuracy: 0.8397 - c3_accuracy: 0.8403 - c4_accuracy: 0.8543 - c5_accuracy: 0.8425 - val_loss: 1.1958 - val_c1_loss: 0.2290 - val_c2_loss: 0.2465 - val_c3_loss: 0.2145 - val_c4_loss: 0.2435 - val_c5_loss: 0.2623 - val_c1_accuracy: 0.9369 - val_c2_accuracy: 0.9317 - val_c3_accuracy: 0.9413 - val_c4_accuracy: 0.9403 - val_c5_accuracy: 0.9252\n",
            "Epoch 45/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 2.1477 - c1_loss: 0.4161 - c2_loss: 0.4431 - c3_loss: 0.4425 - c4_loss: 0.3944 - c5_loss: 0.4517 - c1_accuracy: 0.8574 - c2_accuracy: 0.8456 - c3_accuracy: 0.8480 - c4_accuracy: 0.8635 - c5_accuracy: 0.8516 - val_loss: 1.0834 - val_c1_loss: 0.2128 - val_c2_loss: 0.2259 - val_c3_loss: 0.1982 - val_c4_loss: 0.2179 - val_c5_loss: 0.2287 - val_c1_accuracy: 0.9441 - val_c2_accuracy: 0.9379 - val_c3_accuracy: 0.9444 - val_c4_accuracy: 0.9444 - val_c5_accuracy: 0.9369\n",
            "Epoch 46/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.9761 - c1_loss: 0.4147 - c2_loss: 0.3940 - c3_loss: 0.3684 - c4_loss: 0.3795 - c5_loss: 0.4195 - c1_accuracy: 0.8618 - c2_accuracy: 0.8628 - c3_accuracy: 0.8713 - c4_accuracy: 0.8678 - c5_accuracy: 0.8555 - val_loss: 1.0374 - val_c1_loss: 0.1935 - val_c2_loss: 0.2172 - val_c3_loss: 0.1915 - val_c4_loss: 0.2145 - val_c5_loss: 0.2208 - val_c1_accuracy: 0.9499 - val_c2_accuracy: 0.9427 - val_c3_accuracy: 0.9475 - val_c4_accuracy: 0.9482 - val_c5_accuracy: 0.9386\n",
            "Epoch 47/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 2.0870 - c1_loss: 0.4119 - c2_loss: 0.4218 - c3_loss: 0.3977 - c4_loss: 0.4285 - c5_loss: 0.4271 - c1_accuracy: 0.8613 - c2_accuracy: 0.8533 - c3_accuracy: 0.8649 - c4_accuracy: 0.8549 - c5_accuracy: 0.8540 - val_loss: 1.1388 - val_c1_loss: 0.2162 - val_c2_loss: 0.2451 - val_c3_loss: 0.2039 - val_c4_loss: 0.2297 - val_c5_loss: 0.2439 - val_c1_accuracy: 0.9441 - val_c2_accuracy: 0.9290 - val_c3_accuracy: 0.9444 - val_c4_accuracy: 0.9400 - val_c5_accuracy: 0.9280\n",
            "Epoch 48/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 2.0344 - c1_loss: 0.3980 - c2_loss: 0.4089 - c3_loss: 0.4054 - c4_loss: 0.4109 - c5_loss: 0.4112 - c1_accuracy: 0.8609 - c2_accuracy: 0.8603 - c3_accuracy: 0.8608 - c4_accuracy: 0.8618 - c5_accuracy: 0.8602 - val_loss: 1.0289 - val_c1_loss: 0.1960 - val_c2_loss: 0.2146 - val_c3_loss: 0.1860 - val_c4_loss: 0.2172 - val_c5_loss: 0.2151 - val_c1_accuracy: 0.9561 - val_c2_accuracy: 0.9468 - val_c3_accuracy: 0.9527 - val_c4_accuracy: 0.9451 - val_c5_accuracy: 0.9451\n",
            "Epoch 49/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.7927 - c1_loss: 0.3378 - c2_loss: 0.3613 - c3_loss: 0.3414 - c4_loss: 0.3623 - c5_loss: 0.3898 - c1_accuracy: 0.8853 - c2_accuracy: 0.8733 - c3_accuracy: 0.8819 - c4_accuracy: 0.8802 - c5_accuracy: 0.8665 - val_loss: 0.9936 - val_c1_loss: 0.1971 - val_c2_loss: 0.2037 - val_c3_loss: 0.1756 - val_c4_loss: 0.2044 - val_c5_loss: 0.2128 - val_c1_accuracy: 0.9509 - val_c2_accuracy: 0.9437 - val_c3_accuracy: 0.9581 - val_c4_accuracy: 0.9520 - val_c5_accuracy: 0.9420\n",
            "Epoch 50/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 1.9871 - c1_loss: 0.3764 - c2_loss: 0.4020 - c3_loss: 0.4271 - c4_loss: 0.3806 - c5_loss: 0.4010 - c1_accuracy: 0.8743 - c2_accuracy: 0.8646 - c3_accuracy: 0.8552 - c4_accuracy: 0.8687 - c5_accuracy: 0.8650 - val_loss: 1.0020 - val_c1_loss: 0.1916 - val_c2_loss: 0.1956 - val_c3_loss: 0.1768 - val_c4_loss: 0.2173 - val_c5_loss: 0.2207 - val_c1_accuracy: 0.9520 - val_c2_accuracy: 0.9475 - val_c3_accuracy: 0.9544 - val_c4_accuracy: 0.9455 - val_c5_accuracy: 0.9413\n",
            "Epoch 51/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 2.0182 - c1_loss: 0.4000 - c2_loss: 0.3932 - c3_loss: 0.4047 - c4_loss: 0.4010 - c5_loss: 0.4192 - c1_accuracy: 0.8609 - c2_accuracy: 0.8705 - c3_accuracy: 0.8612 - c4_accuracy: 0.8591 - c5_accuracy: 0.8546 - val_loss: 0.9712 - val_c1_loss: 0.1860 - val_c2_loss: 0.2000 - val_c3_loss: 0.1720 - val_c4_loss: 0.1983 - val_c5_loss: 0.2148 - val_c1_accuracy: 0.9527 - val_c2_accuracy: 0.9458 - val_c3_accuracy: 0.9540 - val_c4_accuracy: 0.9547 - val_c5_accuracy: 0.9410\n",
            "Epoch 52/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 1.6736 - c1_loss: 0.3336 - c2_loss: 0.3347 - c3_loss: 0.3290 - c4_loss: 0.3231 - c5_loss: 0.3532 - c1_accuracy: 0.8868 - c2_accuracy: 0.8849 - c3_accuracy: 0.8868 - c4_accuracy: 0.8934 - c5_accuracy: 0.8805 - val_loss: 0.9412 - val_c1_loss: 0.1792 - val_c2_loss: 0.1936 - val_c3_loss: 0.1722 - val_c4_loss: 0.1927 - val_c5_loss: 0.2035 - val_c1_accuracy: 0.9568 - val_c2_accuracy: 0.9492 - val_c3_accuracy: 0.9588 - val_c4_accuracy: 0.9533 - val_c5_accuracy: 0.9431\n",
            "Epoch 53/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.8264 - c1_loss: 0.3518 - c2_loss: 0.3789 - c3_loss: 0.3404 - c4_loss: 0.3693 - c5_loss: 0.3859 - c1_accuracy: 0.8784 - c2_accuracy: 0.8716 - c3_accuracy: 0.8853 - c4_accuracy: 0.8756 - c5_accuracy: 0.8658 - val_loss: 0.9354 - val_c1_loss: 0.1831 - val_c2_loss: 0.1892 - val_c3_loss: 0.1681 - val_c4_loss: 0.1983 - val_c5_loss: 0.1967 - val_c1_accuracy: 0.9568 - val_c2_accuracy: 0.9551 - val_c3_accuracy: 0.9599 - val_c4_accuracy: 0.9523 - val_c5_accuracy: 0.9485\n",
            "Epoch 54/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 1.6894 - c1_loss: 0.3371 - c2_loss: 0.3229 - c3_loss: 0.3379 - c4_loss: 0.3341 - c5_loss: 0.3574 - c1_accuracy: 0.8855 - c2_accuracy: 0.8916 - c3_accuracy: 0.8834 - c4_accuracy: 0.8877 - c5_accuracy: 0.8768 - val_loss: 0.9031 - val_c1_loss: 0.1792 - val_c2_loss: 0.1880 - val_c3_loss: 0.1566 - val_c4_loss: 0.1901 - val_c5_loss: 0.1892 - val_c1_accuracy: 0.9571 - val_c2_accuracy: 0.9530 - val_c3_accuracy: 0.9605 - val_c4_accuracy: 0.9523 - val_c5_accuracy: 0.9492\n",
            "Epoch 55/500\n",
            "69/69 [==============================] - 2s 29ms/step - loss: 1.7529 - c1_loss: 0.3608 - c2_loss: 0.3568 - c3_loss: 0.3670 - c4_loss: 0.3411 - c5_loss: 0.3272 - c1_accuracy: 0.8780 - c2_accuracy: 0.8766 - c3_accuracy: 0.8762 - c4_accuracy: 0.8822 - c5_accuracy: 0.8869 - val_loss: 0.8830 - val_c1_loss: 0.1737 - val_c2_loss: 0.1834 - val_c3_loss: 0.1541 - val_c4_loss: 0.1842 - val_c5_loss: 0.1877 - val_c1_accuracy: 0.9554 - val_c2_accuracy: 0.9575 - val_c3_accuracy: 0.9657 - val_c4_accuracy: 0.9527 - val_c5_accuracy: 0.9482\n",
            "Epoch 56/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.5030 - c1_loss: 0.3017 - c2_loss: 0.3049 - c3_loss: 0.2863 - c4_loss: 0.2928 - c5_loss: 0.3174 - c1_accuracy: 0.9003 - c2_accuracy: 0.8897 - c3_accuracy: 0.9006 - c4_accuracy: 0.9007 - c5_accuracy: 0.8896 - val_loss: 0.8428 - val_c1_loss: 0.1667 - val_c2_loss: 0.1717 - val_c3_loss: 0.1509 - val_c4_loss: 0.1788 - val_c5_loss: 0.1747 - val_c1_accuracy: 0.9571 - val_c2_accuracy: 0.9561 - val_c3_accuracy: 0.9578 - val_c4_accuracy: 0.9561 - val_c5_accuracy: 0.9499\n",
            "Epoch 57/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.5450 - c1_loss: 0.3103 - c2_loss: 0.3248 - c3_loss: 0.2881 - c4_loss: 0.2967 - c5_loss: 0.3250 - c1_accuracy: 0.8944 - c2_accuracy: 0.8875 - c3_accuracy: 0.9025 - c4_accuracy: 0.8985 - c5_accuracy: 0.8893 - val_loss: 0.8302 - val_c1_loss: 0.1722 - val_c2_loss: 0.1677 - val_c3_loss: 0.1491 - val_c4_loss: 0.1784 - val_c5_loss: 0.1627 - val_c1_accuracy: 0.9581 - val_c2_accuracy: 0.9592 - val_c3_accuracy: 0.9595 - val_c4_accuracy: 0.9578 - val_c5_accuracy: 0.9554\n",
            "Epoch 58/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.4025 - c1_loss: 0.2751 - c2_loss: 0.2873 - c3_loss: 0.2729 - c4_loss: 0.2808 - c5_loss: 0.2865 - c1_accuracy: 0.9078 - c2_accuracy: 0.9016 - c3_accuracy: 0.9053 - c4_accuracy: 0.9075 - c5_accuracy: 0.9016 - val_loss: 0.8365 - val_c1_loss: 0.1657 - val_c2_loss: 0.1786 - val_c3_loss: 0.1477 - val_c4_loss: 0.1743 - val_c5_loss: 0.1701 - val_c1_accuracy: 0.9592 - val_c2_accuracy: 0.9537 - val_c3_accuracy: 0.9630 - val_c4_accuracy: 0.9588 - val_c5_accuracy: 0.9551\n",
            "Epoch 59/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.5700 - c1_loss: 0.3231 - c2_loss: 0.3250 - c3_loss: 0.2879 - c4_loss: 0.3127 - c5_loss: 0.3213 - c1_accuracy: 0.8935 - c2_accuracy: 0.8902 - c3_accuracy: 0.9006 - c4_accuracy: 0.8940 - c5_accuracy: 0.8852 - val_loss: 0.8560 - val_c1_loss: 0.1710 - val_c2_loss: 0.1788 - val_c3_loss: 0.1446 - val_c4_loss: 0.1897 - val_c5_loss: 0.1719 - val_c1_accuracy: 0.9571 - val_c2_accuracy: 0.9551 - val_c3_accuracy: 0.9605 - val_c4_accuracy: 0.9513 - val_c5_accuracy: 0.9516\n",
            "Epoch 60/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.3515 - c1_loss: 0.2659 - c2_loss: 0.2906 - c3_loss: 0.2719 - c4_loss: 0.2528 - c5_loss: 0.2702 - c1_accuracy: 0.9096 - c2_accuracy: 0.8981 - c3_accuracy: 0.9065 - c4_accuracy: 0.9112 - c5_accuracy: 0.9043 - val_loss: 0.7883 - val_c1_loss: 0.1543 - val_c2_loss: 0.1687 - val_c3_loss: 0.1368 - val_c4_loss: 0.1676 - val_c5_loss: 0.1609 - val_c1_accuracy: 0.9585 - val_c2_accuracy: 0.9588 - val_c3_accuracy: 0.9633 - val_c4_accuracy: 0.9609 - val_c5_accuracy: 0.9544\n",
            "Epoch 61/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.3450 - c1_loss: 0.2526 - c2_loss: 0.2759 - c3_loss: 0.2789 - c4_loss: 0.2602 - c5_loss: 0.2774 - c1_accuracy: 0.9182 - c2_accuracy: 0.9060 - c3_accuracy: 0.9037 - c4_accuracy: 0.9071 - c5_accuracy: 0.9060 - val_loss: 0.7573 - val_c1_loss: 0.1510 - val_c2_loss: 0.1632 - val_c3_loss: 0.1356 - val_c4_loss: 0.1612 - val_c5_loss: 0.1463 - val_c1_accuracy: 0.9616 - val_c2_accuracy: 0.9578 - val_c3_accuracy: 0.9643 - val_c4_accuracy: 0.9647 - val_c5_accuracy: 0.9605\n",
            "Epoch 62/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.2848 - c1_loss: 0.2541 - c2_loss: 0.2661 - c3_loss: 0.2542 - c4_loss: 0.2551 - c5_loss: 0.2553 - c1_accuracy: 0.9160 - c2_accuracy: 0.9122 - c3_accuracy: 0.9132 - c4_accuracy: 0.9103 - c5_accuracy: 0.9125 - val_loss: 0.7745 - val_c1_loss: 0.1470 - val_c2_loss: 0.1670 - val_c3_loss: 0.1481 - val_c4_loss: 0.1590 - val_c5_loss: 0.1535 - val_c1_accuracy: 0.9633 - val_c2_accuracy: 0.9595 - val_c3_accuracy: 0.9568 - val_c4_accuracy: 0.9602 - val_c5_accuracy: 0.9571\n",
            "Epoch 63/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.4595 - c1_loss: 0.2860 - c2_loss: 0.3201 - c3_loss: 0.2804 - c4_loss: 0.2782 - c5_loss: 0.2947 - c1_accuracy: 0.9021 - c2_accuracy: 0.8912 - c3_accuracy: 0.9044 - c4_accuracy: 0.9075 - c5_accuracy: 0.8985 - val_loss: 0.8282 - val_c1_loss: 0.1566 - val_c2_loss: 0.1826 - val_c3_loss: 0.1325 - val_c4_loss: 0.1771 - val_c5_loss: 0.1793 - val_c1_accuracy: 0.9605 - val_c2_accuracy: 0.9503 - val_c3_accuracy: 0.9657 - val_c4_accuracy: 0.9561 - val_c5_accuracy: 0.9461\n",
            "Epoch 64/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.5036 - c1_loss: 0.2933 - c2_loss: 0.3015 - c3_loss: 0.2921 - c4_loss: 0.2924 - c5_loss: 0.3244 - c1_accuracy: 0.9013 - c2_accuracy: 0.9000 - c3_accuracy: 0.9038 - c4_accuracy: 0.9021 - c5_accuracy: 0.8938 - val_loss: 0.7777 - val_c1_loss: 0.1470 - val_c2_loss: 0.1733 - val_c3_loss: 0.1309 - val_c4_loss: 0.1652 - val_c5_loss: 0.1613 - val_c1_accuracy: 0.9643 - val_c2_accuracy: 0.9533 - val_c3_accuracy: 0.9633 - val_c4_accuracy: 0.9602 - val_c5_accuracy: 0.9554\n",
            "Epoch 65/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.3779 - c1_loss: 0.2670 - c2_loss: 0.2702 - c3_loss: 0.2810 - c4_loss: 0.2807 - c5_loss: 0.2790 - c1_accuracy: 0.9113 - c2_accuracy: 0.9037 - c3_accuracy: 0.9075 - c4_accuracy: 0.9043 - c5_accuracy: 0.9034 - val_loss: 0.7926 - val_c1_loss: 0.1511 - val_c2_loss: 0.1697 - val_c3_loss: 0.1339 - val_c4_loss: 0.1703 - val_c5_loss: 0.1675 - val_c1_accuracy: 0.9633 - val_c2_accuracy: 0.9561 - val_c3_accuracy: 0.9667 - val_c4_accuracy: 0.9605 - val_c5_accuracy: 0.9554\n",
            "Epoch 66/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.2424 - c1_loss: 0.2338 - c2_loss: 0.2730 - c3_loss: 0.2434 - c4_loss: 0.2342 - c5_loss: 0.2582 - c1_accuracy: 0.9240 - c2_accuracy: 0.9094 - c3_accuracy: 0.9213 - c4_accuracy: 0.9203 - c5_accuracy: 0.9124 - val_loss: 0.7588 - val_c1_loss: 0.1472 - val_c2_loss: 0.1596 - val_c3_loss: 0.1322 - val_c4_loss: 0.1601 - val_c5_loss: 0.1598 - val_c1_accuracy: 0.9664 - val_c2_accuracy: 0.9612 - val_c3_accuracy: 0.9667 - val_c4_accuracy: 0.9602 - val_c5_accuracy: 0.9571\n",
            "Epoch 67/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.4700 - c1_loss: 0.3216 - c2_loss: 0.2872 - c3_loss: 0.2847 - c4_loss: 0.2833 - c5_loss: 0.2931 - c1_accuracy: 0.8952 - c2_accuracy: 0.9043 - c3_accuracy: 0.9049 - c4_accuracy: 0.9052 - c5_accuracy: 0.9024 - val_loss: 0.7760 - val_c1_loss: 0.1507 - val_c2_loss: 0.1660 - val_c3_loss: 0.1354 - val_c4_loss: 0.1654 - val_c5_loss: 0.1585 - val_c1_accuracy: 0.9654 - val_c2_accuracy: 0.9612 - val_c3_accuracy: 0.9643 - val_c4_accuracy: 0.9578 - val_c5_accuracy: 0.9575\n",
            "Epoch 68/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.3349 - c1_loss: 0.2505 - c2_loss: 0.2732 - c3_loss: 0.2669 - c4_loss: 0.2622 - c5_loss: 0.2821 - c1_accuracy: 0.9153 - c2_accuracy: 0.9081 - c3_accuracy: 0.9087 - c4_accuracy: 0.9131 - c5_accuracy: 0.9044 - val_loss: 0.7805 - val_c1_loss: 0.1470 - val_c2_loss: 0.1703 - val_c3_loss: 0.1468 - val_c4_loss: 0.1620 - val_c5_loss: 0.1545 - val_c1_accuracy: 0.9650 - val_c2_accuracy: 0.9561 - val_c3_accuracy: 0.9619 - val_c4_accuracy: 0.9643 - val_c5_accuracy: 0.9537\n",
            "Epoch 69/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.5341 - c1_loss: 0.3308 - c2_loss: 0.3011 - c3_loss: 0.2864 - c4_loss: 0.3171 - c5_loss: 0.2987 - c1_accuracy: 0.8931 - c2_accuracy: 0.8972 - c3_accuracy: 0.9040 - c4_accuracy: 0.8949 - c5_accuracy: 0.9010 - val_loss: 0.7918 - val_c1_loss: 0.1594 - val_c2_loss: 0.1751 - val_c3_loss: 0.1403 - val_c4_loss: 0.1637 - val_c5_loss: 0.1533 - val_c1_accuracy: 0.9592 - val_c2_accuracy: 0.9551 - val_c3_accuracy: 0.9643 - val_c4_accuracy: 0.9633 - val_c5_accuracy: 0.9588\n",
            "Epoch 70/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.2431 - c1_loss: 0.2397 - c2_loss: 0.2515 - c3_loss: 0.2456 - c4_loss: 0.2357 - c5_loss: 0.2705 - c1_accuracy: 0.9197 - c2_accuracy: 0.9121 - c3_accuracy: 0.9152 - c4_accuracy: 0.9209 - c5_accuracy: 0.9103 - val_loss: 0.7780 - val_c1_loss: 0.1502 - val_c2_loss: 0.1732 - val_c3_loss: 0.1404 - val_c4_loss: 0.1673 - val_c5_loss: 0.1469 - val_c1_accuracy: 0.9640 - val_c2_accuracy: 0.9571 - val_c3_accuracy: 0.9647 - val_c4_accuracy: 0.9575 - val_c5_accuracy: 0.9588\n",
