{
  "cells": [
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "tn7RUWXiJg_Z"
      },
      "source": [
        "# ResNet网络可视化\n",
        "\n",
        "**目标：**\n",
        "\n",
        "\n",
        "1.   学习神经网络在训练过程中，隐含层的特征变化机制\n",
        "\n",
        "代码中的学习器为残差网络(residual network, res)\n",
        "\n",
        "**参考：**\n",
        "\n",
        "葛家驿,杨乃森,唐宏,徐朋磊,纪超.端到端的梯度提升网络分类过程可视化[J].信号处理,2022,38(02):355-366.DOI:10.16798/j.issn.1003-0530.2022.02.015."
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## 研究准备"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XWAfbFHIJW3x"
      },
      "source": [
        "### **环境配置**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1e9oxni89ilt"
      },
      "outputs": [],
      "source": [
        "'''\n",
        "残差网络分类过程可视化\n",
        "'''\n",
        "import numpy as np\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import backend as K\n",
        "from tensorflow.keras import layers\n",
        "from tensorflow.keras.callbacks import CSVLogger\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "from matplotlib.colors import ListedColormap\n",
        "from matplotlib import cm\n",
        "import matplotlib as mpl\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "from sklearn.datasets import make_circles #同心圆数据\n",
        "from sklearn.model_selection import train_test_split\n",
        "import sys\n",
        "sys.setrecursionlimit(500000)\n",
        "import imageio\n",
        "%pylab inline "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rzOHrosxJYin"
      },
      "source": [
        "### **设置模拟数据**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "9mc1hvBT-J3e"
      },
      "outputs": [],
      "source": [
        "n_samples = 1000 #样本点数\n",
        "X, y = make_circles(n_samples=1000,factor=.4,noise=.06,random_state=0) #生成同心圆数据\n",
        "test_size = 0.5\n",
        "\n",
        "#划分训练集、测试集\n",
        "X_train, X_test, Y_train, Y_test = train_test_split(X, y, test_size=test_size, random_state=2)\n",
        "\n",
        "c,r = np.mgrid[[slice(X.min()- .2,X.max() + .2,50j)]*2]\n",
        "p = np.c_[c.flat,r.flat]\n",
        "\n",
        "ss = StandardScaler().fit(X_train)\n",
        "X = ss.transform(X)\n",
        "p = ss.transform(p)\n",
        "X_train = ss.transform(X_train)\n",
        "X_test = ss.transform(X_test)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "s6PVC2h2Jabo"
      },
      "source": [
        "### **配置绘图环境**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 600
        },
        "id": "TeBvTz76-QdY",
        "outputId": "000f425f-4895-4fec-9584-4547e866e2d5"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 648x216 with 3 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 504x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "#设置画布大小和颜色\n",
        "fig = plt.figure(figsize = (9,3))\n",
        "top = cm.get_cmap('Oranges_r', 512)\n",
        "bottom = cm.get_cmap('Blues', 512)\n",
        "newcolors = np.vstack((top(np.linspace(0.55, 1, 512)),\n",
        "                       bottom(np.linspace(0, 0.75, 512))))\n",
        "cm_bright = ListedColormap(newcolors, name='OrangeBlue')\n",
        "\n",
        "#训练数据可视化\n",
        "plt.subplot(121)\n",
        "m1 = plt.scatter(*X_train.T,c = Y_train,cmap = cm_bright,edgecolors='white',s = 20,linewidths = 0.5)\n",
        "plt.title(f'train data ({int(n_samples*(1-test_size))} points)')\n",
        "plt.axis('equal')\n",
        "\n",
        "#测试数据可视化\n",
        "plt.subplot(122)\n",
        "m2 = plt.scatter(*X_test.T,c = Y_test,cmap = cm_bright,edgecolors='white',s = 20,linewidths = 0.5);\n",
        "plt.title(f'test data ({int(n_samples*test_size)} points)')\n",
        "plt.axis('equal')\n",
        "ax = fig.get_axes()\n",
        "plt.colorbar(ax = ax)\n",
        "#plt.savefig(f'data_{n_samples}_points.png')\n",
        "#plt.savefig(f'data_{n_samples}_points.pdf')\n",
        "plt.show()\n",
        "\n",
        "#全部数据可视化\n",
        "fig = plt.figure(figsize = (7,6))\n",
        "plt.scatter(*X.T,c = y,cmap = cm_bright,edgecolors='white',s = 20,linewidths = 0.5)\n",
        "plt.title(f'Raw data ({n_samples} points)')\n",
        "plt.axis('equal')\n",
        "#plt.savefig(f'Raw data ({n_samples} points)')\n",
        "#plt.savefig(f'Raw data ({n_samples} points).pdf')\n",
        "plt.axis('equal')\n",
        "#plt.colorbar(ax = ax)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "L99HdhJ4Jcji"
      },
      "source": [
        "### **配置数据与损失函数**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "jTZei6w3-VOf"
      },
      "outputs": [],
      "source": [
        "num_classes=2 #设置类别数\n",
        "y_train=keras.utils.to_categorical(Y_train,num_classes) #类别标签转换为onehot编码\n",
        "y_test=keras.utils.to_categorical(Y_test,num_classes)\n",
        "#定义损失曲线绘制函数\n",
        "\n",
        "def plot_loss_accuracy(history, title_text, file_name):\n",
        "\n",
        "    fig, ax1 = plt.subplots()\n",
        "    ax2 = ax1.twinx()\n",
        "    ax1.set_xlabel(\"Epoch\")\n",
        "    ax1.set_ylabel(\"Accuracy\")\n",
        "    ax2.set_ylabel(\"Loss\")\n",
        "    \n",
        "    #ax1.set_ylim(-0.01,1.01)\n",
        "    ax1.plot(history.epoch,\n",
        "             history.history['accuracy'],\n",
        "             label=\"Training Accuracy\")\n",
        "    ax1.plot(history.epoch,\n",
        "             history.history['val_accuracy'],\n",
        "             linestyle='--',\n",
        "             label=\"Test Accuracy\")\n",
        "    \n",
        "    #ax2.set_ylim(-0.01,1.01)\n",
        "    ax2._get_lines.prop_cycler = ax1._get_lines.prop_cycler \n",
        "    \n",
        "    ax2.plot(history.epoch, \n",
        "             history.history['loss'], \n",
        "             label=\"Training Loss\")\n",
        "    ax2.plot(history.epoch,\n",
        "             history.history['val_loss'],\n",
        "             linestyle='--',\n",
        "             label=\"Test Loss\")\n",
        "\n",
        "    ax1.legend()\n",
        "    ax2.legend()\n",
        "    plt.suptitle(title_text)\n",
        "    plt.savefig(file_name)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "t_CR5Kx3Jgrl"
      },
      "source": [
        "### **配置残差模型**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "eeo-G9f6-fA6"
      },
      "outputs": [],
      "source": [
        "#定义残差块\n",
        "\n",
        "def ResMLP_Block(number_of_res_blocks,\n",
        "                 number_dense_of_res_block,\n",
        "                 number_neurons_of_dense,\n",
        "                 inputs):\n",
        " \n",
        "    x = inputs\n",
        "    for i in range(number_of_res_blocks):\n",
        "        inputs = x\n",
        "        for j in range(number_dense_of_res_block):\n",
        "            x = layers.Dense(number_neurons_of_dense,\n",
        "                             activation=tf.nn.relu,\n",
        "                             name = f'{i}th-resBlock_{j}th-hidden')(x)\n",
        "        x = layers.Dense(2, name = f'{i}th-resBlock_linear')(x)\n",
        "        x = layers.Add(name = f'{i}th-resBlock_Add')([x,inputs]) \n",
        "        \n",
        "        \n",
        "    return x\n",
        "#构建残差网络\n",
        "\n",
        "def ResMLP_Model(number_of_res_blocks,\n",
        "                 number_dense_of_res_block,\n",
        "                 number_neurons_of_dense,\n",
        "                 num_of_classes=2):\n",
        "    inputs = keras.Input(shape=(2, ))\n",
        "    x = ResMLP_Block(number_of_res_blocks,\n",
        "                     number_dense_of_res_block,\n",
        "                     number_neurons_of_dense,\n",
        "                     inputs)\n",
        "    outputs = layers.Dense(2,activation='softmax')(x)\n",
        "    res_model = keras.Model(inputs, outputs)\n",
        "    res_model.compile(loss=keras.losses.categorical_crossentropy,\n",
        "                      optimizer=keras.optimizers.Adam(lr=3e-4),\n",
        "                      metrics=['accuracy'])\n",
        "    #csv_logger = CSVLogger(f'training_res_model.log')\n",
        "    history = res_model.fit(X_train,\n",
        "                            y_train,\n",
        "                            batch_size=batch_size,\n",
        "                            epochs=epochs,\n",
        "                            verbose=2,\n",
        "                            validation_data=(X_test, y_test))\n",
        "    return res_model,history"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## 进行实验"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6YZA52E5KXvp"
      },
      "source": [
        "### **设置训练参数**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "R2dFCSwO-q7g",
        "outputId": "12190b7d-e40a-4efa-d3b7-0bb749447fe4"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/keras/optimizers/optimizer_v2/adam.py:110: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.\n",
            "  super(Adam, self).__init__(name, **kwargs)\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[1;30;43m流式输出内容被截断，只能显示最后 5000 行内容。\u001b[0m\n",
            "Epoch 7501/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4890e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7502/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4913e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7503/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4866e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 7504/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4890e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7505/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4890e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7506/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4890e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7507/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4866e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7508/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4890e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7509/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4842e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 7510/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4866e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7511/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4866e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7512/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4818e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7513/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4842e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7514/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4866e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7515/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4842e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7516/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4842e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7517/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4818e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7518/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4818e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7519/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4794e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7520/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4842e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7521/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4818e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7522/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4794e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7523/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4842e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7524/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4866e-07 - val_accuracy: 1.0000 - 73ms/epoch - 5ms/step\n",
            "Epoch 7525/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4770e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7526/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4794e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7527/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4794e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7528/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4794e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7529/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4818e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7530/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4747e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7531/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7532/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7533/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4723e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7534/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4747e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7535/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4770e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7536/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4723e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7537/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4699e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7538/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7539/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7540/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7541/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4580e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7542/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4604e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7543/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7544/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4580e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7545/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7546/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7547/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7548/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7549/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7550/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7551/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7552/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4723e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7553/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7554/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7555/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7556/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4699e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7557/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4699e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7558/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4699e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7559/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7560/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4699e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7561/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4699e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7562/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4699e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7563/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7564/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7565/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7566/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7567/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7568/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4699e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7569/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7570/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7571/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7572/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4580e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7573/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4556e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7574/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7575/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4580e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7576/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4604e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7577/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4580e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7578/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4532e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7579/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7580/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7581/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4604e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7582/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4556e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7583/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4604e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7584/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7585/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4651e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7586/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7587/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7588/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4675e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7589/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7590/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4556e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7591/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4604e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7592/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4556e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7593/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4580e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7594/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4508e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7595/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4580e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7596/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4556e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7597/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4508e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7598/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4484e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7599/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4532e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7600/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4556e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7601/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4604e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7602/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4580e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7603/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4627e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7604/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4604e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7605/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4580e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7606/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4508e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7607/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4556e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7608/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4604e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7609/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4532e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 7610/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4508e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 7611/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4508e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7612/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7613/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4484e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7614/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4413e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7615/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7616/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4484e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7617/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7618/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4484e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7619/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7620/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7621/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7622/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7623/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7624/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4437e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7625/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4484e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7626/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7627/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4437e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7628/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4508e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7629/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4484e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 7630/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4508e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7631/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4437e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7632/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4413e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7633/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4508e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7634/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 73ms/epoch - 5ms/step\n",
            "Epoch 7635/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4437e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7636/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7637/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7638/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4437e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7639/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4461e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7640/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4413e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7641/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4389e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7642/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4365e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7643/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4389e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7644/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4365e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7645/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4341e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7646/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4389e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7647/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4365e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7648/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4341e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7649/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4341e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7650/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4318e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7651/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4341e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7652/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4318e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7653/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4318e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7654/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4222e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7655/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4318e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7656/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4222e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7657/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4270e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7658/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4294e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7659/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4246e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7660/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4198e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7661/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4198e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7662/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4222e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7663/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4270e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7664/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4222e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7665/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4222e-07 - val_accuracy: 1.0000 - 72ms/epoch - 4ms/step\n",