            "Epoch 71/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.2702 - c1_loss: 0.2675 - c2_loss: 0.2490 - c3_loss: 0.2575 - c4_loss: 0.2463 - c5_loss: 0.2499 - c1_accuracy: 0.9106 - c2_accuracy: 0.9140 - c3_accuracy: 0.9107 - c4_accuracy: 0.9144 - c5_accuracy: 0.9157 - val_loss: 0.7536 - val_c1_loss: 0.1483 - val_c2_loss: 0.1625 - val_c3_loss: 0.1367 - val_c4_loss: 0.1669 - val_c5_loss: 0.1393 - val_c1_accuracy: 0.9647 - val_c2_accuracy: 0.9602 - val_c3_accuracy: 0.9654 - val_c4_accuracy: 0.9585 - val_c5_accuracy: 0.9605\n",
            "Epoch 72/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.1142 - c1_loss: 0.2131 - c2_loss: 0.2223 - c3_loss: 0.2304 - c4_loss: 0.2246 - c5_loss: 0.2238 - c1_accuracy: 0.9274 - c2_accuracy: 0.9234 - c3_accuracy: 0.9209 - c4_accuracy: 0.9237 - c5_accuracy: 0.9230 - val_loss: 0.7296 - val_c1_loss: 0.1481 - val_c2_loss: 0.1525 - val_c3_loss: 0.1304 - val_c4_loss: 0.1588 - val_c5_loss: 0.1399 - val_c1_accuracy: 0.9623 - val_c2_accuracy: 0.9616 - val_c3_accuracy: 0.9650 - val_c4_accuracy: 0.9640 - val_c5_accuracy: 0.9605\n",
            "Epoch 73/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.1956 - c1_loss: 0.2369 - c2_loss: 0.2468 - c3_loss: 0.2307 - c4_loss: 0.2333 - c5_loss: 0.2479 - c1_accuracy: 0.9169 - c2_accuracy: 0.9191 - c3_accuracy: 0.9209 - c4_accuracy: 0.9222 - c5_accuracy: 0.9162 - val_loss: 0.7125 - val_c1_loss: 0.1441 - val_c2_loss: 0.1550 - val_c3_loss: 0.1243 - val_c4_loss: 0.1556 - val_c5_loss: 0.1336 - val_c1_accuracy: 0.9657 - val_c2_accuracy: 0.9581 - val_c3_accuracy: 0.9691 - val_c4_accuracy: 0.9616 - val_c5_accuracy: 0.9643\n",
            "Epoch 74/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.1123 - c1_loss: 0.2086 - c2_loss: 0.2411 - c3_loss: 0.2105 - c4_loss: 0.2189 - c5_loss: 0.2332 - c1_accuracy: 0.9275 - c2_accuracy: 0.9159 - c3_accuracy: 0.9285 - c4_accuracy: 0.9240 - c5_accuracy: 0.9207 - val_loss: 0.6878 - val_c1_loss: 0.1320 - val_c2_loss: 0.1496 - val_c3_loss: 0.1236 - val_c4_loss: 0.1518 - val_c5_loss: 0.1309 - val_c1_accuracy: 0.9712 - val_c2_accuracy: 0.9650 - val_c3_accuracy: 0.9684 - val_c4_accuracy: 0.9630 - val_c5_accuracy: 0.9630\n",
            "Epoch 75/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.0430 - c1_loss: 0.2043 - c2_loss: 0.2117 - c3_loss: 0.2107 - c4_loss: 0.2115 - c5_loss: 0.2048 - c1_accuracy: 0.9328 - c2_accuracy: 0.9288 - c3_accuracy: 0.9306 - c4_accuracy: 0.9260 - c5_accuracy: 0.9290 - val_loss: 0.6781 - val_c1_loss: 0.1315 - val_c2_loss: 0.1479 - val_c3_loss: 0.1233 - val_c4_loss: 0.1420 - val_c5_loss: 0.1334 - val_c1_accuracy: 0.9708 - val_c2_accuracy: 0.9630 - val_c3_accuracy: 0.9664 - val_c4_accuracy: 0.9667 - val_c5_accuracy: 0.9647\n",
            "Epoch 76/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.0352 - c1_loss: 0.1963 - c2_loss: 0.2196 - c3_loss: 0.2016 - c4_loss: 0.2141 - c5_loss: 0.2035 - c1_accuracy: 0.9349 - c2_accuracy: 0.9262 - c3_accuracy: 0.9324 - c4_accuracy: 0.9253 - c5_accuracy: 0.9282 - val_loss: 0.6984 - val_c1_loss: 0.1391 - val_c2_loss: 0.1487 - val_c3_loss: 0.1307 - val_c4_loss: 0.1438 - val_c5_loss: 0.1360 - val_c1_accuracy: 0.9667 - val_c2_accuracy: 0.9650 - val_c3_accuracy: 0.9667 - val_c4_accuracy: 0.9681 - val_c5_accuracy: 0.9633\n",
            "Epoch 77/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.0710 - c1_loss: 0.2264 - c2_loss: 0.2084 - c3_loss: 0.2058 - c4_loss: 0.2190 - c5_loss: 0.2114 - c1_accuracy: 0.9222 - c2_accuracy: 0.9324 - c3_accuracy: 0.9299 - c4_accuracy: 0.9260 - c5_accuracy: 0.9288 - val_loss: 0.6875 - val_c1_loss: 0.1358 - val_c2_loss: 0.1474 - val_c3_loss: 0.1226 - val_c4_loss: 0.1479 - val_c5_loss: 0.1337 - val_c1_accuracy: 0.9688 - val_c2_accuracy: 0.9667 - val_c3_accuracy: 0.9708 - val_c4_accuracy: 0.9664 - val_c5_accuracy: 0.9664\n",
            "Epoch 78/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.2579 - c1_loss: 0.2605 - c2_loss: 0.2654 - c3_loss: 0.2388 - c4_loss: 0.2391 - c5_loss: 0.2540 - c1_accuracy: 0.9191 - c2_accuracy: 0.9096 - c3_accuracy: 0.9177 - c4_accuracy: 0.9200 - c5_accuracy: 0.9137 - val_loss: 0.7340 - val_c1_loss: 0.1520 - val_c2_loss: 0.1592 - val_c3_loss: 0.1305 - val_c4_loss: 0.1509 - val_c5_loss: 0.1414 - val_c1_accuracy: 0.9657 - val_c2_accuracy: 0.9599 - val_c3_accuracy: 0.9654 - val_c4_accuracy: 0.9643 - val_c5_accuracy: 0.9602\n",
            "Epoch 79/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.1213 - c1_loss: 0.2333 - c2_loss: 0.2293 - c3_loss: 0.2347 - c4_loss: 0.2102 - c5_loss: 0.2137 - c1_accuracy: 0.9262 - c2_accuracy: 0.9222 - c3_accuracy: 0.9224 - c4_accuracy: 0.9247 - c5_accuracy: 0.9266 - val_loss: 0.6994 - val_c1_loss: 0.1370 - val_c2_loss: 0.1513 - val_c3_loss: 0.1277 - val_c4_loss: 0.1468 - val_c5_loss: 0.1366 - val_c1_accuracy: 0.9695 - val_c2_accuracy: 0.9616 - val_c3_accuracy: 0.9688 - val_c4_accuracy: 0.9640 - val_c5_accuracy: 0.9633\n",
            "Epoch 80/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.1982 - c1_loss: 0.2404 - c2_loss: 0.2508 - c3_loss: 0.2309 - c4_loss: 0.2245 - c5_loss: 0.2516 - c1_accuracy: 0.9184 - c2_accuracy: 0.9153 - c3_accuracy: 0.9228 - c4_accuracy: 0.9240 - c5_accuracy: 0.9175 - val_loss: 0.7190 - val_c1_loss: 0.1386 - val_c2_loss: 0.1545 - val_c3_loss: 0.1245 - val_c4_loss: 0.1507 - val_c5_loss: 0.1506 - val_c1_accuracy: 0.9688 - val_c2_accuracy: 0.9612 - val_c3_accuracy: 0.9688 - val_c4_accuracy: 0.9636 - val_c5_accuracy: 0.9588\n",
            "Epoch 81/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.0609 - c1_loss: 0.1916 - c2_loss: 0.2106 - c3_loss: 0.2056 - c4_loss: 0.2242 - c5_loss: 0.2289 - c1_accuracy: 0.9328 - c2_accuracy: 0.9277 - c3_accuracy: 0.9316 - c4_accuracy: 0.9268 - c5_accuracy: 0.9238 - val_loss: 0.6957 - val_c1_loss: 0.1383 - val_c2_loss: 0.1524 - val_c3_loss: 0.1242 - val_c4_loss: 0.1416 - val_c5_loss: 0.1392 - val_c1_accuracy: 0.9678 - val_c2_accuracy: 0.9623 - val_c3_accuracy: 0.9667 - val_c4_accuracy: 0.9671 - val_c5_accuracy: 0.9654\n",
            "Epoch 82/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9914 - c1_loss: 0.2115 - c2_loss: 0.1932 - c3_loss: 0.1923 - c4_loss: 0.1843 - c5_loss: 0.2101 - c1_accuracy: 0.9275 - c2_accuracy: 0.9327 - c3_accuracy: 0.9377 - c4_accuracy: 0.9375 - c5_accuracy: 0.9290 - val_loss: 0.6948 - val_c1_loss: 0.1361 - val_c2_loss: 0.1544 - val_c3_loss: 0.1203 - val_c4_loss: 0.1404 - val_c5_loss: 0.1437 - val_c1_accuracy: 0.9681 - val_c2_accuracy: 0.9578 - val_c3_accuracy: 0.9698 - val_c4_accuracy: 0.9664 - val_c5_accuracy: 0.9643\n",
            "Epoch 83/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.1935 - c1_loss: 0.2258 - c2_loss: 0.2527 - c3_loss: 0.2379 - c4_loss: 0.2452 - c5_loss: 0.2319 - c1_accuracy: 0.9219 - c2_accuracy: 0.9163 - c3_accuracy: 0.9168 - c4_accuracy: 0.9141 - c5_accuracy: 0.9218 - val_loss: 0.7189 - val_c1_loss: 0.1385 - val_c2_loss: 0.1564 - val_c3_loss: 0.1342 - val_c4_loss: 0.1446 - val_c5_loss: 0.1453 - val_c1_accuracy: 0.9654 - val_c2_accuracy: 0.9588 - val_c3_accuracy: 0.9640 - val_c4_accuracy: 0.9678 - val_c5_accuracy: 0.9609\n",
            "Epoch 84/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.0653 - c1_loss: 0.1923 - c2_loss: 0.2229 - c3_loss: 0.2114 - c4_loss: 0.2236 - c5_loss: 0.2150 - c1_accuracy: 0.9343 - c2_accuracy: 0.9240 - c3_accuracy: 0.9284 - c4_accuracy: 0.9224 - c5_accuracy: 0.9263 - val_loss: 0.6832 - val_c1_loss: 0.1384 - val_c2_loss: 0.1469 - val_c3_loss: 0.1256 - val_c4_loss: 0.1439 - val_c5_loss: 0.1284 - val_c1_accuracy: 0.9640 - val_c2_accuracy: 0.9636 - val_c3_accuracy: 0.9640 - val_c4_accuracy: 0.9691 - val_c5_accuracy: 0.9671\n",
            "Epoch 85/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.1846 - c1_loss: 0.2406 - c2_loss: 0.2260 - c3_loss: 0.2605 - c4_loss: 0.2303 - c5_loss: 0.2273 - c1_accuracy: 0.9196 - c2_accuracy: 0.9215 - c3_accuracy: 0.9132 - c4_accuracy: 0.9219 - c5_accuracy: 0.9224 - val_loss: 0.7466 - val_c1_loss: 0.1382 - val_c2_loss: 0.1682 - val_c3_loss: 0.1377 - val_c4_loss: 0.1596 - val_c5_loss: 0.1429 - val_c1_accuracy: 0.9674 - val_c2_accuracy: 0.9554 - val_c3_accuracy: 0.9650 - val_c4_accuracy: 0.9630 - val_c5_accuracy: 0.9605\n",
            "Epoch 86/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.1218 - c1_loss: 0.2290 - c2_loss: 0.2317 - c3_loss: 0.2419 - c4_loss: 0.2184 - c5_loss: 0.2008 - c1_accuracy: 0.9271 - c2_accuracy: 0.9252 - c3_accuracy: 0.9185 - c4_accuracy: 0.9303 - c5_accuracy: 0.9312 - val_loss: 0.6909 - val_c1_loss: 0.1300 - val_c2_loss: 0.1544 - val_c3_loss: 0.1310 - val_c4_loss: 0.1398 - val_c5_loss: 0.1357 - val_c1_accuracy: 0.9671 - val_c2_accuracy: 0.9612 - val_c3_accuracy: 0.9664 - val_c4_accuracy: 0.9684 - val_c5_accuracy: 0.9626\n",
            "Epoch 87/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.0008 - c1_loss: 0.1873 - c2_loss: 0.2055 - c3_loss: 0.1908 - c4_loss: 0.2051 - c5_loss: 0.2122 - c1_accuracy: 0.9366 - c2_accuracy: 0.9307 - c3_accuracy: 0.9347 - c4_accuracy: 0.9296 - c5_accuracy: 0.9278 - val_loss: 0.6501 - val_c1_loss: 0.1208 - val_c2_loss: 0.1545 - val_c3_loss: 0.1155 - val_c4_loss: 0.1375 - val_c5_loss: 0.1218 - val_c1_accuracy: 0.9736 - val_c2_accuracy: 0.9612 - val_c3_accuracy: 0.9688 - val_c4_accuracy: 0.9698 - val_c5_accuracy: 0.9678\n",
            "Epoch 88/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.1200 - c1_loss: 0.2125 - c2_loss: 0.2480 - c3_loss: 0.2204 - c4_loss: 0.2160 - c5_loss: 0.2232 - c1_accuracy: 0.9287 - c2_accuracy: 0.9180 - c3_accuracy: 0.9288 - c4_accuracy: 0.9228 - c5_accuracy: 0.9285 - val_loss: 0.6598 - val_c1_loss: 0.1269 - val_c2_loss: 0.1473 - val_c3_loss: 0.1174 - val_c4_loss: 0.1395 - val_c5_loss: 0.1287 - val_c1_accuracy: 0.9708 - val_c2_accuracy: 0.9626 - val_c3_accuracy: 0.9667 - val_c4_accuracy: 0.9671 - val_c5_accuracy: 0.9660\n",
            "Epoch 89/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.1747 - c1_loss: 0.2363 - c2_loss: 0.2408 - c3_loss: 0.2228 - c4_loss: 0.2406 - c5_loss: 0.2343 - c1_accuracy: 0.9188 - c2_accuracy: 0.9196 - c3_accuracy: 0.9215 - c4_accuracy: 0.9219 - c5_accuracy: 0.9218 - val_loss: 0.6840 - val_c1_loss: 0.1396 - val_c2_loss: 0.1442 - val_c3_loss: 0.1242 - val_c4_loss: 0.1430 - val_c5_loss: 0.1330 - val_c1_accuracy: 0.9647 - val_c2_accuracy: 0.9633 - val_c3_accuracy: 0.9702 - val_c4_accuracy: 0.9678 - val_c5_accuracy: 0.9605\n",
            "Epoch 90/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.0380 - c1_loss: 0.2075 - c2_loss: 0.1952 - c3_loss: 0.2113 - c4_loss: 0.2093 - c5_loss: 0.2147 - c1_accuracy: 0.9299 - c2_accuracy: 0.9352 - c3_accuracy: 0.9288 - c4_accuracy: 0.9293 - c5_accuracy: 0.9278 - val_loss: 0.6548 - val_c1_loss: 0.1279 - val_c2_loss: 0.1449 - val_c3_loss: 0.1206 - val_c4_loss: 0.1350 - val_c5_loss: 0.1264 - val_c1_accuracy: 0.9691 - val_c2_accuracy: 0.9667 - val_c3_accuracy: 0.9695 - val_c4_accuracy: 0.9691 - val_c5_accuracy: 0.9636\n",
            "Epoch 91/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.0662 - c1_loss: 0.2329 - c2_loss: 0.2100 - c3_loss: 0.2149 - c4_loss: 0.2032 - c5_loss: 0.2052 - c1_accuracy: 0.9243 - c2_accuracy: 0.9253 - c3_accuracy: 0.9278 - c4_accuracy: 0.9319 - c5_accuracy: 0.9300 - val_loss: 0.6946 - val_c1_loss: 0.1400 - val_c2_loss: 0.1519 - val_c3_loss: 0.1314 - val_c4_loss: 0.1437 - val_c5_loss: 0.1278 - val_c1_accuracy: 0.9667 - val_c2_accuracy: 0.9623 - val_c3_accuracy: 0.9674 - val_c4_accuracy: 0.9647 - val_c5_accuracy: 0.9657\n",
            "Epoch 92/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.1779 - c1_loss: 0.2401 - c2_loss: 0.2364 - c3_loss: 0.2240 - c4_loss: 0.2449 - c5_loss: 0.2325 - c1_accuracy: 0.9185 - c2_accuracy: 0.9240 - c3_accuracy: 0.9263 - c4_accuracy: 0.9203 - c5_accuracy: 0.9250 - val_loss: 0.6897 - val_c1_loss: 0.1282 - val_c2_loss: 0.1603 - val_c3_loss: 0.1222 - val_c4_loss: 0.1476 - val_c5_loss: 0.1314 - val_c1_accuracy: 0.9691 - val_c2_accuracy: 0.9609 - val_c3_accuracy: 0.9702 - val_c4_accuracy: 0.9657 - val_c5_accuracy: 0.9674\n",
            "Epoch 93/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.2551 - c1_loss: 0.2388 - c2_loss: 0.2620 - c3_loss: 0.2514 - c4_loss: 0.2569 - c5_loss: 0.2461 - c1_accuracy: 0.9210 - c2_accuracy: 0.9155 - c3_accuracy: 0.9124 - c4_accuracy: 0.9181 - c5_accuracy: 0.9184 - val_loss: 0.7044 - val_c1_loss: 0.1398 - val_c2_loss: 0.1598 - val_c3_loss: 0.1232 - val_c4_loss: 0.1478 - val_c5_loss: 0.1337 - val_c1_accuracy: 0.9674 - val_c2_accuracy: 0.9599 - val_c3_accuracy: 0.9702 - val_c4_accuracy: 0.9674 - val_c5_accuracy: 0.9602\n",
            "Epoch 94/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.0249 - c1_loss: 0.2022 - c2_loss: 0.2047 - c3_loss: 0.2116 - c4_loss: 0.2016 - c5_loss: 0.2047 - c1_accuracy: 0.9318 - c2_accuracy: 0.9293 - c3_accuracy: 0.9288 - c4_accuracy: 0.9319 - c5_accuracy: 0.9303 - val_loss: 0.6591 - val_c1_loss: 0.1252 - val_c2_loss: 0.1465 - val_c3_loss: 0.1150 - val_c4_loss: 0.1430 - val_c5_loss: 0.1295 - val_c1_accuracy: 0.9726 - val_c2_accuracy: 0.9633 - val_c3_accuracy: 0.9708 - val_c4_accuracy: 0.9681 - val_c5_accuracy: 0.9643\n",
            "Epoch 95/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9630 - c1_loss: 0.1962 - c2_loss: 0.1980 - c3_loss: 0.1839 - c4_loss: 0.1955 - c5_loss: 0.1895 - c1_accuracy: 0.9340 - c2_accuracy: 0.9356 - c3_accuracy: 0.9363 - c4_accuracy: 0.9331 - c5_accuracy: 0.9360 - val_loss: 0.6564 - val_c1_loss: 0.1237 - val_c2_loss: 0.1455 - val_c3_loss: 0.1206 - val_c4_loss: 0.1411 - val_c5_loss: 0.1255 - val_c1_accuracy: 0.9732 - val_c2_accuracy: 0.9681 - val_c3_accuracy: 0.9671 - val_c4_accuracy: 0.9660 - val_c5_accuracy: 0.9657\n",
            "Epoch 96/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9108 - c1_loss: 0.1742 - c2_loss: 0.1863 - c3_loss: 0.1810 - c4_loss: 0.1867 - c5_loss: 0.1826 - c1_accuracy: 0.9428 - c2_accuracy: 0.9375 - c3_accuracy: 0.9397 - c4_accuracy: 0.9390 - c5_accuracy: 0.9366 - val_loss: 0.6208 - val_c1_loss: 0.1191 - val_c2_loss: 0.1396 - val_c3_loss: 0.1109 - val_c4_loss: 0.1336 - val_c5_loss: 0.1175 - val_c1_accuracy: 0.9736 - val_c2_accuracy: 0.9671 - val_c3_accuracy: 0.9732 - val_c4_accuracy: 0.9708 - val_c5_accuracy: 0.9678\n",