            "Epoch 7666/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4198e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7667/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4246e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 7668/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4246e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7669/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4246e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7670/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4222e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7671/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4198e-07 - val_accuracy: 1.0000 - 96ms/epoch - 6ms/step\n",
            "Epoch 7672/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4198e-07 - val_accuracy: 1.0000 - 80ms/epoch - 5ms/step\n",
            "Epoch 7673/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4174e-07 - val_accuracy: 1.0000 - 108ms/epoch - 7ms/step\n",
            "Epoch 7674/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4294e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7675/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4222e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7676/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4198e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 7677/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4246e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 7678/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4198e-07 - val_accuracy: 1.0000 - 108ms/epoch - 7ms/step\n",
            "Epoch 7679/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4270e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7680/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4246e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 7681/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4222e-07 - val_accuracy: 1.0000 - 111ms/epoch - 7ms/step\n",
            "Epoch 7682/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4222e-07 - val_accuracy: 1.0000 - 86ms/epoch - 5ms/step\n",
            "Epoch 7683/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4246e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 7684/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4127e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 7685/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4198e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7686/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4151e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7687/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4151e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7688/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4127e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7689/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4151e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7690/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4127e-07 - val_accuracy: 1.0000 - 72ms/epoch - 4ms/step\n",
            "Epoch 7691/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4174e-07 - val_accuracy: 1.0000 - 90ms/epoch - 6ms/step\n",
            "Epoch 7692/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4127e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 7693/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4151e-07 - val_accuracy: 1.0000 - 81ms/epoch - 5ms/step\n",
            "Epoch 7694/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 78ms/epoch - 5ms/step\n",
            "Epoch 7695/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4127e-07 - val_accuracy: 1.0000 - 75ms/epoch - 5ms/step\n",
            "Epoch 7696/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7697/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 74ms/epoch - 5ms/step\n",
            "Epoch 7698/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4079e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7699/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7700/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4079e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7701/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7702/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7703/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7704/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4079e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 7705/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7706/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4079e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7707/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7708/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4079e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7709/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7710/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7711/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7712/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4103e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7713/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4079e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7714/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4079e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7715/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4079e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7716/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4103e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7717/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4103e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7718/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4103e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7719/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4103e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7720/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 7721/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 7722/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7723/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 7724/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7725/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7726/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7727/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7728/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7729/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7730/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7731/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7732/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7733/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7734/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7735/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7736/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7737/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7738/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4079e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7739/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7740/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7741/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7742/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7743/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7744/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7745/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7746/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7747/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7748/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7749/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7750/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4055e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7751/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7752/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4031e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7753/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7754/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7755/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 7756/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7757/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7758/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7759/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7760/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 7761/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7762/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7763/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7764/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7765/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7766/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 7767/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7768/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7769/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7770/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7771/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7772/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7773/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7774/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7775/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7776/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7777/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.4008e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7778/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7779/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7780/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7781/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3912e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7782/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7783/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7784/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7785/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7786/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7787/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7788/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3912e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7789/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7790/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7791/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7792/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7793/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7794/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7795/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3912e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7796/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7797/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7798/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7799/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 77ms/epoch - 5ms/step\n",
            "Epoch 7800/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3912e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7801/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3912e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7802/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7803/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7804/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7805/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7806/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7807/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3936e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7808/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3984e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7809/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3960e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7810/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3912e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7811/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3912e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7812/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3888e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7813/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3912e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7814/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3865e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7815/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3912e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7816/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3888e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7817/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3841e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7818/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3793e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7819/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3865e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7820/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3841e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7821/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3841e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 7822/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3865e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7823/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3841e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7824/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3841e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7825/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3841e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7826/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3865e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7827/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3841e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7828/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3841e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7829/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3888e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7830/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3865e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7831/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3841e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 7832/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3865e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7833/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3865e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7834/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3817e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7835/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3865e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7836/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7837/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7838/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3817e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 7839/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3865e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7840/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3817e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7841/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3841e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7842/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3817e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7843/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3817e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7844/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3793e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7845/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3793e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7846/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 72ms/epoch - 4ms/step\n",
            "Epoch 7847/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7848/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7849/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7850/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7851/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7852/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3793e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7853/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3698e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7854/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3793e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7855/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3793e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7856/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7857/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7858/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3817e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7859/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7860/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7861/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7862/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 77ms/epoch - 5ms/step\n",
            "Epoch 7863/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7864/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7865/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7866/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3793e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7867/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7868/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7869/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7870/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7871/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3722e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7872/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7873/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3769e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7874/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3722e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7875/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3722e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7876/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7877/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3722e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7878/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3698e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7879/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3698e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7880/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3722e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7881/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7882/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3698e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7883/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7884/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7885/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3722e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 7886/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3698e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7887/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3745e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7888/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3722e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7889/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 7890/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3698e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7891/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3722e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7892/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3674e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7893/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3722e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7894/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3674e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7895/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3674e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7896/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7897/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7898/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3674e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7899/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3602e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7900/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3674e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7901/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3578e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7902/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3602e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7903/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3626e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 7904/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3602e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7905/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3602e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7906/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3578e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 7907/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7908/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3578e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7909/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3602e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7910/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3602e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 7911/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7912/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3674e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7913/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3626e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7914/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7915/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7916/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7917/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7918/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3626e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7919/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3626e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7920/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3674e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7921/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3626e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7922/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7923/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3674e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7924/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3674e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7925/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 7926/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3602e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7927/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3650e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7928/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3674e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7929/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3602e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7930/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3626e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7931/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3602e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7932/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3578e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 7933/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3555e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7934/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3507e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7935/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7936/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3555e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7937/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7938/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7939/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7940/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7941/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3555e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7942/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3507e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7943/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3578e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7944/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3555e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7945/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7946/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3507e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7947/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7948/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7949/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3507e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7950/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3459e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 7951/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7952/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3531e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7953/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3459e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7954/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3507e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7955/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3483e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 7956/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3459e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 7957/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3459e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7958/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3483e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7959/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3483e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7960/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3483e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7961/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3435e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7962/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 7963/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7964/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7965/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3388e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7966/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3388e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7967/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3459e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7968/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7969/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3388e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7970/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3435e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7971/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3435e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7972/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3483e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7973/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3507e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7974/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3459e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7975/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3459e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 7976/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3435e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 7977/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3435e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7978/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 7979/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3388e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7980/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3364e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7981/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7982/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3364e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 7983/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3388e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7984/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3340e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7985/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3364e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 7986/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3340e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 7987/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3388e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 7988/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3364e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 7989/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3340e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 7990/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3316e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 7991/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3340e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7992/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7993/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3364e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 7994/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 7995/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 7996/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3364e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 7997/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3388e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 7998/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3364e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 7999/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8000/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3435e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8001/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3435e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8002/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3412e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 8003/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3364e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8004/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3316e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8005/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3316e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8006/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3340e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8007/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3340e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8008/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3364e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8009/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3316e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8010/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3292e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8011/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3316e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8012/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3269e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8013/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3269e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8014/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3245e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8015/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3245e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8016/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8017/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8018/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8019/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8020/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8021/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8022/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8023/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8024/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8025/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8026/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8027/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8028/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8029/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8030/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8031/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8032/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8033/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8034/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3269e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8035/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3245e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8036/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3269e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8037/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3269e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8038/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8039/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3269e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8040/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8041/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3245e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8042/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8043/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3269e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 8044/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8045/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8046/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8047/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8048/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8049/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8050/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8051/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8052/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8053/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3221e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8054/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8055/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8056/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8057/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8058/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8059/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8060/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8061/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8062/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8063/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 72ms/epoch - 4ms/step\n",
            "Epoch 8064/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8065/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8066/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8067/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8068/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3245e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8069/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8070/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8071/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8072/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8073/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3197e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8074/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8075/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8076/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8077/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8078/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 72ms/epoch - 4ms/step\n",
            "Epoch 8079/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8080/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 72ms/epoch - 5ms/step\n",
            "Epoch 8081/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8082/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8083/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3102e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8084/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8085/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8086/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8087/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3102e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8088/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8089/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8090/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8091/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8092/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8093/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3078e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8094/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3030e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8095/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8096/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3078e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8097/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8098/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3078e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8099/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8100/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8101/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8102/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3030e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8103/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3030e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8104/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8105/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8106/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3006e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8107/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3030e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8108/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3006e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8109/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8110/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3078e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8111/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3102e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8112/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8113/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8114/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8115/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3078e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8116/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8117/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8118/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8119/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3078e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8120/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8121/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8122/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8123/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3173e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8124/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3149e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8125/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8126/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3102e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8127/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8128/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8129/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3126e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8130/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3030e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 8131/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3078e-07 - val_accuracy: 1.0000 - 52ms/epoch - 3ms/step\n",
            "Epoch 8132/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8133/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3102e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8134/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3078e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8135/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8136/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8137/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3078e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8138/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8139/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3006e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8140/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3006e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8141/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2982e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8142/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3030e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8143/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3006e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8144/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3054e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8145/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2982e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8146/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3006e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8147/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2982e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8148/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2982e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8149/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3006e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8150/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.3006e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8151/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2959e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8152/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2982e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8153/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2911e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8154/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2982e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8155/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2887e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8156/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2935e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8157/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2982e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8158/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2935e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8159/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2935e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8160/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2935e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8161/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2959e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8162/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2959e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8163/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2982e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8164/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2911e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8165/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2911e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8166/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2911e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8167/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2911e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8168/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2863e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8169/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2887e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8170/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2839e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8171/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2911e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8172/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2863e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8173/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2863e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8174/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2911e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8175/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2839e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8176/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2816e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8177/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2816e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8178/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2863e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8179/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2839e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8180/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2768e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8181/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2768e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8182/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2768e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8183/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2744e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8184/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2792e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8185/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2792e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8186/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2744e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8187/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2768e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8188/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2720e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8189/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2768e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8190/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2744e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8191/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2768e-07 - val_accuracy: 1.0000 - 52ms/epoch - 3ms/step\n",
            "Epoch 8192/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2768e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8193/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2816e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8194/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2768e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8195/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2744e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8196/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2744e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8197/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2720e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8198/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2768e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8199/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2696e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8200/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2720e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8201/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2673e-07 - val_accuracy: 1.0000 - 81ms/epoch - 5ms/step\n",
            "Epoch 8202/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2649e-07 - val_accuracy: 1.0000 - 86ms/epoch - 5ms/step\n",
            "Epoch 8203/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2696e-07 - val_accuracy: 1.0000 - 77ms/epoch - 5ms/step\n",
            "Epoch 8204/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2696e-07 - val_accuracy: 1.0000 - 84ms/epoch - 5ms/step\n",
            "Epoch 8205/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2696e-07 - val_accuracy: 1.0000 - 113ms/epoch - 7ms/step\n",
            "Epoch 8206/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2744e-07 - val_accuracy: 1.0000 - 73ms/epoch - 5ms/step\n",
            "Epoch 8207/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2720e-07 - val_accuracy: 1.0000 - 76ms/epoch - 5ms/step\n",
            "Epoch 8208/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2720e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8209/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2720e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8210/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2673e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8211/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2696e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8212/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2696e-07 - val_accuracy: 1.0000 - 72ms/epoch - 4ms/step\n",
            "Epoch 8213/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2673e-07 - val_accuracy: 1.0000 - 107ms/epoch - 7ms/step\n",
            "Epoch 8214/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2696e-07 - val_accuracy: 1.0000 - 73ms/epoch - 5ms/step\n",
            "Epoch 8215/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2673e-07 - val_accuracy: 1.0000 - 106ms/epoch - 7ms/step\n",
            "Epoch 8216/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2673e-07 - val_accuracy: 1.0000 - 82ms/epoch - 5ms/step\n",
            "Epoch 8217/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2649e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8218/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 76ms/epoch - 5ms/step\n",
            "Epoch 8219/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2649e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8220/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 73ms/epoch - 5ms/step\n",
            "Epoch 8221/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2625e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8222/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2696e-07 - val_accuracy: 1.0000 - 78ms/epoch - 5ms/step\n",
            "Epoch 8223/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2625e-07 - val_accuracy: 1.0000 - 72ms/epoch - 4ms/step\n",
            "Epoch 8224/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2649e-07 - val_accuracy: 1.0000 - 77ms/epoch - 5ms/step\n",
            "Epoch 8225/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2649e-07 - val_accuracy: 1.0000 - 74ms/epoch - 5ms/step\n",
            "Epoch 8226/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2673e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8227/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2649e-07 - val_accuracy: 1.0000 - 105ms/epoch - 7ms/step\n",