            "Epoch 97/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9271 - c1_loss: 0.1952 - c2_loss: 0.1810 - c3_loss: 0.1781 - c4_loss: 0.1975 - c5_loss: 0.1753 - c1_accuracy: 0.9371 - c2_accuracy: 0.9410 - c3_accuracy: 0.9381 - c4_accuracy: 0.9350 - c5_accuracy: 0.9421 - val_loss: 0.6301 - val_c1_loss: 0.1189 - val_c2_loss: 0.1423 - val_c3_loss: 0.1097 - val_c4_loss: 0.1355 - val_c5_loss: 0.1238 - val_c1_accuracy: 0.9739 - val_c2_accuracy: 0.9650 - val_c3_accuracy: 0.9739 - val_c4_accuracy: 0.9691 - val_c5_accuracy: 0.9650\n",
            "Epoch 98/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9755 - c1_loss: 0.2135 - c2_loss: 0.1946 - c3_loss: 0.1865 - c4_loss: 0.1866 - c5_loss: 0.1942 - c1_accuracy: 0.9296 - c2_accuracy: 0.9299 - c3_accuracy: 0.9382 - c4_accuracy: 0.9415 - c5_accuracy: 0.9315 - val_loss: 0.6403 - val_c1_loss: 0.1256 - val_c2_loss: 0.1435 - val_c3_loss: 0.1105 - val_c4_loss: 0.1439 - val_c5_loss: 0.1168 - val_c1_accuracy: 0.9722 - val_c2_accuracy: 0.9650 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9664 - val_c5_accuracy: 0.9678\n",
            "Epoch 99/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9019 - c1_loss: 0.1754 - c2_loss: 0.1856 - c3_loss: 0.1643 - c4_loss: 0.1845 - c5_loss: 0.1921 - c1_accuracy: 0.9394 - c2_accuracy: 0.9403 - c3_accuracy: 0.9466 - c4_accuracy: 0.9365 - c5_accuracy: 0.9335 - val_loss: 0.6330 - val_c1_loss: 0.1197 - val_c2_loss: 0.1424 - val_c3_loss: 0.1117 - val_c4_loss: 0.1398 - val_c5_loss: 0.1194 - val_c1_accuracy: 0.9739 - val_c2_accuracy: 0.9702 - val_c3_accuracy: 0.9708 - val_c4_accuracy: 0.9702 - val_c5_accuracy: 0.9681\n",
            "Epoch 100/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.8514 - c1_loss: 0.1705 - c2_loss: 0.1686 - c3_loss: 0.1679 - c4_loss: 0.1771 - c5_loss: 0.1672 - c1_accuracy: 0.9419 - c2_accuracy: 0.9428 - c3_accuracy: 0.9434 - c4_accuracy: 0.9399 - c5_accuracy: 0.9434 - val_loss: 0.6060 - val_c1_loss: 0.1198 - val_c2_loss: 0.1318 - val_c3_loss: 0.1054 - val_c4_loss: 0.1309 - val_c5_loss: 0.1180 - val_c1_accuracy: 0.9746 - val_c2_accuracy: 0.9698 - val_c3_accuracy: 0.9732 - val_c4_accuracy: 0.9715 - val_c5_accuracy: 0.9698\n",
            "Epoch 101/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.8022 - c1_loss: 0.1416 - c2_loss: 0.1712 - c3_loss: 0.1699 - c4_loss: 0.1611 - c5_loss: 0.1583 - c1_accuracy: 0.9524 - c2_accuracy: 0.9409 - c3_accuracy: 0.9415 - c4_accuracy: 0.9444 - c5_accuracy: 0.9457 - val_loss: 0.6293 - val_c1_loss: 0.1203 - val_c2_loss: 0.1312 - val_c3_loss: 0.1181 - val_c4_loss: 0.1350 - val_c5_loss: 0.1248 - val_c1_accuracy: 0.9756 - val_c2_accuracy: 0.9702 - val_c3_accuracy: 0.9660 - val_c4_accuracy: 0.9698 - val_c5_accuracy: 0.9684\n",
            "Epoch 102/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.8639 - c1_loss: 0.1640 - c2_loss: 0.1622 - c3_loss: 0.1817 - c4_loss: 0.1783 - c5_loss: 0.1777 - c1_accuracy: 0.9419 - c2_accuracy: 0.9441 - c3_accuracy: 0.9413 - c4_accuracy: 0.9412 - c5_accuracy: 0.9394 - val_loss: 0.6157 - val_c1_loss: 0.1227 - val_c2_loss: 0.1367 - val_c3_loss: 0.1130 - val_c4_loss: 0.1255 - val_c5_loss: 0.1177 - val_c1_accuracy: 0.9736 - val_c2_accuracy: 0.9654 - val_c3_accuracy: 0.9722 - val_c4_accuracy: 0.9746 - val_c5_accuracy: 0.9695\n",
            "Epoch 103/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7808 - c1_loss: 0.1513 - c2_loss: 0.1702 - c3_loss: 0.1549 - c4_loss: 0.1472 - c5_loss: 0.1572 - c1_accuracy: 0.9499 - c2_accuracy: 0.9437 - c3_accuracy: 0.9500 - c4_accuracy: 0.9484 - c5_accuracy: 0.9468 - val_loss: 0.5984 - val_c1_loss: 0.1168 - val_c2_loss: 0.1309 - val_c3_loss: 0.1075 - val_c4_loss: 0.1302 - val_c5_loss: 0.1130 - val_c1_accuracy: 0.9770 - val_c2_accuracy: 0.9650 - val_c3_accuracy: 0.9726 - val_c4_accuracy: 0.9760 - val_c5_accuracy: 0.9719\n",
            "Epoch 104/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.9553 - c1_loss: 0.1898 - c2_loss: 0.2092 - c3_loss: 0.1895 - c4_loss: 0.1783 - c5_loss: 0.1887 - c1_accuracy: 0.9382 - c2_accuracy: 0.9299 - c3_accuracy: 0.9375 - c4_accuracy: 0.9357 - c5_accuracy: 0.9344 - val_loss: 0.6419 - val_c1_loss: 0.1270 - val_c2_loss: 0.1390 - val_c3_loss: 0.1166 - val_c4_loss: 0.1427 - val_c5_loss: 0.1166 - val_c1_accuracy: 0.9729 - val_c2_accuracy: 0.9667 - val_c3_accuracy: 0.9726 - val_c4_accuracy: 0.9671 - val_c5_accuracy: 0.9691\n",
            "Epoch 105/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 1.0528 - c1_loss: 0.2208 - c2_loss: 0.2140 - c3_loss: 0.2000 - c4_loss: 0.2088 - c5_loss: 0.2093 - c1_accuracy: 0.9293 - c2_accuracy: 0.9275 - c3_accuracy: 0.9341 - c4_accuracy: 0.9291 - c5_accuracy: 0.9340 - val_loss: 0.6775 - val_c1_loss: 0.1322 - val_c2_loss: 0.1572 - val_c3_loss: 0.1188 - val_c4_loss: 0.1480 - val_c5_loss: 0.1212 - val_c1_accuracy: 0.9702 - val_c2_accuracy: 0.9575 - val_c3_accuracy: 0.9708 - val_c4_accuracy: 0.9643 - val_c5_accuracy: 0.9691\n",
            "Epoch 106/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9169 - c1_loss: 0.1834 - c2_loss: 0.1862 - c3_loss: 0.1893 - c4_loss: 0.1773 - c5_loss: 0.1807 - c1_accuracy: 0.9384 - c2_accuracy: 0.9355 - c3_accuracy: 0.9350 - c4_accuracy: 0.9424 - c5_accuracy: 0.9357 - val_loss: 0.6101 - val_c1_loss: 0.1154 - val_c2_loss: 0.1367 - val_c3_loss: 0.1102 - val_c4_loss: 0.1319 - val_c5_loss: 0.1159 - val_c1_accuracy: 0.9770 - val_c2_accuracy: 0.9667 - val_c3_accuracy: 0.9729 - val_c4_accuracy: 0.9746 - val_c5_accuracy: 0.9708\n",
            "Epoch 107/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9745 - c1_loss: 0.1844 - c2_loss: 0.2076 - c3_loss: 0.2071 - c4_loss: 0.1905 - c5_loss: 0.1850 - c1_accuracy: 0.9343 - c2_accuracy: 0.9280 - c3_accuracy: 0.9318 - c4_accuracy: 0.9366 - c5_accuracy: 0.9375 - val_loss: 0.6266 - val_c1_loss: 0.1196 - val_c2_loss: 0.1324 - val_c3_loss: 0.1106 - val_c4_loss: 0.1380 - val_c5_loss: 0.1261 - val_c1_accuracy: 0.9746 - val_c2_accuracy: 0.9633 - val_c3_accuracy: 0.9712 - val_c4_accuracy: 0.9719 - val_c5_accuracy: 0.9657\n",
            "Epoch 108/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.9304 - c1_loss: 0.1846 - c2_loss: 0.1836 - c3_loss: 0.1856 - c4_loss: 0.1843 - c5_loss: 0.1922 - c1_accuracy: 0.9357 - c2_accuracy: 0.9391 - c3_accuracy: 0.9380 - c4_accuracy: 0.9369 - c5_accuracy: 0.9374 - val_loss: 0.6238 - val_c1_loss: 0.1231 - val_c2_loss: 0.1314 - val_c3_loss: 0.1082 - val_c4_loss: 0.1446 - val_c5_loss: 0.1165 - val_c1_accuracy: 0.9756 - val_c2_accuracy: 0.9691 - val_c3_accuracy: 0.9736 - val_c4_accuracy: 0.9674 - val_c5_accuracy: 0.9712\n",
            "Epoch 109/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.8335 - c1_loss: 0.1623 - c2_loss: 0.1709 - c3_loss: 0.1640 - c4_loss: 0.1627 - c5_loss: 0.1737 - c1_accuracy: 0.9456 - c2_accuracy: 0.9422 - c3_accuracy: 0.9457 - c4_accuracy: 0.9435 - c5_accuracy: 0.9422 - val_loss: 0.5912 - val_c1_loss: 0.1096 - val_c2_loss: 0.1355 - val_c3_loss: 0.1017 - val_c4_loss: 0.1338 - val_c5_loss: 0.1104 - val_c1_accuracy: 0.9791 - val_c2_accuracy: 0.9640 - val_c3_accuracy: 0.9777 - val_c4_accuracy: 0.9708 - val_c5_accuracy: 0.9736\n",
            "Epoch 110/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.8104 - c1_loss: 0.1589 - c2_loss: 0.1657 - c3_loss: 0.1632 - c4_loss: 0.1662 - c5_loss: 0.1563 - c1_accuracy: 0.9469 - c2_accuracy: 0.9432 - c3_accuracy: 0.9440 - c4_accuracy: 0.9419 - c5_accuracy: 0.9424 - val_loss: 0.6280 - val_c1_loss: 0.1204 - val_c2_loss: 0.1347 - val_c3_loss: 0.1115 - val_c4_loss: 0.1430 - val_c5_loss: 0.1184 - val_c1_accuracy: 0.9774 - val_c2_accuracy: 0.9667 - val_c3_accuracy: 0.9726 - val_c4_accuracy: 0.9715 - val_c5_accuracy: 0.9691\n",
            "Epoch 111/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7983 - c1_loss: 0.1627 - c2_loss: 0.1784 - c3_loss: 0.1449 - c4_loss: 0.1576 - c5_loss: 0.1546 - c1_accuracy: 0.9450 - c2_accuracy: 0.9407 - c3_accuracy: 0.9488 - c4_accuracy: 0.9485 - c5_accuracy: 0.9512 - val_loss: 0.6034 - val_c1_loss: 0.1183 - val_c2_loss: 0.1339 - val_c3_loss: 0.1024 - val_c4_loss: 0.1343 - val_c5_loss: 0.1145 - val_c1_accuracy: 0.9767 - val_c2_accuracy: 0.9643 - val_c3_accuracy: 0.9746 - val_c4_accuracy: 0.9695 - val_c5_accuracy: 0.9712\n",
            "Epoch 112/500\n",
            "69/69 [==============================] - 2s 32ms/step - loss: 0.8077 - c1_loss: 0.1477 - c2_loss: 0.1442 - c3_loss: 0.1686 - c4_loss: 0.1791 - c5_loss: 0.1681 - c1_accuracy: 0.9521 - c2_accuracy: 0.9515 - c3_accuracy: 0.9457 - c4_accuracy: 0.9410 - c5_accuracy: 0.9421 - val_loss: 0.5876 - val_c1_loss: 0.1109 - val_c2_loss: 0.1351 - val_c3_loss: 0.1008 - val_c4_loss: 0.1287 - val_c5_loss: 0.1121 - val_c1_accuracy: 0.9760 - val_c2_accuracy: 0.9688 - val_c3_accuracy: 0.9739 - val_c4_accuracy: 0.9750 - val_c5_accuracy: 0.9715\n",
            "Epoch 113/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.8271 - c1_loss: 0.1558 - c2_loss: 0.1876 - c3_loss: 0.1580 - c4_loss: 0.1617 - c5_loss: 0.1640 - c1_accuracy: 0.9457 - c2_accuracy: 0.9360 - c3_accuracy: 0.9474 - c4_accuracy: 0.9441 - c5_accuracy: 0.9444 - val_loss: 0.6320 - val_c1_loss: 0.1167 - val_c2_loss: 0.1455 - val_c3_loss: 0.1024 - val_c4_loss: 0.1387 - val_c5_loss: 0.1287 - val_c1_accuracy: 0.9746 - val_c2_accuracy: 0.9667 - val_c3_accuracy: 0.9760 - val_c4_accuracy: 0.9715 - val_c5_accuracy: 0.9654\n",
            "Epoch 114/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7323 - c1_loss: 0.1488 - c2_loss: 0.1518 - c3_loss: 0.1348 - c4_loss: 0.1372 - c5_loss: 0.1597 - c1_accuracy: 0.9490 - c2_accuracy: 0.9479 - c3_accuracy: 0.9546 - c4_accuracy: 0.9507 - c5_accuracy: 0.9446 - val_loss: 0.6343 - val_c1_loss: 0.1206 - val_c2_loss: 0.1385 - val_c3_loss: 0.1023 - val_c4_loss: 0.1408 - val_c5_loss: 0.1321 - val_c1_accuracy: 0.9729 - val_c2_accuracy: 0.9667 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9715 - val_c5_accuracy: 0.9650\n",
            "Epoch 115/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7530 - c1_loss: 0.1477 - c2_loss: 0.1465 - c3_loss: 0.1510 - c4_loss: 0.1566 - c5_loss: 0.1512 - c1_accuracy: 0.9487 - c2_accuracy: 0.9497 - c3_accuracy: 0.9484 - c4_accuracy: 0.9481 - c5_accuracy: 0.9482 - val_loss: 0.6117 - val_c1_loss: 0.1222 - val_c2_loss: 0.1341 - val_c3_loss: 0.1102 - val_c4_loss: 0.1310 - val_c5_loss: 0.1141 - val_c1_accuracy: 0.9756 - val_c2_accuracy: 0.9674 - val_c3_accuracy: 0.9726 - val_c4_accuracy: 0.9767 - val_c5_accuracy: 0.9695\n",
            "Epoch 116/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7839 - c1_loss: 0.1454 - c2_loss: 0.1590 - c3_loss: 0.1623 - c4_loss: 0.1709 - c5_loss: 0.1462 - c1_accuracy: 0.9504 - c2_accuracy: 0.9485 - c3_accuracy: 0.9474 - c4_accuracy: 0.9452 - c5_accuracy: 0.9496 - val_loss: 0.6115 - val_c1_loss: 0.1231 - val_c2_loss: 0.1390 - val_c3_loss: 0.1079 - val_c4_loss: 0.1319 - val_c5_loss: 0.1096 - val_c1_accuracy: 0.9756 - val_c2_accuracy: 0.9654 - val_c3_accuracy: 0.9736 - val_c4_accuracy: 0.9708 - val_c5_accuracy: 0.9691\n",
            "Epoch 117/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7505 - c1_loss: 0.1476 - c2_loss: 0.1482 - c3_loss: 0.1470 - c4_loss: 0.1550 - c5_loss: 0.1527 - c1_accuracy: 0.9497 - c2_accuracy: 0.9479 - c3_accuracy: 0.9529 - c4_accuracy: 0.9469 - c5_accuracy: 0.9466 - val_loss: 0.6036 - val_c1_loss: 0.1199 - val_c2_loss: 0.1367 - val_c3_loss: 0.1032 - val_c4_loss: 0.1345 - val_c5_loss: 0.1092 - val_c1_accuracy: 0.9746 - val_c2_accuracy: 0.9691 - val_c3_accuracy: 0.9750 - val_c4_accuracy: 0.9708 - val_c5_accuracy: 0.9688\n",
            "Epoch 118/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7414 - c1_loss: 0.1465 - c2_loss: 0.1485 - c3_loss: 0.1445 - c4_loss: 0.1486 - c5_loss: 0.1532 - c1_accuracy: 0.9497 - c2_accuracy: 0.9500 - c3_accuracy: 0.9521 - c4_accuracy: 0.9535 - c5_accuracy: 0.9497 - val_loss: 0.6005 - val_c1_loss: 0.1111 - val_c2_loss: 0.1374 - val_c3_loss: 0.1011 - val_c4_loss: 0.1334 - val_c5_loss: 0.1175 - val_c1_accuracy: 0.9774 - val_c2_accuracy: 0.9695 - val_c3_accuracy: 0.9760 - val_c4_accuracy: 0.9719 - val_c5_accuracy: 0.9674\n",
            "Epoch 119/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7302 - c1_loss: 0.1387 - c2_loss: 0.1451 - c3_loss: 0.1502 - c4_loss: 0.1497 - c5_loss: 0.1465 - c1_accuracy: 0.9529 - c2_accuracy: 0.9500 - c3_accuracy: 0.9477 - c4_accuracy: 0.9478 - c5_accuracy: 0.9531 - val_loss: 0.6013 - val_c1_loss: 0.1187 - val_c2_loss: 0.1349 - val_c3_loss: 0.1050 - val_c4_loss: 0.1241 - val_c5_loss: 0.1187 - val_c1_accuracy: 0.9777 - val_c2_accuracy: 0.9681 - val_c3_accuracy: 0.9746 - val_c4_accuracy: 0.9760 - val_c5_accuracy: 0.9684\n",
            "Epoch 120/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.8390 - c1_loss: 0.1639 - c2_loss: 0.1921 - c3_loss: 0.1626 - c4_loss: 0.1594 - c5_loss: 0.1610 - c1_accuracy: 0.9456 - c2_accuracy: 0.9403 - c3_accuracy: 0.9457 - c4_accuracy: 0.9446 - c5_accuracy: 0.9452 - val_loss: 0.6495 - val_c1_loss: 0.1247 - val_c2_loss: 0.1490 - val_c3_loss: 0.1219 - val_c4_loss: 0.1321 - val_c5_loss: 0.1219 - val_c1_accuracy: 0.9715 - val_c2_accuracy: 0.9654 - val_c3_accuracy: 0.9688 - val_c4_accuracy: 0.9702 - val_c5_accuracy: 0.9684\n",
            "Epoch 121/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7889 - c1_loss: 0.1456 - c2_loss: 0.1740 - c3_loss: 0.1591 - c4_loss: 0.1494 - c5_loss: 0.1608 - c1_accuracy: 0.9497 - c2_accuracy: 0.9419 - c3_accuracy: 0.9475 - c4_accuracy: 0.9518 - c5_accuracy: 0.9456 - val_loss: 0.5865 - val_c1_loss: 0.1111 - val_c2_loss: 0.1386 - val_c3_loss: 0.0992 - val_c4_loss: 0.1326 - val_c5_loss: 0.1050 - val_c1_accuracy: 0.9753 - val_c2_accuracy: 0.9657 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9719 - val_c5_accuracy: 0.9750\n",
            "Epoch 122/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7570 - c1_loss: 0.1525 - c2_loss: 0.1616 - c3_loss: 0.1455 - c4_loss: 0.1512 - c5_loss: 0.1462 - c1_accuracy: 0.9479 - c2_accuracy: 0.9457 - c3_accuracy: 0.9488 - c4_accuracy: 0.9462 - c5_accuracy: 0.9518 - val_loss: 0.5915 - val_c1_loss: 0.1138 - val_c2_loss: 0.1320 - val_c3_loss: 0.0989 - val_c4_loss: 0.1317 - val_c5_loss: 0.1151 - val_c1_accuracy: 0.9760 - val_c2_accuracy: 0.9684 - val_c3_accuracy: 0.9750 - val_c4_accuracy: 0.9719 - val_c5_accuracy: 0.9719\n",
            "Epoch 123/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7705 - c1_loss: 0.1536 - c2_loss: 0.1547 - c3_loss: 0.1562 - c4_loss: 0.1598 - c5_loss: 0.1463 - c1_accuracy: 0.9466 - c2_accuracy: 0.9494 - c3_accuracy: 0.9496 - c4_accuracy: 0.9463 - c5_accuracy: 0.9500 - val_loss: 0.6090 - val_c1_loss: 0.1138 - val_c2_loss: 0.1399 - val_c3_loss: 0.1026 - val_c4_loss: 0.1321 - val_c5_loss: 0.1206 - val_c1_accuracy: 0.9770 - val_c2_accuracy: 0.9674 - val_c3_accuracy: 0.9739 - val_c4_accuracy: 0.9722 - val_c5_accuracy: 0.9729\n",
            "Epoch 124/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9020 - c1_loss: 0.1734 - c2_loss: 0.1678 - c3_loss: 0.1936 - c4_loss: 0.1867 - c5_loss: 0.1805 - c1_accuracy: 0.9416 - c2_accuracy: 0.9421 - c3_accuracy: 0.9396 - c4_accuracy: 0.9400 - c5_accuracy: 0.9394 - val_loss: 0.6084 - val_c1_loss: 0.1182 - val_c2_loss: 0.1364 - val_c3_loss: 0.1050 - val_c4_loss: 0.1378 - val_c5_loss: 0.1110 - val_c1_accuracy: 0.9753 - val_c2_accuracy: 0.9688 - val_c3_accuracy: 0.9702 - val_c4_accuracy: 0.9712 - val_c5_accuracy: 0.9698\n",
            "Epoch 125/500\n",
            "69/69 [==============================] - 2s 32ms/step - loss: 0.7953 - c1_loss: 0.1534 - c2_loss: 0.1524 - c3_loss: 0.1524 - c4_loss: 0.1685 - c5_loss: 0.1685 - c1_accuracy: 0.9468 - c2_accuracy: 0.9485 - c3_accuracy: 0.9510 - c4_accuracy: 0.9454 - c5_accuracy: 0.9424 - val_loss: 0.5959 - val_c1_loss: 0.1141 - val_c2_loss: 0.1374 - val_c3_loss: 0.1073 - val_c4_loss: 0.1282 - val_c5_loss: 0.1090 - val_c1_accuracy: 0.9763 - val_c2_accuracy: 0.9671 - val_c3_accuracy: 0.9719 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9719\n",
            "Epoch 126/500\n",
            "69/69 [==============================] - 2s 32ms/step - loss: 0.9591 - c1_loss: 0.1858 - c2_loss: 0.1703 - c3_loss: 0.2011 - c4_loss: 0.1933 - c5_loss: 0.2086 - c1_accuracy: 0.9397 - c2_accuracy: 0.9434 - c3_accuracy: 0.9305 - c4_accuracy: 0.9371 - c5_accuracy: 0.9325 - val_loss: 0.6183 - val_c1_loss: 0.1123 - val_c2_loss: 0.1377 - val_c3_loss: 0.1107 - val_c4_loss: 0.1349 - val_c5_loss: 0.1226 - val_c1_accuracy: 0.9753 - val_c2_accuracy: 0.9681 - val_c3_accuracy: 0.9705 - val_c4_accuracy: 0.9705 - val_c5_accuracy: 0.9688\n",
            "Epoch 127/500\n",
            "69/69 [==============================] - 2s 32ms/step - loss: 0.8726 - c1_loss: 0.1730 - c2_loss: 0.1661 - c3_loss: 0.1652 - c4_loss: 0.1797 - c5_loss: 0.1888 - c1_accuracy: 0.9432 - c2_accuracy: 0.9444 - c3_accuracy: 0.9484 - c4_accuracy: 0.9380 - c5_accuracy: 0.9362 - val_loss: 0.6160 - val_c1_loss: 0.1210 - val_c2_loss: 0.1356 - val_c3_loss: 0.1023 - val_c4_loss: 0.1368 - val_c5_loss: 0.1203 - val_c1_accuracy: 0.9726 - val_c2_accuracy: 0.9678 - val_c3_accuracy: 0.9746 - val_c4_accuracy: 0.9695 - val_c5_accuracy: 0.9702\n",
            "Epoch 128/500\n",
            "69/69 [==============================] - 2s 32ms/step - loss: 0.7437 - c1_loss: 0.1495 - c2_loss: 0.1331 - c3_loss: 0.1563 - c4_loss: 0.1605 - c5_loss: 0.1443 - c1_accuracy: 0.9493 - c2_accuracy: 0.9556 - c3_accuracy: 0.9496 - c4_accuracy: 0.9482 - c5_accuracy: 0.9510 - val_loss: 0.5946 - val_c1_loss: 0.1103 - val_c2_loss: 0.1262 - val_c3_loss: 0.1022 - val_c4_loss: 0.1319 - val_c5_loss: 0.1239 - val_c1_accuracy: 0.9760 - val_c2_accuracy: 0.9712 - val_c3_accuracy: 0.9753 - val_c4_accuracy: 0.9695 - val_c5_accuracy: 0.9681\n",
            "Epoch 129/500\n",
            "69/69 [==============================] - 2s 32ms/step - loss: 0.8434 - c1_loss: 0.1746 - c2_loss: 0.1672 - c3_loss: 0.1530 - c4_loss: 0.1707 - c5_loss: 0.1779 - c1_accuracy: 0.9391 - c2_accuracy: 0.9409 - c3_accuracy: 0.9441 - c4_accuracy: 0.9419 - c5_accuracy: 0.9425 - val_loss: 0.6721 - val_c1_loss: 0.1409 - val_c2_loss: 0.1376 - val_c3_loss: 0.1223 - val_c4_loss: 0.1452 - val_c5_loss: 0.1261 - val_c1_accuracy: 0.9681 - val_c2_accuracy: 0.9664 - val_c3_accuracy: 0.9654 - val_c4_accuracy: 0.9681 - val_c5_accuracy: 0.9688\n",
            "Epoch 130/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.0333 - c1_loss: 0.2247 - c2_loss: 0.2094 - c3_loss: 0.2059 - c4_loss: 0.2086 - c5_loss: 0.1847 - c1_accuracy: 0.9269 - c2_accuracy: 0.9285 - c3_accuracy: 0.9297 - c4_accuracy: 0.9281 - c5_accuracy: 0.9363 - val_loss: 0.6352 - val_c1_loss: 0.1260 - val_c2_loss: 0.1325 - val_c3_loss: 0.1119 - val_c4_loss: 0.1400 - val_c5_loss: 0.1247 - val_c1_accuracy: 0.9698 - val_c2_accuracy: 0.9688 - val_c3_accuracy: 0.9678 - val_c4_accuracy: 0.9698 - val_c5_accuracy: 0.9674\n",
            "Epoch 131/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.8403 - c1_loss: 0.1712 - c2_loss: 0.1667 - c3_loss: 0.1693 - c4_loss: 0.1771 - c5_loss: 0.1559 - c1_accuracy: 0.9440 - c2_accuracy: 0.9409 - c3_accuracy: 0.9399 - c4_accuracy: 0.9432 - c5_accuracy: 0.9449 - val_loss: 0.6454 - val_c1_loss: 0.1252 - val_c2_loss: 0.1387 - val_c3_loss: 0.1158 - val_c4_loss: 0.1415 - val_c5_loss: 0.1242 - val_c1_accuracy: 0.9708 - val_c2_accuracy: 0.9643 - val_c3_accuracy: 0.9684 - val_c4_accuracy: 0.9660 - val_c5_accuracy: 0.9650\n",
            "Epoch 132/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 1.0231 - c1_loss: 0.1861 - c2_loss: 0.2162 - c3_loss: 0.2084 - c4_loss: 0.2079 - c5_loss: 0.2046 - c1_accuracy: 0.9384 - c2_accuracy: 0.9290 - c3_accuracy: 0.9275 - c4_accuracy: 0.9310 - c5_accuracy: 0.9302 - val_loss: 0.6340 - val_c1_loss: 0.1254 - val_c2_loss: 0.1258 - val_c3_loss: 0.1124 - val_c4_loss: 0.1414 - val_c5_loss: 0.1291 - val_c1_accuracy: 0.9712 - val_c2_accuracy: 0.9705 - val_c3_accuracy: 0.9705 - val_c4_accuracy: 0.9691 - val_c5_accuracy: 0.9678\n",
            "Epoch 133/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9171 - c1_loss: 0.1764 - c2_loss: 0.1953 - c3_loss: 0.1805 - c4_loss: 0.1852 - c5_loss: 0.1798 - c1_accuracy: 0.9403 - c2_accuracy: 0.9353 - c3_accuracy: 0.9422 - c4_accuracy: 0.9418 - c5_accuracy: 0.9384 - val_loss: 0.6326 - val_c1_loss: 0.1272 - val_c2_loss: 0.1267 - val_c3_loss: 0.1103 - val_c4_loss: 0.1450 - val_c5_loss: 0.1233 - val_c1_accuracy: 0.9691 - val_c2_accuracy: 0.9695 - val_c3_accuracy: 0.9729 - val_c4_accuracy: 0.9671 - val_c5_accuracy: 0.9698\n",
            "Epoch 134/500\n",
            "69/69 [==============================] - 2s 32ms/step - loss: 0.7895 - c1_loss: 0.1467 - c2_loss: 0.1689 - c3_loss: 0.1659 - c4_loss: 0.1466 - c5_loss: 0.1615 - c1_accuracy: 0.9500 - c2_accuracy: 0.9434 - c3_accuracy: 0.9475 - c4_accuracy: 0.9499 - c5_accuracy: 0.9468 - val_loss: 0.5795 - val_c1_loss: 0.1090 - val_c2_loss: 0.1215 - val_c3_loss: 0.0980 - val_c4_loss: 0.1365 - val_c5_loss: 0.1146 - val_c1_accuracy: 0.9753 - val_c2_accuracy: 0.9708 - val_c3_accuracy: 0.9760 - val_c4_accuracy: 0.9729 - val_c5_accuracy: 0.9705\n",
            "Epoch 135/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7434 - c1_loss: 0.1480 - c2_loss: 0.1583 - c3_loss: 0.1423 - c4_loss: 0.1436 - c5_loss: 0.1511 - c1_accuracy: 0.9497 - c2_accuracy: 0.9469 - c3_accuracy: 0.9502 - c4_accuracy: 0.9532 - c5_accuracy: 0.9488 - val_loss: 0.5812 - val_c1_loss: 0.1091 - val_c2_loss: 0.1243 - val_c3_loss: 0.0965 - val_c4_loss: 0.1343 - val_c5_loss: 0.1170 - val_c1_accuracy: 0.9763 - val_c2_accuracy: 0.9691 - val_c3_accuracy: 0.9719 - val_c4_accuracy: 0.9715 - val_c5_accuracy: 0.9678\n",
            "Epoch 136/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7457 - c1_loss: 0.1488 - c2_loss: 0.1406 - c3_loss: 0.1484 - c4_loss: 0.1571 - c5_loss: 0.1508 - c1_accuracy: 0.9524 - c2_accuracy: 0.9534 - c3_accuracy: 0.9494 - c4_accuracy: 0.9504 - c5_accuracy: 0.9510 - val_loss: 0.5705 - val_c1_loss: 0.1066 - val_c2_loss: 0.1242 - val_c3_loss: 0.0971 - val_c4_loss: 0.1301 - val_c5_loss: 0.1124 - val_c1_accuracy: 0.9763 - val_c2_accuracy: 0.9712 - val_c3_accuracy: 0.9753 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9702\n",
            "Epoch 137/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.8355 - c1_loss: 0.1500 - c2_loss: 0.1728 - c3_loss: 0.1610 - c4_loss: 0.1585 - c5_loss: 0.1932 - c1_accuracy: 0.9465 - c2_accuracy: 0.9403 - c3_accuracy: 0.9474 - c4_accuracy: 0.9491 - c5_accuracy: 0.9362 - val_loss: 0.6213 - val_c1_loss: 0.1186 - val_c2_loss: 0.1296 - val_c3_loss: 0.1028 - val_c4_loss: 0.1409 - val_c5_loss: 0.1294 - val_c1_accuracy: 0.9767 - val_c2_accuracy: 0.9681 - val_c3_accuracy: 0.9739 - val_c4_accuracy: 0.9722 - val_c5_accuracy: 0.9650\n",
            "Epoch 138/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7283 - c1_loss: 0.1441 - c2_loss: 0.1504 - c3_loss: 0.1431 - c4_loss: 0.1567 - c5_loss: 0.1339 - c1_accuracy: 0.9521 - c2_accuracy: 0.9475 - c3_accuracy: 0.9485 - c4_accuracy: 0.9479 - c5_accuracy: 0.9531 - val_loss: 0.5921 - val_c1_loss: 0.1107 - val_c2_loss: 0.1230 - val_c3_loss: 0.1017 - val_c4_loss: 0.1321 - val_c5_loss: 0.1247 - val_c1_accuracy: 0.9756 - val_c2_accuracy: 0.9719 - val_c3_accuracy: 0.9739 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9695\n",
            "Epoch 139/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9785 - c1_loss: 0.2001 - c2_loss: 0.1891 - c3_loss: 0.2022 - c4_loss: 0.1869 - c5_loss: 0.2003 - c1_accuracy: 0.9330 - c2_accuracy: 0.9372 - c3_accuracy: 0.9365 - c4_accuracy: 0.9378 - c5_accuracy: 0.9290 - val_loss: 0.6259 - val_c1_loss: 0.1222 - val_c2_loss: 0.1302 - val_c3_loss: 0.1160 - val_c4_loss: 0.1428 - val_c5_loss: 0.1147 - val_c1_accuracy: 0.9736 - val_c2_accuracy: 0.9715 - val_c3_accuracy: 0.9729 - val_c4_accuracy: 0.9712 - val_c5_accuracy: 0.9715\n",
            "Epoch 140/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7791 - c1_loss: 0.1667 - c2_loss: 0.1473 - c3_loss: 0.1705 - c4_loss: 0.1517 - c5_loss: 0.1429 - c1_accuracy: 0.9457 - c2_accuracy: 0.9504 - c3_accuracy: 0.9447 - c4_accuracy: 0.9502 - c5_accuracy: 0.9515 - val_loss: 0.5999 - val_c1_loss: 0.1164 - val_c2_loss: 0.1264 - val_c3_loss: 0.0994 - val_c4_loss: 0.1291 - val_c5_loss: 0.1287 - val_c1_accuracy: 0.9732 - val_c2_accuracy: 0.9722 - val_c3_accuracy: 0.9719 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9667\n",
            "Epoch 141/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.8648 - c1_loss: 0.1858 - c2_loss: 0.1722 - c3_loss: 0.1656 - c4_loss: 0.1568 - c5_loss: 0.1844 - c1_accuracy: 0.9429 - c2_accuracy: 0.9422 - c3_accuracy: 0.9454 - c4_accuracy: 0.9460 - c5_accuracy: 0.9338 - val_loss: 0.5998 - val_c1_loss: 0.1131 - val_c2_loss: 0.1254 - val_c3_loss: 0.1029 - val_c4_loss: 0.1423 - val_c5_loss: 0.1160 - val_c1_accuracy: 0.9743 - val_c2_accuracy: 0.9698 - val_c3_accuracy: 0.9753 - val_c4_accuracy: 0.9702 - val_c5_accuracy: 0.9698\n",
            "Epoch 142/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7317 - c1_loss: 0.1504 - c2_loss: 0.1483 - c3_loss: 0.1494 - c4_loss: 0.1416 - c5_loss: 0.1421 - c1_accuracy: 0.9502 - c2_accuracy: 0.9509 - c3_accuracy: 0.9474 - c4_accuracy: 0.9538 - c5_accuracy: 0.9515 - val_loss: 0.5921 - val_c1_loss: 0.1167 - val_c2_loss: 0.1199 - val_c3_loss: 0.1020 - val_c4_loss: 0.1380 - val_c5_loss: 0.1154 - val_c1_accuracy: 0.9743 - val_c2_accuracy: 0.9722 - val_c3_accuracy: 0.9729 - val_c4_accuracy: 0.9708 - val_c5_accuracy: 0.9705\n",
            "Epoch 143/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.8768 - c1_loss: 0.1918 - c2_loss: 0.1787 - c3_loss: 0.1679 - c4_loss: 0.1594 - c5_loss: 0.1790 - c1_accuracy: 0.9371 - c2_accuracy: 0.9397 - c3_accuracy: 0.9413 - c4_accuracy: 0.9479 - c5_accuracy: 0.9391 - val_loss: 0.6258 - val_c1_loss: 0.1199 - val_c2_loss: 0.1340 - val_c3_loss: 0.1061 - val_c4_loss: 0.1453 - val_c5_loss: 0.1205 - val_c1_accuracy: 0.9753 - val_c2_accuracy: 0.9691 - val_c3_accuracy: 0.9732 - val_c4_accuracy: 0.9708 - val_c5_accuracy: 0.9650\n",
            "Epoch 144/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.8655 - c1_loss: 0.1819 - c2_loss: 0.1818 - c3_loss: 0.1714 - c4_loss: 0.1624 - c5_loss: 0.1680 - c1_accuracy: 0.9427 - c2_accuracy: 0.9407 - c3_accuracy: 0.9421 - c4_accuracy: 0.9453 - c5_accuracy: 0.9434 - val_loss: 0.6008 - val_c1_loss: 0.1188 - val_c2_loss: 0.1307 - val_c3_loss: 0.1040 - val_c4_loss: 0.1411 - val_c5_loss: 0.1062 - val_c1_accuracy: 0.9743 - val_c2_accuracy: 0.9698 - val_c3_accuracy: 0.9753 - val_c4_accuracy: 0.9705 - val_c5_accuracy: 0.9732\n",
            "Epoch 145/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7490 - c1_loss: 0.1571 - c2_loss: 0.1549 - c3_loss: 0.1573 - c4_loss: 0.1378 - c5_loss: 0.1420 - c1_accuracy: 0.9482 - c2_accuracy: 0.9503 - c3_accuracy: 0.9485 - c4_accuracy: 0.9529 - c5_accuracy: 0.9525 - val_loss: 0.5711 - val_c1_loss: 0.1051 - val_c2_loss: 0.1246 - val_c3_loss: 0.0961 - val_c4_loss: 0.1381 - val_c5_loss: 0.1073 - val_c1_accuracy: 0.9787 - val_c2_accuracy: 0.9712 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9722 - val_c5_accuracy: 0.9722\n",
            "Epoch 146/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7346 - c1_loss: 0.1296 - c2_loss: 0.1524 - c3_loss: 0.1535 - c4_loss: 0.1401 - c5_loss: 0.1591 - c1_accuracy: 0.9550 - c2_accuracy: 0.9481 - c3_accuracy: 0.9496 - c4_accuracy: 0.9493 - c5_accuracy: 0.9456 - val_loss: 0.5649 - val_c1_loss: 0.1064 - val_c2_loss: 0.1220 - val_c3_loss: 0.1002 - val_c4_loss: 0.1331 - val_c5_loss: 0.1032 - val_c1_accuracy: 0.9780 - val_c2_accuracy: 0.9743 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9729 - val_c5_accuracy: 0.9719\n",
            "Epoch 147/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7119 - c1_loss: 0.1431 - c2_loss: 0.1418 - c3_loss: 0.1390 - c4_loss: 0.1445 - c5_loss: 0.1434 - c1_accuracy: 0.9535 - c2_accuracy: 0.9525 - c3_accuracy: 0.9524 - c4_accuracy: 0.9494 - c5_accuracy: 0.9518 - val_loss: 0.5836 - val_c1_loss: 0.1146 - val_c2_loss: 0.1218 - val_c3_loss: 0.1031 - val_c4_loss: 0.1320 - val_c5_loss: 0.1121 - val_c1_accuracy: 0.9774 - val_c2_accuracy: 0.9736 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9736 - val_c5_accuracy: 0.9688\n",
            "Epoch 148/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6831 - c1_loss: 0.1375 - c2_loss: 0.1313 - c3_loss: 0.1368 - c4_loss: 0.1400 - c5_loss: 0.1375 - c1_accuracy: 0.9534 - c2_accuracy: 0.9581 - c3_accuracy: 0.9537 - c4_accuracy: 0.9543 - c5_accuracy: 0.9537 - val_loss: 0.5656 - val_c1_loss: 0.1107 - val_c2_loss: 0.1191 - val_c3_loss: 0.1017 - val_c4_loss: 0.1326 - val_c5_loss: 0.1015 - val_c1_accuracy: 0.9808 - val_c2_accuracy: 0.9719 - val_c3_accuracy: 0.9732 - val_c4_accuracy: 0.9756 - val_c5_accuracy: 0.9726\n",
            "Epoch 149/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6960 - c1_loss: 0.1357 - c2_loss: 0.1409 - c3_loss: 0.1448 - c4_loss: 0.1434 - c5_loss: 0.1312 - c1_accuracy: 0.9524 - c2_accuracy: 0.9510 - c3_accuracy: 0.9478 - c4_accuracy: 0.9502 - c5_accuracy: 0.9571 - val_loss: 0.5661 - val_c1_loss: 0.1087 - val_c2_loss: 0.1207 - val_c3_loss: 0.1029 - val_c4_loss: 0.1362 - val_c5_loss: 0.0975 - val_c1_accuracy: 0.9794 - val_c2_accuracy: 0.9726 - val_c3_accuracy: 0.9774 - val_c4_accuracy: 0.9736 - val_c5_accuracy: 0.9774\n",