            "Epoch 8228/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2649e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8229/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8230/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2625e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8231/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2625e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8232/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2577e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8233/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8234/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8235/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8236/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8237/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2577e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8238/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2577e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8239/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8240/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8241/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8242/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2553e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8243/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8244/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2601e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8245/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2530e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8246/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2553e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8247/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2506e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8248/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2506e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8249/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2506e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8250/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2458e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8251/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2458e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8252/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2434e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8253/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2458e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8254/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2553e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8255/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2458e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8256/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2458e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8257/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2434e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8258/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2458e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8259/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2434e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8260/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2387e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8261/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2363e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8262/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2363e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8263/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8264/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2387e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8265/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8266/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8267/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8268/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8269/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8270/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8271/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2363e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8272/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8273/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8274/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8275/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 106ms/epoch - 7ms/step\n",
            "Epoch 8276/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8277/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8278/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8279/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8280/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8281/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8282/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8283/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2387e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8284/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2363e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8285/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2363e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8286/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2387e-07 - val_accuracy: 1.0000 - 52ms/epoch - 3ms/step\n",
            "Epoch 8287/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8288/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8289/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8290/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8291/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8292/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2363e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8293/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8294/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8295/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8296/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8297/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8298/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8299/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8300/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2339e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8301/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8302/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8303/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2267e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8304/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8305/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2315e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8306/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8307/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2267e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8308/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2267e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8309/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8310/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2267e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8311/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2267e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8312/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2243e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8313/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2291e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8314/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2267e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8315/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2243e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8316/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2267e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8317/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2267e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8318/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2220e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8319/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2220e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8320/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2196e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8321/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2172e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8322/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2196e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8323/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2196e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8324/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2148e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8325/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2148e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8326/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2148e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8327/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2148e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8328/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2124e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8329/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2172e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8330/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2148e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8331/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2100e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8332/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2077e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8333/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2053e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8334/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2053e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8335/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2029e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8336/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2100e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8337/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2077e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8338/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2124e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8339/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2100e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8340/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2124e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8341/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2124e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8342/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2100e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 8343/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2100e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8344/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2053e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8345/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2077e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8346/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2077e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8347/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2100e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8348/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2029e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8349/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2053e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8350/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2100e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8351/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2053e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8352/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2029e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8353/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2005e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8354/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2053e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8355/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2053e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8356/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2053e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8357/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2053e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8358/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2029e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8359/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2005e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8360/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1934e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8361/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1981e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8362/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1981e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8363/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1957e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8364/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1957e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8365/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1934e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8366/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2005e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8367/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.2005e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8368/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1981e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8369/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1934e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8370/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1957e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8371/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1957e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8372/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1957e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8373/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1981e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8374/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1957e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8375/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1957e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8376/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1934e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8377/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1934e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8378/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8379/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8380/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8381/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1886e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8382/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8383/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8384/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8385/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1957e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8386/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1934e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8387/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8388/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1934e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8389/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1886e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8390/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1886e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8391/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1838e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8392/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1934e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8393/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8394/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8395/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1886e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8396/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8397/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8398/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1934e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8399/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1910e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8400/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1838e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8401/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1862e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8402/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1862e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8403/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1838e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8404/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1862e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8405/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1886e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8406/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1814e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8407/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1814e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8408/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1838e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8409/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1791e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8410/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1791e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8411/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1814e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8412/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1791e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8413/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1767e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8414/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1743e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 8415/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1767e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8416/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1791e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8417/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1791e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8418/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1767e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8419/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1791e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 8420/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1743e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8421/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1743e-07 - val_accuracy: 1.0000 - 74ms/epoch - 5ms/step\n",
            "Epoch 8422/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8423/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8424/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8425/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8426/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8427/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8428/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8429/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 108ms/epoch - 7ms/step\n",
            "Epoch 8430/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8431/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8432/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1743e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8433/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8434/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8435/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8436/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8437/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8438/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8439/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8440/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 52ms/epoch - 3ms/step\n",
            "Epoch 8441/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8442/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8443/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8444/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8445/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8446/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8447/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8448/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8449/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8450/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8451/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8452/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8453/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8454/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8455/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8456/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8457/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8458/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8459/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8460/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8461/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8462/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8463/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1719e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8464/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8465/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8466/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8467/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8468/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8469/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8470/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8471/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8472/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8473/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8474/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8475/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1695e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8476/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8477/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8478/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8479/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8480/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8481/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1624e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8482/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8483/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1624e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8484/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1624e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8485/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1671e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8486/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 52ms/epoch - 3ms/step\n",
            "Epoch 8487/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1647e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8488/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8489/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8490/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8491/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8492/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8493/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8494/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8495/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8496/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8497/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8498/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8499/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8500/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8501/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8502/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8503/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8504/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8505/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8506/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8507/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8508/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8509/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8510/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8511/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8512/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8513/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8514/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8515/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8516/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8517/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1624e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8518/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8519/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8520/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8521/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8522/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8523/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1624e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8524/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8525/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8526/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1624e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8527/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1624e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8528/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8529/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8530/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8531/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8532/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1528e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8533/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8534/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8535/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8536/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1624e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8537/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8538/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 73ms/epoch - 5ms/step\n",