            "Epoch 150/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6819 - c1_loss: 0.1244 - c2_loss: 0.1477 - c3_loss: 0.1371 - c4_loss: 0.1363 - c5_loss: 0.1364 - c1_accuracy: 0.9587 - c2_accuracy: 0.9518 - c3_accuracy: 0.9552 - c4_accuracy: 0.9571 - c5_accuracy: 0.9547 - val_loss: 0.5972 - val_c1_loss: 0.1194 - val_c2_loss: 0.1311 - val_c3_loss: 0.1089 - val_c4_loss: 0.1340 - val_c5_loss: 0.1039 - val_c1_accuracy: 0.9753 - val_c2_accuracy: 0.9708 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9726\n",
            "Epoch 151/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.8531 - c1_loss: 0.1700 - c2_loss: 0.1650 - c3_loss: 0.1624 - c4_loss: 0.1989 - c5_loss: 0.1568 - c1_accuracy: 0.9427 - c2_accuracy: 0.9465 - c3_accuracy: 0.9450 - c4_accuracy: 0.9384 - c5_accuracy: 0.9479 - val_loss: 0.6086 - val_c1_loss: 0.1153 - val_c2_loss: 0.1295 - val_c3_loss: 0.1084 - val_c4_loss: 0.1413 - val_c5_loss: 0.1141 - val_c1_accuracy: 0.9750 - val_c2_accuracy: 0.9719 - val_c3_accuracy: 0.9736 - val_c4_accuracy: 0.9722 - val_c5_accuracy: 0.9698\n",
            "Epoch 152/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7658 - c1_loss: 0.1461 - c2_loss: 0.1669 - c3_loss: 0.1515 - c4_loss: 0.1560 - c5_loss: 0.1453 - c1_accuracy: 0.9538 - c2_accuracy: 0.9444 - c3_accuracy: 0.9509 - c4_accuracy: 0.9493 - c5_accuracy: 0.9544 - val_loss: 0.6047 - val_c1_loss: 0.1157 - val_c2_loss: 0.1334 - val_c3_loss: 0.1038 - val_c4_loss: 0.1446 - val_c5_loss: 0.1073 - val_c1_accuracy: 0.9756 - val_c2_accuracy: 0.9695 - val_c3_accuracy: 0.9770 - val_c4_accuracy: 0.9719 - val_c5_accuracy: 0.9722\n",
            "Epoch 153/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9222 - c1_loss: 0.1951 - c2_loss: 0.1885 - c3_loss: 0.1872 - c4_loss: 0.1728 - c5_loss: 0.1786 - c1_accuracy: 0.9340 - c2_accuracy: 0.9393 - c3_accuracy: 0.9368 - c4_accuracy: 0.9443 - c5_accuracy: 0.9402 - val_loss: 0.6217 - val_c1_loss: 0.1355 - val_c2_loss: 0.1388 - val_c3_loss: 0.1028 - val_c4_loss: 0.1393 - val_c5_loss: 0.1054 - val_c1_accuracy: 0.9678 - val_c2_accuracy: 0.9684 - val_c3_accuracy: 0.9750 - val_c4_accuracy: 0.9736 - val_c5_accuracy: 0.9729\n",
            "Epoch 154/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.9096 - c1_loss: 0.2112 - c2_loss: 0.1880 - c3_loss: 0.1802 - c4_loss: 0.1705 - c5_loss: 0.1598 - c1_accuracy: 0.9291 - c2_accuracy: 0.9385 - c3_accuracy: 0.9413 - c4_accuracy: 0.9446 - c5_accuracy: 0.9475 - val_loss: 0.6183 - val_c1_loss: 0.1279 - val_c2_loss: 0.1177 - val_c3_loss: 0.1063 - val_c4_loss: 0.1488 - val_c5_loss: 0.1177 - val_c1_accuracy: 0.9729 - val_c2_accuracy: 0.9746 - val_c3_accuracy: 0.9763 - val_c4_accuracy: 0.9722 - val_c5_accuracy: 0.9678\n",
            "Epoch 155/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7190 - c1_loss: 0.1446 - c2_loss: 0.1371 - c3_loss: 0.1501 - c4_loss: 0.1421 - c5_loss: 0.1451 - c1_accuracy: 0.9516 - c2_accuracy: 0.9522 - c3_accuracy: 0.9474 - c4_accuracy: 0.9538 - c5_accuracy: 0.9503 - val_loss: 0.5835 - val_c1_loss: 0.1171 - val_c2_loss: 0.1218 - val_c3_loss: 0.1016 - val_c4_loss: 0.1376 - val_c5_loss: 0.1054 - val_c1_accuracy: 0.9760 - val_c2_accuracy: 0.9736 - val_c3_accuracy: 0.9760 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9736\n",
            "Epoch 156/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7089 - c1_loss: 0.1394 - c2_loss: 0.1392 - c3_loss: 0.1441 - c4_loss: 0.1516 - c5_loss: 0.1347 - c1_accuracy: 0.9535 - c2_accuracy: 0.9532 - c3_accuracy: 0.9506 - c4_accuracy: 0.9503 - c5_accuracy: 0.9537 - val_loss: 0.5749 - val_c1_loss: 0.1139 - val_c2_loss: 0.1205 - val_c3_loss: 0.1022 - val_c4_loss: 0.1368 - val_c5_loss: 0.1015 - val_c1_accuracy: 0.9760 - val_c2_accuracy: 0.9729 - val_c3_accuracy: 0.9774 - val_c4_accuracy: 0.9746 - val_c5_accuracy: 0.9736\n",
            "Epoch 157/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6508 - c1_loss: 0.1311 - c2_loss: 0.1358 - c3_loss: 0.1315 - c4_loss: 0.1180 - c5_loss: 0.1344 - c1_accuracy: 0.9572 - c2_accuracy: 0.9547 - c3_accuracy: 0.9535 - c4_accuracy: 0.9607 - c5_accuracy: 0.9527 - val_loss: 0.5763 - val_c1_loss: 0.1147 - val_c2_loss: 0.1234 - val_c3_loss: 0.1016 - val_c4_loss: 0.1360 - val_c5_loss: 0.1006 - val_c1_accuracy: 0.9770 - val_c2_accuracy: 0.9743 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9743 - val_c5_accuracy: 0.9739\n",
            "Epoch 158/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6705 - c1_loss: 0.1306 - c2_loss: 0.1363 - c3_loss: 0.1322 - c4_loss: 0.1291 - c5_loss: 0.1423 - c1_accuracy: 0.9562 - c2_accuracy: 0.9535 - c3_accuracy: 0.9557 - c4_accuracy: 0.9575 - c5_accuracy: 0.9554 - val_loss: 0.5839 - val_c1_loss: 0.1147 - val_c2_loss: 0.1223 - val_c3_loss: 0.1053 - val_c4_loss: 0.1384 - val_c5_loss: 0.1033 - val_c1_accuracy: 0.9760 - val_c2_accuracy: 0.9708 - val_c3_accuracy: 0.9732 - val_c4_accuracy: 0.9732 - val_c5_accuracy: 0.9743\n",
            "Epoch 159/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.6074 - c1_loss: 0.1252 - c2_loss: 0.1278 - c3_loss: 0.1151 - c4_loss: 0.1280 - c5_loss: 0.1114 - c1_accuracy: 0.9599 - c2_accuracy: 0.9565 - c3_accuracy: 0.9613 - c4_accuracy: 0.9568 - c5_accuracy: 0.9635 - val_loss: 0.5680 - val_c1_loss: 0.1108 - val_c2_loss: 0.1273 - val_c3_loss: 0.1005 - val_c4_loss: 0.1315 - val_c5_loss: 0.0979 - val_c1_accuracy: 0.9801 - val_c2_accuracy: 0.9726 - val_c3_accuracy: 0.9770 - val_c4_accuracy: 0.9732 - val_c5_accuracy: 0.9746\n",
            "Epoch 160/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7591 - c1_loss: 0.1706 - c2_loss: 0.1454 - c3_loss: 0.1501 - c4_loss: 0.1398 - c5_loss: 0.1534 - c1_accuracy: 0.9478 - c2_accuracy: 0.9522 - c3_accuracy: 0.9474 - c4_accuracy: 0.9500 - c5_accuracy: 0.9487 - val_loss: 0.6000 - val_c1_loss: 0.1190 - val_c2_loss: 0.1223 - val_c3_loss: 0.1043 - val_c4_loss: 0.1372 - val_c5_loss: 0.1172 - val_c1_accuracy: 0.9753 - val_c2_accuracy: 0.9750 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9729 - val_c5_accuracy: 0.9722\n",
            "Epoch 161/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.8139 - c1_loss: 0.1555 - c2_loss: 0.1518 - c3_loss: 0.1718 - c4_loss: 0.1607 - c5_loss: 0.1741 - c1_accuracy: 0.9465 - c2_accuracy: 0.9468 - c3_accuracy: 0.9425 - c4_accuracy: 0.9472 - c5_accuracy: 0.9431 - val_loss: 0.5947 - val_c1_loss: 0.1147 - val_c2_loss: 0.1213 - val_c3_loss: 0.1087 - val_c4_loss: 0.1416 - val_c5_loss: 0.1083 - val_c1_accuracy: 0.9770 - val_c2_accuracy: 0.9753 - val_c3_accuracy: 0.9708 - val_c4_accuracy: 0.9739 - val_c5_accuracy: 0.9698\n",
            "Epoch 162/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7703 - c1_loss: 0.1314 - c2_loss: 0.1601 - c3_loss: 0.1546 - c4_loss: 0.1475 - c5_loss: 0.1768 - c1_accuracy: 0.9588 - c2_accuracy: 0.9471 - c3_accuracy: 0.9478 - c4_accuracy: 0.9540 - c5_accuracy: 0.9457 - val_loss: 0.5871 - val_c1_loss: 0.1106 - val_c2_loss: 0.1302 - val_c3_loss: 0.0971 - val_c4_loss: 0.1442 - val_c5_loss: 0.1050 - val_c1_accuracy: 0.9787 - val_c2_accuracy: 0.9688 - val_c3_accuracy: 0.9763 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9732\n",
            "Epoch 163/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6801 - c1_loss: 0.1395 - c2_loss: 0.1405 - c3_loss: 0.1376 - c4_loss: 0.1261 - c5_loss: 0.1364 - c1_accuracy: 0.9521 - c2_accuracy: 0.9507 - c3_accuracy: 0.9540 - c4_accuracy: 0.9603 - c5_accuracy: 0.9538 - val_loss: 0.5732 - val_c1_loss: 0.1152 - val_c2_loss: 0.1217 - val_c3_loss: 0.0947 - val_c4_loss: 0.1398 - val_c5_loss: 0.1018 - val_c1_accuracy: 0.9798 - val_c2_accuracy: 0.9722 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9736 - val_c5_accuracy: 0.9732\n",
            "Epoch 164/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6462 - c1_loss: 0.1192 - c2_loss: 0.1271 - c3_loss: 0.1320 - c4_loss: 0.1348 - c5_loss: 0.1332 - c1_accuracy: 0.9593 - c2_accuracy: 0.9554 - c3_accuracy: 0.9537 - c4_accuracy: 0.9552 - c5_accuracy: 0.9565 - val_loss: 0.5736 - val_c1_loss: 0.1071 - val_c2_loss: 0.1240 - val_c3_loss: 0.1024 - val_c4_loss: 0.1339 - val_c5_loss: 0.1061 - val_c1_accuracy: 0.9804 - val_c2_accuracy: 0.9726 - val_c3_accuracy: 0.9736 - val_c4_accuracy: 0.9756 - val_c5_accuracy: 0.9722\n",
            "Epoch 165/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7074 - c1_loss: 0.1406 - c2_loss: 0.1470 - c3_loss: 0.1406 - c4_loss: 0.1429 - c5_loss: 0.1363 - c1_accuracy: 0.9493 - c2_accuracy: 0.9518 - c3_accuracy: 0.9538 - c4_accuracy: 0.9528 - c5_accuracy: 0.9550 - val_loss: 0.5900 - val_c1_loss: 0.1123 - val_c2_loss: 0.1297 - val_c3_loss: 0.0994 - val_c4_loss: 0.1403 - val_c5_loss: 0.1083 - val_c1_accuracy: 0.9770 - val_c2_accuracy: 0.9695 - val_c3_accuracy: 0.9750 - val_c4_accuracy: 0.9736 - val_c5_accuracy: 0.9722\n",
            "Epoch 166/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6897 - c1_loss: 0.1370 - c2_loss: 0.1440 - c3_loss: 0.1337 - c4_loss: 0.1334 - c5_loss: 0.1416 - c1_accuracy: 0.9540 - c2_accuracy: 0.9509 - c3_accuracy: 0.9529 - c4_accuracy: 0.9540 - c5_accuracy: 0.9527 - val_loss: 0.5888 - val_c1_loss: 0.1197 - val_c2_loss: 0.1225 - val_c3_loss: 0.0997 - val_c4_loss: 0.1448 - val_c5_loss: 0.1021 - val_c1_accuracy: 0.9760 - val_c2_accuracy: 0.9732 - val_c3_accuracy: 0.9750 - val_c4_accuracy: 0.9719 - val_c5_accuracy: 0.9736\n",
            "Epoch 167/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6135 - c1_loss: 0.1248 - c2_loss: 0.1184 - c3_loss: 0.1246 - c4_loss: 0.1193 - c5_loss: 0.1264 - c1_accuracy: 0.9594 - c2_accuracy: 0.9591 - c3_accuracy: 0.9584 - c4_accuracy: 0.9602 - c5_accuracy: 0.9582 - val_loss: 0.5793 - val_c1_loss: 0.1121 - val_c2_loss: 0.1229 - val_c3_loss: 0.0984 - val_c4_loss: 0.1477 - val_c5_loss: 0.0982 - val_c1_accuracy: 0.9791 - val_c2_accuracy: 0.9756 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9746\n",
            "Epoch 168/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6120 - c1_loss: 0.1202 - c2_loss: 0.1352 - c3_loss: 0.1311 - c4_loss: 0.1153 - c5_loss: 0.1102 - c1_accuracy: 0.9603 - c2_accuracy: 0.9563 - c3_accuracy: 0.9577 - c4_accuracy: 0.9616 - c5_accuracy: 0.9631 - val_loss: 0.5773 - val_c1_loss: 0.1143 - val_c2_loss: 0.1230 - val_c3_loss: 0.0958 - val_c4_loss: 0.1424 - val_c5_loss: 0.1018 - val_c1_accuracy: 0.9770 - val_c2_accuracy: 0.9739 - val_c3_accuracy: 0.9774 - val_c4_accuracy: 0.9756 - val_c5_accuracy: 0.9750\n",
            "Epoch 169/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.5945 - c1_loss: 0.1051 - c2_loss: 0.1264 - c3_loss: 0.1192 - c4_loss: 0.1235 - c5_loss: 0.1202 - c1_accuracy: 0.9607 - c2_accuracy: 0.9569 - c3_accuracy: 0.9629 - c4_accuracy: 0.9618 - c5_accuracy: 0.9604 - val_loss: 0.5687 - val_c1_loss: 0.1048 - val_c2_loss: 0.1270 - val_c3_loss: 0.0989 - val_c4_loss: 0.1363 - val_c5_loss: 0.1017 - val_c1_accuracy: 0.9808 - val_c2_accuracy: 0.9732 - val_c3_accuracy: 0.9777 - val_c4_accuracy: 0.9743 - val_c5_accuracy: 0.9767\n",
            "Epoch 170/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5860 - c1_loss: 0.1161 - c2_loss: 0.1233 - c3_loss: 0.1266 - c4_loss: 0.1095 - c5_loss: 0.1106 - c1_accuracy: 0.9615 - c2_accuracy: 0.9577 - c3_accuracy: 0.9577 - c4_accuracy: 0.9654 - c5_accuracy: 0.9627 - val_loss: 0.5524 - val_c1_loss: 0.1039 - val_c2_loss: 0.1249 - val_c3_loss: 0.0938 - val_c4_loss: 0.1312 - val_c5_loss: 0.0987 - val_c1_accuracy: 0.9815 - val_c2_accuracy: 0.9743 - val_c3_accuracy: 0.9774 - val_c4_accuracy: 0.9763 - val_c5_accuracy: 0.9743\n",
            "Epoch 171/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7975 - c1_loss: 0.1574 - c2_loss: 0.1584 - c3_loss: 0.1538 - c4_loss: 0.1535 - c5_loss: 0.1743 - c1_accuracy: 0.9484 - c2_accuracy: 0.9500 - c3_accuracy: 0.9485 - c4_accuracy: 0.9509 - c5_accuracy: 0.9441 - val_loss: 0.5717 - val_c1_loss: 0.1072 - val_c2_loss: 0.1215 - val_c3_loss: 0.0996 - val_c4_loss: 0.1376 - val_c5_loss: 0.1058 - val_c1_accuracy: 0.9808 - val_c2_accuracy: 0.9750 - val_c3_accuracy: 0.9732 - val_c4_accuracy: 0.9753 - val_c5_accuracy: 0.9729\n",
            "Epoch 172/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6567 - c1_loss: 0.1331 - c2_loss: 0.1293 - c3_loss: 0.1261 - c4_loss: 0.1389 - c5_loss: 0.1294 - c1_accuracy: 0.9554 - c2_accuracy: 0.9544 - c3_accuracy: 0.9619 - c4_accuracy: 0.9549 - c5_accuracy: 0.9566 - val_loss: 0.5503 - val_c1_loss: 0.1084 - val_c2_loss: 0.1130 - val_c3_loss: 0.0967 - val_c4_loss: 0.1330 - val_c5_loss: 0.0991 - val_c1_accuracy: 0.9808 - val_c2_accuracy: 0.9763 - val_c3_accuracy: 0.9780 - val_c4_accuracy: 0.9746 - val_c5_accuracy: 0.9756\n",
            "Epoch 173/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5732 - c1_loss: 0.1152 - c2_loss: 0.1182 - c3_loss: 0.1190 - c4_loss: 0.1107 - c5_loss: 0.1100 - c1_accuracy: 0.9621 - c2_accuracy: 0.9599 - c3_accuracy: 0.9625 - c4_accuracy: 0.9607 - c5_accuracy: 0.9641 - val_loss: 0.5566 - val_c1_loss: 0.1007 - val_c2_loss: 0.1161 - val_c3_loss: 0.1024 - val_c4_loss: 0.1388 - val_c5_loss: 0.0985 - val_c1_accuracy: 0.9828 - val_c2_accuracy: 0.9753 - val_c3_accuracy: 0.9746 - val_c4_accuracy: 0.9732 - val_c5_accuracy: 0.9746\n",
            "Epoch 174/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.5721 - c1_loss: 0.1142 - c2_loss: 0.1129 - c3_loss: 0.1115 - c4_loss: 0.1176 - c5_loss: 0.1160 - c1_accuracy: 0.9641 - c2_accuracy: 0.9649 - c3_accuracy: 0.9637 - c4_accuracy: 0.9593 - c5_accuracy: 0.9619 - val_loss: 0.5588 - val_c1_loss: 0.1087 - val_c2_loss: 0.1139 - val_c3_loss: 0.1042 - val_c4_loss: 0.1317 - val_c5_loss: 0.1003 - val_c1_accuracy: 0.9791 - val_c2_accuracy: 0.9753 - val_c3_accuracy: 0.9715 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9756\n",
            "Epoch 175/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.5679 - c1_loss: 0.1130 - c2_loss: 0.1203 - c3_loss: 0.1070 - c4_loss: 0.1141 - c5_loss: 0.1134 - c1_accuracy: 0.9628 - c2_accuracy: 0.9606 - c3_accuracy: 0.9637 - c4_accuracy: 0.9618 - c5_accuracy: 0.9632 - val_loss: 0.5478 - val_c1_loss: 0.1057 - val_c2_loss: 0.1158 - val_c3_loss: 0.0999 - val_c4_loss: 0.1301 - val_c5_loss: 0.0963 - val_c1_accuracy: 0.9787 - val_c2_accuracy: 0.9746 - val_c3_accuracy: 0.9767 - val_c4_accuracy: 0.9743 - val_c5_accuracy: 0.9739\n",
            "Epoch 176/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5869 - c1_loss: 0.1180 - c2_loss: 0.1146 - c3_loss: 0.1214 - c4_loss: 0.1189 - c5_loss: 0.1141 - c1_accuracy: 0.9602 - c2_accuracy: 0.9603 - c3_accuracy: 0.9600 - c4_accuracy: 0.9584 - c5_accuracy: 0.9621 - val_loss: 0.5435 - val_c1_loss: 0.0968 - val_c2_loss: 0.1160 - val_c3_loss: 0.1031 - val_c4_loss: 0.1280 - val_c5_loss: 0.0995 - val_c1_accuracy: 0.9815 - val_c2_accuracy: 0.9753 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9743\n",
            "Epoch 177/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5584 - c1_loss: 0.1092 - c2_loss: 0.1114 - c3_loss: 0.1057 - c4_loss: 0.1161 - c5_loss: 0.1159 - c1_accuracy: 0.9643 - c2_accuracy: 0.9640 - c3_accuracy: 0.9619 - c4_accuracy: 0.9618 - c5_accuracy: 0.9599 - val_loss: 0.5361 - val_c1_loss: 0.1046 - val_c2_loss: 0.1198 - val_c3_loss: 0.0968 - val_c4_loss: 0.1278 - val_c5_loss: 0.0871 - val_c1_accuracy: 0.9822 - val_c2_accuracy: 0.9736 - val_c3_accuracy: 0.9760 - val_c4_accuracy: 0.9780 - val_c5_accuracy: 0.9794\n",
            "Epoch 178/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5332 - c1_loss: 0.0986 - c2_loss: 0.1086 - c3_loss: 0.1100 - c4_loss: 0.1102 - c5_loss: 0.1059 - c1_accuracy: 0.9684 - c2_accuracy: 0.9627 - c3_accuracy: 0.9616 - c4_accuracy: 0.9616 - c5_accuracy: 0.9607 - val_loss: 0.5443 - val_c1_loss: 0.1080 - val_c2_loss: 0.1199 - val_c3_loss: 0.0974 - val_c4_loss: 0.1236 - val_c5_loss: 0.0953 - val_c1_accuracy: 0.9801 - val_c2_accuracy: 0.9760 - val_c3_accuracy: 0.9760 - val_c4_accuracy: 0.9756 - val_c5_accuracy: 0.9777\n",
            "Epoch 179/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5308 - c1_loss: 0.1030 - c2_loss: 0.1191 - c3_loss: 0.1101 - c4_loss: 0.1057 - c5_loss: 0.0928 - c1_accuracy: 0.9640 - c2_accuracy: 0.9588 - c3_accuracy: 0.9612 - c4_accuracy: 0.9654 - c5_accuracy: 0.9659 - val_loss: 0.5541 - val_c1_loss: 0.1090 - val_c2_loss: 0.1147 - val_c3_loss: 0.0987 - val_c4_loss: 0.1344 - val_c5_loss: 0.0972 - val_c1_accuracy: 0.9798 - val_c2_accuracy: 0.9746 - val_c3_accuracy: 0.9732 - val_c4_accuracy: 0.9732 - val_c5_accuracy: 0.9767\n",
            "Epoch 180/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.5418 - c1_loss: 0.0980 - c2_loss: 0.1141 - c3_loss: 0.1086 - c4_loss: 0.1144 - c5_loss: 0.1067 - c1_accuracy: 0.9637 - c2_accuracy: 0.9624 - c3_accuracy: 0.9644 - c4_accuracy: 0.9625 - c5_accuracy: 0.9650 - val_loss: 0.5656 - val_c1_loss: 0.1111 - val_c2_loss: 0.1180 - val_c3_loss: 0.1037 - val_c4_loss: 0.1389 - val_c5_loss: 0.0939 - val_c1_accuracy: 0.9787 - val_c2_accuracy: 0.9791 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9780\n",
            "Epoch 181/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.6832 - c1_loss: 0.1375 - c2_loss: 0.1475 - c3_loss: 0.1341 - c4_loss: 0.1207 - c5_loss: 0.1434 - c1_accuracy: 0.9547 - c2_accuracy: 0.9546 - c3_accuracy: 0.9547 - c4_accuracy: 0.9563 - c5_accuracy: 0.9538 - val_loss: 0.5726 - val_c1_loss: 0.1106 - val_c2_loss: 0.1271 - val_c3_loss: 0.1002 - val_c4_loss: 0.1363 - val_c5_loss: 0.0984 - val_c1_accuracy: 0.9801 - val_c2_accuracy: 0.9729 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9729 - val_c5_accuracy: 0.9763\n",
            "Epoch 182/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7527 - c1_loss: 0.1609 - c2_loss: 0.1412 - c3_loss: 0.1417 - c4_loss: 0.1432 - c5_loss: 0.1658 - c1_accuracy: 0.9491 - c2_accuracy: 0.9528 - c3_accuracy: 0.9554 - c4_accuracy: 0.9516 - c5_accuracy: 0.9437 - val_loss: 0.5891 - val_c1_loss: 0.1134 - val_c2_loss: 0.1319 - val_c3_loss: 0.1044 - val_c4_loss: 0.1337 - val_c5_loss: 0.1056 - val_c1_accuracy: 0.9770 - val_c2_accuracy: 0.9726 - val_c3_accuracy: 0.9722 - val_c4_accuracy: 0.9753 - val_c5_accuracy: 0.9729\n",
            "Epoch 183/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.6324 - c1_loss: 0.1169 - c2_loss: 0.1275 - c3_loss: 0.1239 - c4_loss: 0.1255 - c5_loss: 0.1387 - c1_accuracy: 0.9607 - c2_accuracy: 0.9607 - c3_accuracy: 0.9600 - c4_accuracy: 0.9571 - c5_accuracy: 0.9541 - val_loss: 0.5703 - val_c1_loss: 0.1070 - val_c2_loss: 0.1287 - val_c3_loss: 0.0997 - val_c4_loss: 0.1275 - val_c5_loss: 0.1074 - val_c1_accuracy: 0.9791 - val_c2_accuracy: 0.9729 - val_c3_accuracy: 0.9763 - val_c4_accuracy: 0.9770 - val_c5_accuracy: 0.9729\n",
            "Epoch 184/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.5805 - c1_loss: 0.1179 - c2_loss: 0.1177 - c3_loss: 0.1126 - c4_loss: 0.1163 - c5_loss: 0.1160 - c1_accuracy: 0.9612 - c2_accuracy: 0.9587 - c3_accuracy: 0.9644 - c4_accuracy: 0.9603 - c5_accuracy: 0.9622 - val_loss: 0.5795 - val_c1_loss: 0.1086 - val_c2_loss: 0.1222 - val_c3_loss: 0.1084 - val_c4_loss: 0.1360 - val_c5_loss: 0.1043 - val_c1_accuracy: 0.9794 - val_c2_accuracy: 0.9743 - val_c3_accuracy: 0.9729 - val_c4_accuracy: 0.9729 - val_c5_accuracy: 0.9722\n",
            "Epoch 185/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.5915 - c1_loss: 0.1160 - c2_loss: 0.1149 - c3_loss: 0.1287 - c4_loss: 0.1182 - c5_loss: 0.1137 - c1_accuracy: 0.9629 - c2_accuracy: 0.9622 - c3_accuracy: 0.9579 - c4_accuracy: 0.9603 - c5_accuracy: 0.9621 - val_loss: 0.5618 - val_c1_loss: 0.0997 - val_c2_loss: 0.1241 - val_c3_loss: 0.0995 - val_c4_loss: 0.1355 - val_c5_loss: 0.1029 - val_c1_accuracy: 0.9801 - val_c2_accuracy: 0.9732 - val_c3_accuracy: 0.9760 - val_c4_accuracy: 0.9722 - val_c5_accuracy: 0.9729\n",
            "Epoch 186/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5855 - c1_loss: 0.1236 - c2_loss: 0.1111 - c3_loss: 0.1170 - c4_loss: 0.1181 - c5_loss: 0.1157 - c1_accuracy: 0.9606 - c2_accuracy: 0.9624 - c3_accuracy: 0.9627 - c4_accuracy: 0.9609 - c5_accuracy: 0.9577 - val_loss: 0.5545 - val_c1_loss: 0.1031 - val_c2_loss: 0.1250 - val_c3_loss: 0.1029 - val_c4_loss: 0.1305 - val_c5_loss: 0.0930 - val_c1_accuracy: 0.9787 - val_c2_accuracy: 0.9750 - val_c3_accuracy: 0.9739 - val_c4_accuracy: 0.9753 - val_c5_accuracy: 0.9767\n",
            "Epoch 187/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.6362 - c1_loss: 0.1192 - c2_loss: 0.1307 - c3_loss: 0.1190 - c4_loss: 0.1310 - c5_loss: 0.1363 - c1_accuracy: 0.9625 - c2_accuracy: 0.9556 - c3_accuracy: 0.9606 - c4_accuracy: 0.9538 - c5_accuracy: 0.9515 - val_loss: 0.5953 - val_c1_loss: 0.1113 - val_c2_loss: 0.1245 - val_c3_loss: 0.1089 - val_c4_loss: 0.1479 - val_c5_loss: 0.1026 - val_c1_accuracy: 0.9767 - val_c2_accuracy: 0.9753 - val_c3_accuracy: 0.9726 - val_c4_accuracy: 0.9739 - val_c5_accuracy: 0.9760\n",
            "Epoch 188/500\n",
            "69/69 [==============================] - 2s 32ms/step - loss: 0.6907 - c1_loss: 0.1468 - c2_loss: 0.1336 - c3_loss: 0.1472 - c4_loss: 0.1346 - c5_loss: 0.1285 - c1_accuracy: 0.9507 - c2_accuracy: 0.9556 - c3_accuracy: 0.9507 - c4_accuracy: 0.9537 - c5_accuracy: 0.9556 - val_loss: 0.5894 - val_c1_loss: 0.1131 - val_c2_loss: 0.1144 - val_c3_loss: 0.1166 - val_c4_loss: 0.1357 - val_c5_loss: 0.1097 - val_c1_accuracy: 0.9732 - val_c2_accuracy: 0.9753 - val_c3_accuracy: 0.9702 - val_c4_accuracy: 0.9695 - val_c5_accuracy: 0.9729\n",
            "Epoch 189/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6277 - c1_loss: 0.1173 - c2_loss: 0.1308 - c3_loss: 0.1356 - c4_loss: 0.1312 - c5_loss: 0.1129 - c1_accuracy: 0.9602 - c2_accuracy: 0.9568 - c3_accuracy: 0.9553 - c4_accuracy: 0.9543 - c5_accuracy: 0.9613 - val_loss: 0.5751 - val_c1_loss: 0.1008 - val_c2_loss: 0.1231 - val_c3_loss: 0.1076 - val_c4_loss: 0.1395 - val_c5_loss: 0.1042 - val_c1_accuracy: 0.9801 - val_c2_accuracy: 0.9743 - val_c3_accuracy: 0.9719 - val_c4_accuracy: 0.9753 - val_c5_accuracy: 0.9722\n",
            "Epoch 190/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6128 - c1_loss: 0.1168 - c2_loss: 0.1297 - c3_loss: 0.1243 - c4_loss: 0.1129 - c5_loss: 0.1291 - c1_accuracy: 0.9596 - c2_accuracy: 0.9585 - c3_accuracy: 0.9596 - c4_accuracy: 0.9606 - c5_accuracy: 0.9582 - val_loss: 0.5805 - val_c1_loss: 0.1022 - val_c2_loss: 0.1276 - val_c3_loss: 0.1129 - val_c4_loss: 0.1369 - val_c5_loss: 0.1010 - val_c1_accuracy: 0.9794 - val_c2_accuracy: 0.9712 - val_c3_accuracy: 0.9722 - val_c4_accuracy: 0.9760 - val_c5_accuracy: 0.9753\n",
            "Epoch 191/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.5576 - c1_loss: 0.1155 - c2_loss: 0.1105 - c3_loss: 0.1198 - c4_loss: 0.1045 - c5_loss: 0.1072 - c1_accuracy: 0.9596 - c2_accuracy: 0.9641 - c3_accuracy: 0.9597 - c4_accuracy: 0.9650 - c5_accuracy: 0.9629 - val_loss: 0.5756 - val_c1_loss: 0.1066 - val_c2_loss: 0.1247 - val_c3_loss: 0.1051 - val_c4_loss: 0.1299 - val_c5_loss: 0.1093 - val_c1_accuracy: 0.9791 - val_c2_accuracy: 0.9719 - val_c3_accuracy: 0.9763 - val_c4_accuracy: 0.9743 - val_c5_accuracy: 0.9719\n",
            "Epoch 192/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.6664 - c1_loss: 0.1217 - c2_loss: 0.1254 - c3_loss: 0.1271 - c4_loss: 0.1557 - c5_loss: 0.1366 - c1_accuracy: 0.9587 - c2_accuracy: 0.9553 - c3_accuracy: 0.9604 - c4_accuracy: 0.9499 - c5_accuracy: 0.9535 - val_loss: 0.5739 - val_c1_loss: 0.1016 - val_c2_loss: 0.1223 - val_c3_loss: 0.1067 - val_c4_loss: 0.1417 - val_c5_loss: 0.1016 - val_c1_accuracy: 0.9801 - val_c2_accuracy: 0.9719 - val_c3_accuracy: 0.9708 - val_c4_accuracy: 0.9746 - val_c5_accuracy: 0.9756\n",
            "Epoch 193/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7375 - c1_loss: 0.1553 - c2_loss: 0.1461 - c3_loss: 0.1457 - c4_loss: 0.1386 - c5_loss: 0.1518 - c1_accuracy: 0.9494 - c2_accuracy: 0.9488 - c3_accuracy: 0.9541 - c4_accuracy: 0.9543 - c5_accuracy: 0.9509 - val_loss: 0.5679 - val_c1_loss: 0.1072 - val_c2_loss: 0.1188 - val_c3_loss: 0.1021 - val_c4_loss: 0.1398 - val_c5_loss: 0.1000 - val_c1_accuracy: 0.9791 - val_c2_accuracy: 0.9729 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9753\n",
            "Epoch 194/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5904 - c1_loss: 0.1183 - c2_loss: 0.1308 - c3_loss: 0.1190 - c4_loss: 0.1145 - c5_loss: 0.1077 - c1_accuracy: 0.9590 - c2_accuracy: 0.9575 - c3_accuracy: 0.9609 - c4_accuracy: 0.9596 - c5_accuracy: 0.9635 - val_loss: 0.5752 - val_c1_loss: 0.1082 - val_c2_loss: 0.1239 - val_c3_loss: 0.1012 - val_c4_loss: 0.1458 - val_c5_loss: 0.0961 - val_c1_accuracy: 0.9767 - val_c2_accuracy: 0.9736 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9729 - val_c5_accuracy: 0.9767\n",
            "Epoch 195/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.5444 - c1_loss: 0.1077 - c2_loss: 0.1143 - c3_loss: 0.1060 - c4_loss: 0.1065 - c5_loss: 0.1099 - c1_accuracy: 0.9659 - c2_accuracy: 0.9618 - c3_accuracy: 0.9640 - c4_accuracy: 0.9628 - c5_accuracy: 0.9613 - val_loss: 0.5482 - val_c1_loss: 0.0958 - val_c2_loss: 0.1205 - val_c3_loss: 0.0984 - val_c4_loss: 0.1360 - val_c5_loss: 0.0975 - val_c1_accuracy: 0.9804 - val_c2_accuracy: 0.9746 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9760 - val_c5_accuracy: 0.9770\n",
            "Epoch 196/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5352 - c1_loss: 0.1093 - c2_loss: 0.1001 - c3_loss: 0.1113 - c4_loss: 0.1060 - c5_loss: 0.1085 - c1_accuracy: 0.9629 - c2_accuracy: 0.9650 - c3_accuracy: 0.9649 - c4_accuracy: 0.9662 - c5_accuracy: 0.9637 - val_loss: 0.5427 - val_c1_loss: 0.1004 - val_c2_loss: 0.1160 - val_c3_loss: 0.0985 - val_c4_loss: 0.1289 - val_c5_loss: 0.0988 - val_c1_accuracy: 0.9798 - val_c2_accuracy: 0.9750 - val_c3_accuracy: 0.9750 - val_c4_accuracy: 0.9784 - val_c5_accuracy: 0.9780\n",
            "Epoch 197/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.5253 - c1_loss: 0.1015 - c2_loss: 0.1194 - c3_loss: 0.0994 - c4_loss: 0.1030 - c5_loss: 0.1020 - c1_accuracy: 0.9640 - c2_accuracy: 0.9613 - c3_accuracy: 0.9693 - c4_accuracy: 0.9635 - c5_accuracy: 0.9671 - val_loss: 0.5597 - val_c1_loss: 0.1030 - val_c2_loss: 0.1220 - val_c3_loss: 0.0986 - val_c4_loss: 0.1398 - val_c5_loss: 0.0962 - val_c1_accuracy: 0.9791 - val_c2_accuracy: 0.9743 - val_c3_accuracy: 0.9746 - val_c4_accuracy: 0.9780 - val_c5_accuracy: 0.9784\n",
            "Epoch 198/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5360 - c1_loss: 0.1060 - c2_loss: 0.1133 - c3_loss: 0.1128 - c4_loss: 0.1058 - c5_loss: 0.0982 - c1_accuracy: 0.9635 - c2_accuracy: 0.9615 - c3_accuracy: 0.9624 - c4_accuracy: 0.9621 - c5_accuracy: 0.9665 - val_loss: 0.5601 - val_c1_loss: 0.1063 - val_c2_loss: 0.1227 - val_c3_loss: 0.0948 - val_c4_loss: 0.1381 - val_c5_loss: 0.0981 - val_c1_accuracy: 0.9794 - val_c2_accuracy: 0.9739 - val_c3_accuracy: 0.9774 - val_c4_accuracy: 0.9760 - val_c5_accuracy: 0.9770\n",
            "Epoch 199/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5960 - c1_loss: 0.1150 - c2_loss: 0.1120 - c3_loss: 0.1205 - c4_loss: 0.1239 - c5_loss: 0.1245 - c1_accuracy: 0.9594 - c2_accuracy: 0.9632 - c3_accuracy: 0.9587 - c4_accuracy: 0.9603 - c5_accuracy: 0.9591 - val_loss: 0.5864 - val_c1_loss: 0.1139 - val_c2_loss: 0.1346 - val_c3_loss: 0.1000 - val_c4_loss: 0.1312 - val_c5_loss: 0.1067 - val_c1_accuracy: 0.9767 - val_c2_accuracy: 0.9736 - val_c3_accuracy: 0.9791 - val_c4_accuracy: 0.9767 - val_c5_accuracy: 0.9750\n",
            "Epoch 200/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.8210 - c1_loss: 0.1546 - c2_loss: 0.1642 - c3_loss: 0.1647 - c4_loss: 0.1580 - c5_loss: 0.1795 - c1_accuracy: 0.9482 - c2_accuracy: 0.9460 - c3_accuracy: 0.9465 - c4_accuracy: 0.9449 - c5_accuracy: 0.9410 - val_loss: 0.5931 - val_c1_loss: 0.1130 - val_c2_loss: 0.1296 - val_c3_loss: 0.0994 - val_c4_loss: 0.1466 - val_c5_loss: 0.1045 - val_c1_accuracy: 0.9774 - val_c2_accuracy: 0.9726 - val_c3_accuracy: 0.9750 - val_c4_accuracy: 0.9732 - val_c5_accuracy: 0.9739\n",
            "Epoch 201/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7125 - c1_loss: 0.1371 - c2_loss: 0.1529 - c3_loss: 0.1413 - c4_loss: 0.1449 - c5_loss: 0.1362 - c1_accuracy: 0.9522 - c2_accuracy: 0.9503 - c3_accuracy: 0.9529 - c4_accuracy: 0.9535 - c5_accuracy: 0.9568 - val_loss: 0.5778 - val_c1_loss: 0.1118 - val_c2_loss: 0.1218 - val_c3_loss: 0.1030 - val_c4_loss: 0.1412 - val_c5_loss: 0.1000 - val_c1_accuracy: 0.9784 - val_c2_accuracy: 0.9726 - val_c3_accuracy: 0.9760 - val_c4_accuracy: 0.9750 - val_c5_accuracy: 0.9732\n",
            "Epoch 202/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6169 - c1_loss: 0.1193 - c2_loss: 0.1246 - c3_loss: 0.1301 - c4_loss: 0.1194 - c5_loss: 0.1235 - c1_accuracy: 0.9596 - c2_accuracy: 0.9604 - c3_accuracy: 0.9568 - c4_accuracy: 0.9619 - c5_accuracy: 0.9575 - val_loss: 0.5815 - val_c1_loss: 0.1090 - val_c2_loss: 0.1238 - val_c3_loss: 0.1069 - val_c4_loss: 0.1415 - val_c5_loss: 0.1003 - val_c1_accuracy: 0.9791 - val_c2_accuracy: 0.9722 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9760 - val_c5_accuracy: 0.9722\n",
            "Epoch 203/500\n",