            "Epoch 8539/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1552e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8540/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1528e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8541/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1504e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8542/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8543/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1600e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8544/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1576e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8545/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1504e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8546/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1481e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8547/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1457e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8548/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1481e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8549/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1457e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8550/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1528e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8551/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1528e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8552/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1528e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8553/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1504e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8554/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1504e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8555/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1481e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8556/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1481e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8557/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1504e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8558/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1504e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8559/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1433e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8560/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1433e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8561/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1504e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8562/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1528e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8563/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1504e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8564/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1504e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8565/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1504e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8566/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1457e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8567/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1409e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8568/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1457e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8569/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1481e-07 - val_accuracy: 1.0000 - 76ms/epoch - 5ms/step\n",
            "Epoch 8570/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1433e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8571/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8572/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1433e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8573/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1433e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8574/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1433e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8575/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1433e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8576/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8577/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1457e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8578/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8579/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8580/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1409e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8581/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1361e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8582/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8583/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8584/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1361e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8585/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8586/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1361e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8587/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1290e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8588/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8589/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8590/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1361e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8591/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1433e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8592/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1409e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8593/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8594/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8595/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8596/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8597/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8598/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1409e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8599/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1409e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8600/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1361e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8601/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1385e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8602/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1361e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8603/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8604/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1290e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8605/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8606/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8607/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1290e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8608/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8609/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1290e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8610/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8611/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8612/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8613/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8614/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8615/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1409e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8616/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1361e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8617/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1361e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8618/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8619/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1361e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8620/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8621/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8622/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8623/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8624/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1361e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8625/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8626/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8627/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8628/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8629/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1338e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8630/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8631/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1290e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8632/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8633/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8634/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8635/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8636/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8637/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1314e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8638/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1290e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8639/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8640/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1290e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8641/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8642/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8643/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8644/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8645/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8646/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8647/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8648/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8649/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8650/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8651/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8652/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8653/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8654/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8655/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8656/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8657/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8658/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8659/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8660/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8661/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 8662/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 75ms/epoch - 5ms/step\n",
            "Epoch 8663/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8664/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8665/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8666/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8667/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8668/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8669/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8670/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8671/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8672/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8673/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8674/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8675/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8676/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8677/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 104ms/epoch - 7ms/step\n",
            "Epoch 8678/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8679/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8680/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8681/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8682/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8683/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1266e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8684/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8685/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8686/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8687/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8688/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8689/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8690/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8691/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8692/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8693/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8694/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8695/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8696/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1171e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8697/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8698/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8699/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8700/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1242e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8701/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8702/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8703/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8704/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8705/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8706/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1099e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8707/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8708/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8709/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8710/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8711/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8712/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1099e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8713/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8714/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8715/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8716/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8717/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8718/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8719/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1171e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8720/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8721/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8722/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1171e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8723/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8724/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 75ms/epoch - 5ms/step\n",
            "Epoch 8725/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1171e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8726/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8727/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1171e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8728/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8729/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8730/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8731/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8732/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8733/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8734/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1171e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8735/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8736/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1218e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8737/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1171e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8738/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8739/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8740/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1195e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8741/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1171e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8742/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1171e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8743/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8744/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1123e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8745/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1075e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8746/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1099e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8747/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1099e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8748/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1099e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8749/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1123e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8750/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1123e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8751/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1123e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8752/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1147e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8753/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1099e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8754/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1099e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8755/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1051e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8756/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1051e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8757/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1075e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8758/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1051e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8759/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1028e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8760/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1075e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8761/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1028e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8762/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1051e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8763/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1028e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8764/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1004e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8765/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1028e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8766/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1004e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8767/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1004e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8768/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1004e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8769/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0980e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8770/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8771/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0956e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8772/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0956e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8773/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.1004e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8774/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8775/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0956e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8776/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0956e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8777/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8778/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8779/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8780/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8781/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8782/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0861e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8783/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0861e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8784/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8785/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8786/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8787/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8788/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8789/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8790/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8791/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8792/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8793/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8794/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8795/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8796/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8797/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8798/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8799/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8800/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8801/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8802/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8803/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8804/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8805/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8806/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0956e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8807/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8808/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0932e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8809/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8810/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0861e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8811/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8812/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8813/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8814/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8815/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0861e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8816/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0837e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 8817/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0861e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8818/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0861e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8819/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0861e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8820/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0861e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8821/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8822/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0885e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8823/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8824/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0908e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8825/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0837e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8826/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0837e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8827/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0861e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8828/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0837e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8829/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0837e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8830/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0837e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8831/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0861e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8832/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0813e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8833/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0813e-07 - val_accuracy: 1.0000 - 72ms/epoch - 5ms/step\n",