            "69/69 [==============================] - 2s 32ms/step - loss: 0.6082 - c1_loss: 0.1202 - c2_loss: 0.1230 - c3_loss: 0.1334 - c4_loss: 0.1183 - c5_loss: 0.1133 - c1_accuracy: 0.9618 - c2_accuracy: 0.9575 - c3_accuracy: 0.9569 - c4_accuracy: 0.9599 - c5_accuracy: 0.9609 - val_loss: 0.5690 - val_c1_loss: 0.1057 - val_c2_loss: 0.1232 - val_c3_loss: 0.1022 - val_c4_loss: 0.1405 - val_c5_loss: 0.0974 - val_c1_accuracy: 0.9804 - val_c2_accuracy: 0.9732 - val_c3_accuracy: 0.9777 - val_c4_accuracy: 0.9739 - val_c5_accuracy: 0.9746\n",
            "Epoch 204/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5607 - c1_loss: 0.1084 - c2_loss: 0.1012 - c3_loss: 0.1204 - c4_loss: 0.1205 - c5_loss: 0.1102 - c1_accuracy: 0.9641 - c2_accuracy: 0.9638 - c3_accuracy: 0.9594 - c4_accuracy: 0.9602 - c5_accuracy: 0.9616 - val_loss: 0.5681 - val_c1_loss: 0.1050 - val_c2_loss: 0.1232 - val_c3_loss: 0.1032 - val_c4_loss: 0.1374 - val_c5_loss: 0.0993 - val_c1_accuracy: 0.9784 - val_c2_accuracy: 0.9736 - val_c3_accuracy: 0.9763 - val_c4_accuracy: 0.9746 - val_c5_accuracy: 0.9750\n",
            "Epoch 205/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5414 - c1_loss: 0.1122 - c2_loss: 0.1103 - c3_loss: 0.1059 - c4_loss: 0.1058 - c5_loss: 0.1073 - c1_accuracy: 0.9647 - c2_accuracy: 0.9618 - c3_accuracy: 0.9629 - c4_accuracy: 0.9666 - c5_accuracy: 0.9668 - val_loss: 0.5502 - val_c1_loss: 0.1070 - val_c2_loss: 0.1224 - val_c3_loss: 0.0968 - val_c4_loss: 0.1309 - val_c5_loss: 0.0930 - val_c1_accuracy: 0.9784 - val_c2_accuracy: 0.9739 - val_c3_accuracy: 0.9787 - val_c4_accuracy: 0.9729 - val_c5_accuracy: 0.9784\n",
            "Epoch 206/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5366 - c1_loss: 0.1033 - c2_loss: 0.1064 - c3_loss: 0.1140 - c4_loss: 0.1050 - c5_loss: 0.1079 - c1_accuracy: 0.9644 - c2_accuracy: 0.9647 - c3_accuracy: 0.9604 - c4_accuracy: 0.9652 - c5_accuracy: 0.9632 - val_loss: 0.5362 - val_c1_loss: 0.1001 - val_c2_loss: 0.1195 - val_c3_loss: 0.0987 - val_c4_loss: 0.1286 - val_c5_loss: 0.0892 - val_c1_accuracy: 0.9811 - val_c2_accuracy: 0.9739 - val_c3_accuracy: 0.9787 - val_c4_accuracy: 0.9777 - val_c5_accuracy: 0.9808\n",
            "Epoch 207/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.5395 - c1_loss: 0.1192 - c2_loss: 0.1131 - c3_loss: 0.1106 - c4_loss: 0.1000 - c5_loss: 0.0965 - c1_accuracy: 0.9600 - c2_accuracy: 0.9629 - c3_accuracy: 0.9637 - c4_accuracy: 0.9640 - c5_accuracy: 0.9657 - val_loss: 0.5709 - val_c1_loss: 0.1074 - val_c2_loss: 0.1238 - val_c3_loss: 0.1080 - val_c4_loss: 0.1366 - val_c5_loss: 0.0951 - val_c1_accuracy: 0.9787 - val_c2_accuracy: 0.9760 - val_c3_accuracy: 0.9753 - val_c4_accuracy: 0.9756 - val_c5_accuracy: 0.9763\n",
            "Epoch 208/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6324 - c1_loss: 0.1183 - c2_loss: 0.1359 - c3_loss: 0.1227 - c4_loss: 0.1358 - c5_loss: 0.1197 - c1_accuracy: 0.9596 - c2_accuracy: 0.9565 - c3_accuracy: 0.9591 - c4_accuracy: 0.9544 - c5_accuracy: 0.9607 - val_loss: 0.5638 - val_c1_loss: 0.1073 - val_c2_loss: 0.1242 - val_c3_loss: 0.0985 - val_c4_loss: 0.1357 - val_c5_loss: 0.0983 - val_c1_accuracy: 0.9791 - val_c2_accuracy: 0.9746 - val_c3_accuracy: 0.9767 - val_c4_accuracy: 0.9760 - val_c5_accuracy: 0.9756\n",
            "Epoch 209/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5444 - c1_loss: 0.1029 - c2_loss: 0.1084 - c3_loss: 0.1114 - c4_loss: 0.1109 - c5_loss: 0.1109 - c1_accuracy: 0.9671 - c2_accuracy: 0.9629 - c3_accuracy: 0.9618 - c4_accuracy: 0.9625 - c5_accuracy: 0.9644 - val_loss: 0.5577 - val_c1_loss: 0.1064 - val_c2_loss: 0.1211 - val_c3_loss: 0.0993 - val_c4_loss: 0.1330 - val_c5_loss: 0.0979 - val_c1_accuracy: 0.9794 - val_c2_accuracy: 0.9746 - val_c3_accuracy: 0.9770 - val_c4_accuracy: 0.9746 - val_c5_accuracy: 0.9739\n",
            "Epoch 210/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6293 - c1_loss: 0.1237 - c2_loss: 0.1414 - c3_loss: 0.1151 - c4_loss: 0.1188 - c5_loss: 0.1303 - c1_accuracy: 0.9549 - c2_accuracy: 0.9512 - c3_accuracy: 0.9590 - c4_accuracy: 0.9609 - c5_accuracy: 0.9571 - val_loss: 0.5944 - val_c1_loss: 0.1104 - val_c2_loss: 0.1253 - val_c3_loss: 0.1091 - val_c4_loss: 0.1456 - val_c5_loss: 0.1040 - val_c1_accuracy: 0.9770 - val_c2_accuracy: 0.9736 - val_c3_accuracy: 0.9712 - val_c4_accuracy: 0.9743 - val_c5_accuracy: 0.9705\n",
            "Epoch 211/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6542 - c1_loss: 0.1137 - c2_loss: 0.1372 - c3_loss: 0.1367 - c4_loss: 0.1249 - c5_loss: 0.1417 - c1_accuracy: 0.9615 - c2_accuracy: 0.9552 - c3_accuracy: 0.9532 - c4_accuracy: 0.9603 - c5_accuracy: 0.9524 - val_loss: 0.5864 - val_c1_loss: 0.1077 - val_c2_loss: 0.1249 - val_c3_loss: 0.1017 - val_c4_loss: 0.1519 - val_c5_loss: 0.1002 - val_c1_accuracy: 0.9801 - val_c2_accuracy: 0.9750 - val_c3_accuracy: 0.9760 - val_c4_accuracy: 0.9774 - val_c5_accuracy: 0.9753\n",
            "Epoch 212/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5871 - c1_loss: 0.1097 - c2_loss: 0.1156 - c3_loss: 0.1176 - c4_loss: 0.1234 - c5_loss: 0.1208 - c1_accuracy: 0.9622 - c2_accuracy: 0.9604 - c3_accuracy: 0.9604 - c4_accuracy: 0.9581 - c5_accuracy: 0.9579 - val_loss: 0.5569 - val_c1_loss: 0.1019 - val_c2_loss: 0.1231 - val_c3_loss: 0.0964 - val_c4_loss: 0.1340 - val_c5_loss: 0.1015 - val_c1_accuracy: 0.9815 - val_c2_accuracy: 0.9729 - val_c3_accuracy: 0.9774 - val_c4_accuracy: 0.9770 - val_c5_accuracy: 0.9743\n",
            "Epoch 213/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5875 - c1_loss: 0.1278 - c2_loss: 0.1213 - c3_loss: 0.1030 - c4_loss: 0.1136 - c5_loss: 0.1218 - c1_accuracy: 0.9579 - c2_accuracy: 0.9578 - c3_accuracy: 0.9652 - c4_accuracy: 0.9634 - c5_accuracy: 0.9627 - val_loss: 0.5511 - val_c1_loss: 0.1028 - val_c2_loss: 0.1215 - val_c3_loss: 0.1029 - val_c4_loss: 0.1336 - val_c5_loss: 0.0904 - val_c1_accuracy: 0.9808 - val_c2_accuracy: 0.9736 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9770 - val_c5_accuracy: 0.9791\n",
            "Epoch 214/500\n",
            "69/69 [==============================] - 2s 32ms/step - loss: 0.7344 - c1_loss: 0.1325 - c2_loss: 0.1575 - c3_loss: 0.1480 - c4_loss: 0.1626 - c5_loss: 0.1337 - c1_accuracy: 0.9562 - c2_accuracy: 0.9462 - c3_accuracy: 0.9522 - c4_accuracy: 0.9477 - c5_accuracy: 0.9566 - val_loss: 0.5955 - val_c1_loss: 0.1090 - val_c2_loss: 0.1371 - val_c3_loss: 0.0988 - val_c4_loss: 0.1489 - val_c5_loss: 0.1017 - val_c1_accuracy: 0.9794 - val_c2_accuracy: 0.9729 - val_c3_accuracy: 0.9784 - val_c4_accuracy: 0.9726 - val_c5_accuracy: 0.9750\n",
            "Epoch 215/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5720 - c1_loss: 0.1081 - c2_loss: 0.1148 - c3_loss: 0.1163 - c4_loss: 0.1084 - c5_loss: 0.1245 - c1_accuracy: 0.9641 - c2_accuracy: 0.9602 - c3_accuracy: 0.9615 - c4_accuracy: 0.9637 - c5_accuracy: 0.9590 - val_loss: 0.5885 - val_c1_loss: 0.1147 - val_c2_loss: 0.1238 - val_c3_loss: 0.1012 - val_c4_loss: 0.1475 - val_c5_loss: 0.1013 - val_c1_accuracy: 0.9777 - val_c2_accuracy: 0.9763 - val_c3_accuracy: 0.9767 - val_c4_accuracy: 0.9736 - val_c5_accuracy: 0.9743\n",
            "Epoch 216/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.6681 - c1_loss: 0.1380 - c2_loss: 0.1402 - c3_loss: 0.1246 - c4_loss: 0.1307 - c5_loss: 0.1346 - c1_accuracy: 0.9556 - c2_accuracy: 0.9568 - c3_accuracy: 0.9565 - c4_accuracy: 0.9591 - c5_accuracy: 0.9581 - val_loss: 0.5735 - val_c1_loss: 0.1071 - val_c2_loss: 0.1211 - val_c3_loss: 0.1044 - val_c4_loss: 0.1366 - val_c5_loss: 0.1042 - val_c1_accuracy: 0.9780 - val_c2_accuracy: 0.9739 - val_c3_accuracy: 0.9777 - val_c4_accuracy: 0.9760 - val_c5_accuracy: 0.9719\n",
            "Epoch 217/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5793 - c1_loss: 0.1051 - c2_loss: 0.1225 - c3_loss: 0.1234 - c4_loss: 0.1043 - c5_loss: 0.1240 - c1_accuracy: 0.9659 - c2_accuracy: 0.9609 - c3_accuracy: 0.9604 - c4_accuracy: 0.9637 - c5_accuracy: 0.9581 - val_loss: 0.5877 - val_c1_loss: 0.1080 - val_c2_loss: 0.1287 - val_c3_loss: 0.1096 - val_c4_loss: 0.1419 - val_c5_loss: 0.0995 - val_c1_accuracy: 0.9780 - val_c2_accuracy: 0.9722 - val_c3_accuracy: 0.9746 - val_c4_accuracy: 0.9750 - val_c5_accuracy: 0.9746\n",
            "Epoch 218/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.7531 - c1_loss: 0.1419 - c2_loss: 0.1464 - c3_loss: 0.1684 - c4_loss: 0.1452 - c5_loss: 0.1513 - c1_accuracy: 0.9497 - c2_accuracy: 0.9522 - c3_accuracy: 0.9447 - c4_accuracy: 0.9540 - c5_accuracy: 0.9516 - val_loss: 0.6312 - val_c1_loss: 0.1120 - val_c2_loss: 0.1367 - val_c3_loss: 0.1087 - val_c4_loss: 0.1599 - val_c5_loss: 0.1139 - val_c1_accuracy: 0.9798 - val_c2_accuracy: 0.9705 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9705 - val_c5_accuracy: 0.9708\n",
            "Epoch 219/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6674 - c1_loss: 0.1282 - c2_loss: 0.1315 - c3_loss: 0.1264 - c4_loss: 0.1401 - c5_loss: 0.1413 - c1_accuracy: 0.9575 - c2_accuracy: 0.9578 - c3_accuracy: 0.9587 - c4_accuracy: 0.9553 - c5_accuracy: 0.9560 - val_loss: 0.5891 - val_c1_loss: 0.1101 - val_c2_loss: 0.1354 - val_c3_loss: 0.0982 - val_c4_loss: 0.1397 - val_c5_loss: 0.1057 - val_c1_accuracy: 0.9811 - val_c2_accuracy: 0.9719 - val_c3_accuracy: 0.9770 - val_c4_accuracy: 0.9763 - val_c5_accuracy: 0.9732\n",
            "Epoch 220/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.6017 - c1_loss: 0.1219 - c2_loss: 0.1197 - c3_loss: 0.1274 - c4_loss: 0.1178 - c5_loss: 0.1148 - c1_accuracy: 0.9602 - c2_accuracy: 0.9613 - c3_accuracy: 0.9588 - c4_accuracy: 0.9604 - c5_accuracy: 0.9584 - val_loss: 0.5897 - val_c1_loss: 0.1131 - val_c2_loss: 0.1312 - val_c3_loss: 0.1029 - val_c4_loss: 0.1338 - val_c5_loss: 0.1087 - val_c1_accuracy: 0.9777 - val_c2_accuracy: 0.9719 - val_c3_accuracy: 0.9756 - val_c4_accuracy: 0.9760 - val_c5_accuracy: 0.9715\n",
            "Epoch 221/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.5389 - c1_loss: 0.1052 - c2_loss: 0.1095 - c3_loss: 0.1061 - c4_loss: 0.1128 - c5_loss: 0.1053 - c1_accuracy: 0.9674 - c2_accuracy: 0.9627 - c3_accuracy: 0.9638 - c4_accuracy: 0.9640 - c5_accuracy: 0.9666 - val_loss: 0.5919 - val_c1_loss: 0.1084 - val_c2_loss: 0.1345 - val_c3_loss: 0.1027 - val_c4_loss: 0.1372 - val_c5_loss: 0.1090 - val_c1_accuracy: 0.9794 - val_c2_accuracy: 0.9712 - val_c3_accuracy: 0.9750 - val_c4_accuracy: 0.9739 - val_c5_accuracy: 0.9729\n",
            "Epoch 222/500\n",
            "69/69 [==============================] - 2s 30ms/step - loss: 0.6544 - c1_loss: 0.1281 - c2_loss: 0.1306 - c3_loss: 0.1233 - c4_loss: 0.1512 - c5_loss: 0.1213 - c1_accuracy: 0.9575 - c2_accuracy: 0.9577 - c3_accuracy: 0.9562 - c4_accuracy: 0.9510 - c5_accuracy: 0.9596 - val_loss: 0.5844 - val_c1_loss: 0.1041 - val_c2_loss: 0.1304 - val_c3_loss: 0.1044 - val_c4_loss: 0.1426 - val_c5_loss: 0.1030 - val_c1_accuracy: 0.9811 - val_c2_accuracy: 0.9698 - val_c3_accuracy: 0.9746 - val_c4_accuracy: 0.9739 - val_c5_accuracy: 0.9736\n",
            "Epoch 223/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7444 - c1_loss: 0.1407 - c2_loss: 0.1623 - c3_loss: 0.1529 - c4_loss: 0.1382 - c5_loss: 0.1502 - c1_accuracy: 0.9541 - c2_accuracy: 0.9481 - c3_accuracy: 0.9518 - c4_accuracy: 0.9535 - c5_accuracy: 0.9479 - val_loss: 0.5979 - val_c1_loss: 0.1093 - val_c2_loss: 0.1276 - val_c3_loss: 0.1103 - val_c4_loss: 0.1507 - val_c5_loss: 0.1000 - val_c1_accuracy: 0.9798 - val_c2_accuracy: 0.9729 - val_c3_accuracy: 0.9729 - val_c4_accuracy: 0.9753 - val_c5_accuracy: 0.9753\n",
            "Epoch 224/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.7770 - c1_loss: 0.1562 - c2_loss: 0.1325 - c3_loss: 0.1516 - c4_loss: 0.1669 - c5_loss: 0.1698 - c1_accuracy: 0.9491 - c2_accuracy: 0.9575 - c3_accuracy: 0.9493 - c4_accuracy: 0.9460 - c5_accuracy: 0.9481 - val_loss: 0.6024 - val_c1_loss: 0.1078 - val_c2_loss: 0.1285 - val_c3_loss: 0.1066 - val_c4_loss: 0.1490 - val_c5_loss: 0.1105 - val_c1_accuracy: 0.9784 - val_c2_accuracy: 0.9698 - val_c3_accuracy: 0.9715 - val_c4_accuracy: 0.9729 - val_c5_accuracy: 0.9732\n",
            "Epoch 225/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6917 - c1_loss: 0.1368 - c2_loss: 0.1421 - c3_loss: 0.1353 - c4_loss: 0.1332 - c5_loss: 0.1443 - c1_accuracy: 0.9544 - c2_accuracy: 0.9521 - c3_accuracy: 0.9550 - c4_accuracy: 0.9556 - c5_accuracy: 0.9527 - val_loss: 0.6015 - val_c1_loss: 0.1075 - val_c2_loss: 0.1329 - val_c3_loss: 0.1106 - val_c4_loss: 0.1466 - val_c5_loss: 0.1038 - val_c1_accuracy: 0.9791 - val_c2_accuracy: 0.9712 - val_c3_accuracy: 0.9743 - val_c4_accuracy: 0.9719 - val_c5_accuracy: 0.9743\n",
            "Epoch 226/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6811 - c1_loss: 0.1377 - c2_loss: 0.1430 - c3_loss: 0.1260 - c4_loss: 0.1401 - c5_loss: 0.1343 - c1_accuracy: 0.9557 - c2_accuracy: 0.9535 - c3_accuracy: 0.9562 - c4_accuracy: 0.9541 - c5_accuracy: 0.9552 - val_loss: 0.5964 - val_c1_loss: 0.1065 - val_c2_loss: 0.1339 - val_c3_loss: 0.1066 - val_c4_loss: 0.1418 - val_c5_loss: 0.1075 - val_c1_accuracy: 0.9780 - val_c2_accuracy: 0.9684 - val_c3_accuracy: 0.9763 - val_c4_accuracy: 0.9750 - val_c5_accuracy: 0.9763\n",
            "Epoch 227/500\n",
            "69/69 [==============================] - 2s 31ms/step - loss: 0.6694 - c1_loss: 0.1401 - c2_loss: 0.1363 - c3_loss: 0.1268 - c4_loss: 0.1419 - c5_loss: 0.1243 - c1_accuracy: 0.9532 - c2_accuracy: 0.9554 - c3_accuracy: 0.9566 - c4_accuracy: 0.9541 - c5_accuracy: 0.9606 - val_loss: 0.5942 - val_c1_loss: 0.1039 - val_c2_loss: 0.1325 - val_c3_loss: 0.0986 - val_c4_loss: 0.1507 - val_c5_loss: 0.1085 - val_c1_accuracy: 0.9794 - val_c2_accuracy: 0.9702 - val_c3_accuracy: 0.9760 - val_c4_accuracy: 0.9712 - val_c5_accuracy: 0.9753\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Mm_2z1LBsOxo",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "904043a7-39f4-4007-8b03-914c3d9bdd1b"
      },
      "source": [
        "fig, axs = plt.subplots(5, 2, figsize=(15,20))\n",
        "fig.subplots_adjust(hspace=0.5, wspace=0.05)\n",
        "\n",
        "for i in range(1,6):\n",
        "    nacc  = 'c' + str(i) + '_accuracy' \n",
        "    nvacc = 'val_c' + str(i) + '_accuracy'\n",
        "    nlos  = 'c' + str(i) + '_loss'\n",
        "    nvlos = 'val_c' + str(i) + '_loss'\n",
        "\n",
        "    axs[i-1,0].plot(history.history[nacc], label='Acurácia de Treino')\n",
        "    axs[i-1,0].plot(history.history[nvacc], label = 'Acurácia de Validação')\n",