            "Epoch 8834/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0837e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8835/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0813e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8836/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0813e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8837/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0789e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8838/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0765e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8839/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0789e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8840/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0789e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8841/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0813e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8842/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0765e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8843/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0742e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8844/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0765e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8845/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0742e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8846/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0742e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8847/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0765e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8848/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0813e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8849/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0789e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8850/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0765e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8851/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0789e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8852/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0789e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8853/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0765e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8854/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0789e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8855/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0765e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8856/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0765e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8857/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0742e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8858/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0694e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8859/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0694e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8860/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0670e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8861/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0694e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8862/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0694e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8863/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0718e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8864/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0742e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8865/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0718e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8866/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0789e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8867/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0742e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8868/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0694e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8869/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0694e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8870/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0694e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8871/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0694e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8872/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0670e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8873/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0670e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8874/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0646e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8875/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0646e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8876/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0694e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8877/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0646e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8878/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0670e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8879/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0646e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8880/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0646e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8881/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0622e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8882/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0622e-07 - val_accuracy: 1.0000 - 52ms/epoch - 3ms/step\n",
            "Epoch 8883/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0622e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8884/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0622e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8885/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0575e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8886/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0622e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8887/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0646e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8888/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0622e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8889/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0551e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8890/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0575e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8891/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0599e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8892/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0551e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 8893/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0551e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8894/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0527e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8895/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0527e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8896/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0527e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8897/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0503e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8898/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0479e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8899/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0527e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8900/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0479e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8901/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0479e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8902/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0479e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8903/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0479e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8904/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0503e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8905/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0503e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8906/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0551e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8907/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0527e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8908/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0503e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8909/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0527e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8910/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0503e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8911/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0503e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8912/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0503e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8913/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0503e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8914/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0503e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8915/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0455e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8916/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0479e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8917/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0455e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 8918/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0479e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8919/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0455e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8920/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0432e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8921/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0479e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8922/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0432e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8923/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0432e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8924/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0408e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8925/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0408e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8926/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0408e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8927/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0432e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8928/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0384e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8929/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0384e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8930/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0384e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8931/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0384e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8932/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0408e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8933/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0384e-07 - val_accuracy: 1.0000 - 74ms/epoch - 5ms/step\n",
            "Epoch 8934/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0360e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8935/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0360e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 8936/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0360e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8937/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0336e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8938/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0312e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8939/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0289e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8940/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0312e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8941/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0312e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8942/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0289e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8943/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0312e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8944/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0289e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8945/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0384e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8946/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0336e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8947/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0360e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8948/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0384e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8949/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0336e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8950/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0312e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8951/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0312e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8952/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0312e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8953/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0289e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8954/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0312e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8955/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0312e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8956/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0336e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8957/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0289e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8958/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8959/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0289e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8960/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8961/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8962/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0289e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8963/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 8964/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0289e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8965/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0289e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8966/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8967/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8968/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 8969/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8970/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8971/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8972/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8973/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8974/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 8975/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8976/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8977/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 8978/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 8979/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8980/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 8981/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8982/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8983/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 8984/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8985/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8986/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8987/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8988/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 8989/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 8990/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8991/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8992/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 8993/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 8994/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0289e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 8995/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 8996/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 8997/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 8998/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 8999/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9000/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9001/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9002/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9003/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9004/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9005/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9006/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9007/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9008/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9009/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9010/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9011/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9012/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9013/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0336e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9014/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0241e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9015/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0265e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9016/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9017/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9018/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9019/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9020/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9021/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9022/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9023/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9024/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9025/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9026/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9027/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9028/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9029/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9030/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0217e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9031/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9032/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9033/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9034/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0193e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9035/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9036/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9037/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0146e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9038/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9039/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9040/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0169e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9041/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0146e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9042/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0122e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9043/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0146e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9044/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0098e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9045/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0122e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9046/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0122e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9047/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0098e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9048/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0098e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9049/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9050/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9051/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0122e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9052/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0098e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9053/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0098e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9054/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0050e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9055/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0098e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9056/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0050e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9057/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0098e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9058/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9059/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0050e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9060/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0050e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9061/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9062/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9063/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9064/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9065/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9066/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9067/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0050e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9068/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9069/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0074e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9070/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0026e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9071/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0098e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9072/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0098e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9073/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0050e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9074/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0026e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9075/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0050e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9076/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0026e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9077/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0026e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9078/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0026e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9079/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0003e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9080/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0026e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9081/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0026e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9082/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0003e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9083/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0003e-07 - val_accuracy: 1.0000 - 73ms/epoch - 5ms/step\n",
            "Epoch 9084/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9955e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9085/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9979e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9086/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0003e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9087/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 2.0003e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9088/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9979e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9089/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9979e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9090/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9955e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9091/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9979e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9092/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9955e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9093/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9094/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9095/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9096/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9097/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9098/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9979e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9099/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9100/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9955e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9101/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9955e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9102/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9103/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9104/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 74ms/epoch - 5ms/step\n",
            "Epoch 9105/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9106/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9931e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9107/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9907e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9108/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9907e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9109/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9883e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9110/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9859e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9111/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9112/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9883e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9113/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9859e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9114/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9115/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9859e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9116/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9859e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9117/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9883e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9118/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9883e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9119/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9859e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9120/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9859e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9121/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9859e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9122/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9123/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9124/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9125/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9126/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9127/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9128/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9129/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9130/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9131/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9132/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9133/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9134/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9135/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9136/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9137/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9138/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9836e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 9139/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9140/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9141/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9142/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9143/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9144/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9145/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9146/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9788e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9147/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9148/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9149/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9150/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9151/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9152/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9788e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9153/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9812e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9154/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9155/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9788e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9156/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9157/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9158/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9159/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 72ms/epoch - 4ms/step\n",
            "Epoch 9160/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9161/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9162/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9163/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9164/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9165/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9716e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9166/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9716e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9167/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9716e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9168/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9169/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9788e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9170/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9171/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9172/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9788e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9173/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9788e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9174/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9175/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9176/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9177/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9178/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9179/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9788e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9180/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9764e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9181/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9182/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9183/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9716e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9184/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9185/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9186/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9187/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9740e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9188/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9189/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9716e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9190/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9191/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9192/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9716e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9193/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9716e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9194/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9195/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9196/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9645e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9197/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9198/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9199/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9200/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9201/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9669e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9202/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9203/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9693e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9204/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9669e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9205/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9716e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9206/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9669e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9207/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9645e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9208/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9597e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9209/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9621e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9210/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9621e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9211/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9597e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9212/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9597e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9213/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9597e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9214/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9597e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9215/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9573e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9216/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9573e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9217/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9550e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9218/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9550e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9219/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9502e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9220/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9526e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9221/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9550e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9222/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9526e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9223/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9502e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9224/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9502e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9225/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9526e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9226/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9502e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9227/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9502e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9228/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9502e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9229/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9526e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9230/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9502e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 9231/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9502e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9232/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9526e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9233/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9478e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9234/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9478e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9235/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9478e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9236/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9478e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9237/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9478e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9238/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9502e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9239/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9478e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9240/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9454e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9241/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9407e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9242/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9454e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9243/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9244/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9245/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9246/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9478e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9247/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9454e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9248/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9249/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9250/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9251/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9407e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9252/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9253/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9254/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9255/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9256/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9257/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9407e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9258/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9259/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9430e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9260/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9383e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9261/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9407e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9262/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9383e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9263/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9383e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9264/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9359e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9265/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9359e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9266/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9383e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9267/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9268/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9359e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9269/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9270/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9359e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9271/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9287e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9272/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9273/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9274/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9359e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9275/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9276/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9359e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9277/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9359e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9278/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9279/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9280/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9281/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9359e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9282/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9283/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9284/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9285/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9286/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9287/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9288/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9289/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9290/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9291/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9359e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9292/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9293/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9294/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 52ms/epoch - 3ms/step\n",
            "Epoch 9295/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9296/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9297/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9298/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9335e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9299/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9300/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9359e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9301/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9302/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9303/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9287e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9304/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9287e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9305/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9306/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9287e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9307/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9308/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9311e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9309/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9287e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9310/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9263e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9311/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9263e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9312/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9287e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9313/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9287e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9314/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9263e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9315/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9316/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9317/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9263e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9318/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9287e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9319/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9320/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9321/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9263e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9322/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9323/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9324/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9325/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9263e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9326/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9327/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9263e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9328/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9329/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9330/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9331/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9332/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9333/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9334/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9263e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9335/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9336/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9263e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9337/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9338/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9339/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 9340/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9240e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9341/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9342/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9343/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9344/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9345/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9346/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9347/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9348/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9349/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9350/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9351/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9352/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9353/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9354/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9355/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9216e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9356/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9357/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9358/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9359/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9360/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9144e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9361/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9097e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9362/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9097e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 9363/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9120e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9364/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9144e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9365/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9366/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9367/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9368/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9192e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9369/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9144e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9370/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9371/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9372/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9373/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9374/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9144e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9375/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9376/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9144e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9377/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9144e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 9378/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9120e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 9379/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9097e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9380/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9381/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9144e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9382/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9144e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9383/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9144e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9384/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9168e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9385/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9120e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9386/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9120e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9387/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9097e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9388/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9120e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9389/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9120e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9390/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9144e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9391/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9120e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9392/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9120e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9393/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9073e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9394/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9025e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9395/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9049e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9396/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9073e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9397/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9049e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9398/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9049e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9399/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9073e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9400/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9097e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9401/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9049e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9402/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9049e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9403/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9049e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9404/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9049e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9405/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9025e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9406/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9049e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9407/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9025e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9408/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9025e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9409/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9025e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9410/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9049e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 