        "    axs[i-1,0].set_title('Digito'+str(i))\n",
        "    axs[i-1,0].legend(loc='lower right')\n",
        "\n",
        "    axs[i-1,1].plot(history.history[nlos], label='Perda de Treino')\n",
        "    axs[i-1,1].plot(history.history[nvlos], label = 'Perda de Validação')\n",
        "    axs[i-1,1].set_title('Digito'+str(i))\n",
        "    axs[i-1,1].legend(loc='upper right')\n"
      ],
      "execution_count": 82,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x1440 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2cUdOc3PsOrV",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        },
        "outputId": "aa96b351-7c4d-42a3-b10d-3754c8fc012d"
      },
      "source": [
        "# salva os dados para uso posterior\n",
        "data_dir = '/content/drive/My Drive/Colab Notebooks/Data_Colab/Captcha_Break/'\n",
        "if not os.path.exists(data_dir):\n",
        "    os.makedirs(data_dir)\n",
        "\n",
        "with open(data_dir+'LabelBinarizer' +'.pickle', 'wb') as f:\n",
        "    pickle.dump(lb, f, pickle.HIGHEST_PROTOCOL)\n",
        "\n",
        "with open(data_dir+'Best_ModelCaptcha' +'.pickle', 'wb') as f:\n",
        "    pickle.dump(model, f, pickle.HIGHEST_PROTOCOL)"
      ],
      "execution_count": 240,
      "outputs": [
        {
          "output_type": "error",
          "ename": "TypeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-240-6acb7894c8ae>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      8\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      9\u001b[0m \u001b[0;32mwith\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata_dir\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;34m'Best_ModelCaptcha'\u001b[0m \u001b[0;34m+\u001b[0m\u001b[0;34m'.pickle'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'wb'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 10\u001b[0;31m     \u001b[0mpickle\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdump\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpickle\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mHIGHEST_PROTOCOL\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;31mTypeError\u001b[0m: can't pickle _thread.RLock objects"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bpRvZBf-Egrv",
        "colab_type": "text"
      },
      "source": [
        "#### Referência de [Salvar modelos do Keras](https://www.tensorflow.org/tutorials/keras/save_and_load?hl=pt-br)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uQkRfReYERr_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model_dir = '/content/drive/My Drive/Colab Notebooks/Data_Colab/Captcha_Break/'\n",
        "model.save(model_dir+'Best_CaptchaModel.h5') "
      ],
      "execution_count": 242,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cL3wdanfdHYN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "8cb2ef7f-212e-4018-8978-2547d8b98f3a"
      },
      "source": [
        "# salva os dados para uso posterior\n",
        "model_dir = '/content/drive/My Drive/Colab Notebooks/Data_Colab/Captcha_Break/Model'\n",
        "if not os.path.exists(model_dir):\n",
        "    os.makedirs(model_dir)\n",
        "\n",
        "model.save(model_dir)"
      ],
      "execution_count": 90,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Assets written to: /content/drive/My Drive/Colab Notebooks/Data_Colab/Captcha_Break/Model/assets\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_b3JZYWFmFDF",
        "colab_type": "text"
      },
      "source": [
        "## Classificação nos Novos Captchas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SU8lRV7m1AJj",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "b2304736-0060-4ebb-b2c4-0bae0741ed4e"
      },
      "source": [
        "# salva os dados para uso posterior\n",
        "data_dir = '/content/drive/My Drive/Colab Notebooks/Data_Colab/Captcha_Break/'\n",
        "\n",
        "with open(data_dir+'SDBCclean'+'_Data' +'.pickle', 'rb') as f:\n",
        "    data = pickle.load(f)\n",
        "\n",
        "with open(data_dir+'SDBCclean'+'_Labels'+'.pickle', 'rb') as f:\n",
        "    labels = pickle.load(f)\n",
        "\n",
        "data.shape, labels.shape"
      ],
      "execution_count": 91,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((547, 60, 200), (547,))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 91
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zGEp7jZ1pmji",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 101
        },
        "outputId": "3b7b917b-5d48-45ab-f5db-038b6cc972c8"
      },
      "source": [
        "letras_y = np.array([list(x) for x in labels])\n",
        "np.unique(letras_y,return_counts=True)"
      ],
      "execution_count": 94,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(array(['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M',\n",
              "        'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z'],\n",
              "       dtype='<U1'),\n",
              " array([101, 103, 103, 112, 110,  97, 100, 122,  99, 105, 107, 121, 104,\n",
              "        114, 113,  91, 100, 108, 102,  95, 109, 104, 104, 101, 109, 101]))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 94
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "iSpMki2apnAh",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "yI1oBgMBpnAk",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "bdc64c66-87f3-4505-f40b-65023669f66f"
      },
      "source": [
        "lab = []\n",
        "for ltr in letras_y:\n",
        "    t = lb.transform(ltr)\n",
        "    lab.append(t)\n",
        "    \n",
        "labels_lb = np.stack(lab, axis=1)\n",
        "v_y = []\n",
        "for x in labels_lb:\n",
        "    v_y.append(x)\n",
        "\n",
        "\n",
        "x_data = np.expand_dims(data, axis = 3)\n",
        "x_data.shape"
      ],
      "execution_count": 102,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(547, 60, 200, 1)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 102
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6sBi15vKrUFa",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.evaluate(x_data, v_y)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FEeIFyQIoixi",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "label_pred = []\n",
        "labels_prob = []\n",
        "\n",
        "predict = model.predict(x_data, batch_size=1)\n",
        "\n",
        "for img in predict:\n",
        "    pred = img.max(1)\n",
        "    cls_pred = lb.inverse_transform(img)\n",
        "    label_pred.append(cls_pred)\n",
        "    labels_prob.append(pred)"
      ],
      "execution_count": 127,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eCvQm7C_tTCR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "captcha_pred = np.stack(label_pred, axis=1)\n",
        "captcha_pred = np.array([x1+x2+x3+x4+x5 for x1,x2,x3,x4,x5 in captcha_pred])"
      ],
      "execution_count": 134,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ezzqxAFh3LqY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 67
        },
        "outputId": "d02d2b15-7a93-41bb-b59e-772af3e13c3e"
      },
      "source": [
        "# val_erro = lb.inverse_transform(predict) != lb.inverse_transform(labels_lb)\n",
        "val_erro = labels != captcha_pred\n",
        "print('Quantidade total de erros: ', sum(val_erro))\n",
        "print('Quantidade total de dados: ', len(val_erro))\n",
        "print('Percentual de erros: {:.2%}'.format(sum(val_erro)/len(val_erro)) )"
      ],
      "execution_count": 137,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Quantidade total de erros:  31\n",
            "Quantidade total de dados:  547\n",
            "Percentual de erros: 5.67%\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xFapV_017I8U",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "captcha_pred = np.array([x1+x2+x3+x4+x5 for x1,x2,x3,x4,x5 in cls_pred])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5l-rQz-gwLmQ",
        "colab_type": "text"
      },
      "source": [
        "#### Alguns dos Erros"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xx1GKtjcvFxT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 169
        },
        "outputId": "3e986c7f-2a93-499f-eecc-f90144fc3672"
      },
      "source": [
        "j = 0\n",
        "i = np.array(np.where(val_erro))[0,j]\n",
        "\n",
        "print('Label Verdadeiro:' + labels[i] + '\\n' + 'Label Predito:   ' + captcha_pred[i])\n",
        "\n",
        "plt.imshow(data[i], cmap='binary')\n",
        "plt.xticks([]), plt.yticks([])\n",
        "plt.show()"
      ],
      "execution_count": 142,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Label Verdadeiro:LEQOF\n",
            "Label Predito:   IEQOF\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HnfAS7afvFr1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 169
        },
        "outputId": "bbd92e16-cbf3-4275-9f2d-f0f045c12d57"
      },
      "source": [
        "j = 5\n",
        "i = np.array(np.where(val_erro))[0,j]\n",
        "\n",
        "print('Label Verdadeiro:' + labels[i] + '\\n' + 'Label Predito:   ' + captcha_pred[i])\n",
        "\n",
        "plt.imshow(data[i], cmap='binary')\n",
        "plt.xticks([]), plt.yticks([])\n",
        "plt.show()"
      ],
      "execution_count": 143,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Label Verdadeiro:DWDJS\n",
            "Label Predito:   DWDLS\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1IiZf_t8wBuQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 169
        },
        "outputId": "75f7d7e6-5833-4967-c738-e7a7ffdd04fc"
      },
      "source": [
        "j = 30\n",
        "i = np.array(np.where(val_erro))[0,j]\n",
        "\n",
        "print('Label Verdadeiro:' + labels[i] + '\\n' + 'Label Predito:   ' + captcha_pred[i])\n",
        "plt.imshow(data[i], cmap='binary')\n",
        "plt.xticks([]), plt.yticks([])\n",
        "plt.show()"
      ],
      "execution_count": 144,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Label Verdadeiro:QEXKK\n",
            "Label Predito:   GEXKK\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0LkGLIIh2pt7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn import metrics\n",
        "\n",
        "print(metrics.confusion_matrix(y_true, y_pred))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7o1AP6wO3n9P",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q4hZZI-C2pe2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.evaluate(x_data, v_y)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kn959-wV3IIy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "j = 30\n",
        "i = np.array(np.where(val_erro))[0,j]\n",
        "labels[val_erro]\n",
        "\n",
        "print('Label Verdadeiro:' + labels[i] + '\\n' + 'Label Predito:   ' + captcha_pred[i])\n",
        "plt.imshow(data[i], cmap='binary')\n",
        "plt.xticks([]), plt.yticks([])\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mIjxTZmJ3E5s",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "label_pred = []\n",
        "labels_prob = []\n",
        "\n",
        "predict = model.predict(x_data, batch_size=1)\n",
        "\n",
        "for img in predict:\n",
        "    pred = img.max(1)\n",
        "    cls_pred = lb.inverse_transform(img)\n",
        "    label_pred.append(cls_pred)\n",
        "    labels_prob.append(pred)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5qhRXDLa-izo",
        "colab_type": "text"
      },
      "source": [
        "### Tabelas de Confusão de Classificações Corretas e Erradas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iCMUoi6w87Rd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import string  \n",
        "import seaborn as sn"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fN-ZFwm78_Ef",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 443
        },
        "outputId": "ddaca6f5-d3c1-4a0c-b44d-69d865fcab13"
      },
      "source": [
        "pred_cer = np.stack(predict, axis=1)\n",
        "pred_cer = pred_cer.reshape((2735,26))\n",
        "prev = np.stack([list(x) for x in captcha_pred], axis=0).flatten()\n",
        "\n",
        "soma = []\n",
        "for i in string.ascii_uppercase:\n",
        "    letr = (pred_cer[prev == i]).sum(0)\n",
        "    soma.append(letr)\n",
        "soma = np.array(soma)\n",
        "\n",
        "\n",
        "df = pd.DataFrame(np.round(soma, 2), \n",
        "                  index = [i for i in string.ascii_uppercase],\n",
        "                  columns = [i for i in string.ascii_uppercase])\n",
        "plt.figure(figsize = (10,7))\n",
        "sn.heatmap(df)"
      ],
      "execution_count": 218,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f5b2d2be278>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 218
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x504 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4ru_pu6P4M_9",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 443
        },
        "outputId": "e6352503-8322-4296-b799-179aec8fdeaf"
      },
      "source": [
        "pred_err = np.stack(predict, axis=1)[val_erro]\n",
        "pred_err = pred_err.reshape((155,26))\n",
        "prev = np.stack([list(x) for x in captcha_pred[val_erro]], axis=0).flatten()\n",
        "labs = np.stack([list(x) for x in labels[val_erro]], axis=0).flatten()\n",
        "\n",
        "val_err_ltr = prev != labs\n",
        "\n",
        "pred_err = pred_err[val_err_ltr]\n",
        "prev = prev[val_err_ltr]\n",
        "\n",
        "soma = []\n",
        "for i in string.ascii_uppercase:\n",
        "    letr = (pred_err[prev == i]).sum(0)\n",
        "    soma.append(letr)\n",
        "soma = np.array(soma)\n",
        "\n",
        "df = pd.DataFrame(np.round(soma, 2), \n",
        "                  index = [i for i in string.ascii_uppercase],\n",
        "                  columns = [i for i in string.ascii_uppercase])\n",
        "plt.figure(figsize = (10,7))\n",
        "sn.heatmap(df)"
      ],
      "execution_count": 239,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f5b3079c898>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 239
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x504 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OuyXrhw73Eyn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}