9411/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9001e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9412/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9025e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9413/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9025e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9414/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8977e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9415/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9001e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9416/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9001e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9417/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.9025e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9418/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8977e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9419/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8954e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9420/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8954e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9421/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8906e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9422/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8906e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9423/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8930e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9424/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8930e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9425/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8954e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9426/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8954e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9427/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8954e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9428/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8954e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9429/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8930e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9430/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8930e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9431/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8930e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9432/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8906e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9433/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8930e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9434/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8930e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9435/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8930e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9436/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8930e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9437/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8930e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9438/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8906e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9439/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8882e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9440/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8882e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9441/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8882e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9442/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8882e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9443/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8882e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9444/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8882e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9445/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8906e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9446/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8882e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9447/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9448/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8906e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9449/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8882e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9450/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9451/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8882e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9452/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8882e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9453/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9454/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9455/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8906e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9456/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9457/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 72ms/epoch - 4ms/step\n",
            "Epoch 9458/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9459/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8834e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9460/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9461/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8834e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9462/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9463/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9464/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9465/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9466/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9467/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8834e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9468/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9469/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9470/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9471/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9472/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9473/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8858e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9474/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8811e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9475/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8811e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9476/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8834e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9477/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8787e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9478/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8811e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9479/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8811e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9480/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8811e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9481/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8811e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9482/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8787e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9483/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8811e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9484/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8787e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9485/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8787e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9486/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8811e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9487/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8763e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9488/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8787e-07 - val_accuracy: 1.0000 - 105ms/epoch - 7ms/step\n",
            "Epoch 9489/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8763e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9490/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8787e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9491/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8787e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9492/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8787e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9493/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8763e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9494/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8763e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9495/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8763e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9496/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8787e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9497/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8763e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9498/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8739e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9499/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8715e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9500/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8715e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9501/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8739e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9502/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8715e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9503/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8715e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9504/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8691e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9505/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8691e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9506/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8667e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9507/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8715e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9508/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8691e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9509/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8644e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9510/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9511/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8667e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9512/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8644e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9513/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8644e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9514/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8644e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9515/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8644e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9516/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9517/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8644e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9518/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9519/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 9520/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8644e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9521/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9522/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9523/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9524/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9525/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9526/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9527/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9528/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9529/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9530/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9531/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9532/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9533/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9534/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9535/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9536/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9537/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9538/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9539/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9540/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9541/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9542/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9543/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9544/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 52ms/epoch - 3ms/step\n",
            "Epoch 9545/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9546/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9547/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9548/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9549/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9550/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9551/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9552/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9553/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9554/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9555/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9556/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9557/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9558/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9559/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9560/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9561/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9562/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9563/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9564/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9565/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9566/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 78ms/epoch - 5ms/step\n",
            "Epoch 9567/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9568/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9569/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9570/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9571/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9572/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9573/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9574/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9575/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9576/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9577/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9578/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9579/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9580/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9581/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9582/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9583/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9584/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9585/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9586/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9587/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9588/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9589/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9590/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9591/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9592/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9593/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9594/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9595/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9596/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9597/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9598/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9599/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9600/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9601/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9602/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9603/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9604/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9605/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9606/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9607/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9608/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8620e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9609/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9610/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8596e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9611/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8572e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9612/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8524e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9613/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8501e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9614/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8524e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9615/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8548e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9616/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8501e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9617/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8524e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9618/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8524e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9619/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8477e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9620/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8477e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9621/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8477e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9622/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8524e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9623/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8477e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9624/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8501e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9625/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8477e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9626/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8453e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 9627/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8477e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9628/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8477e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9629/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8477e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9630/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8453e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9631/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8453e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9632/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8453e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9633/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8429e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9634/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8429e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9635/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8429e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9636/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8429e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9637/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8453e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9638/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8429e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9639/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8429e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9640/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8405e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9641/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8453e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9642/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8405e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9643/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 52ms/epoch - 3ms/step\n",
            "Epoch 9644/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9645/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9646/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9647/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8405e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9648/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9649/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9650/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9651/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9652/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9653/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9654/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9655/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9656/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9657/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8405e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9658/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9659/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9660/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9661/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9662/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9663/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9664/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9665/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9666/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9667/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9668/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9669/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9670/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9671/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9672/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9673/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9674/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8381e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9675/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9676/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9677/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9678/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9679/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9680/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9681/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9682/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9683/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9684/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9685/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9686/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9687/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9688/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9689/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9690/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9691/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9692/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9693/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9694/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9695/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9696/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9697/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9698/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9699/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9700/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9701/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9702/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9703/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9704/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 9705/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9706/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9707/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 76ms/epoch - 5ms/step\n",
            "Epoch 9708/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9709/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9710/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9711/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9712/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9713/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9714/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9715/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9716/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9717/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9718/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9719/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9720/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9721/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9722/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9723/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9724/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9725/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9726/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9727/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9728/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9729/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9730/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9731/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9732/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9733/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9734/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9735/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9736/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9737/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8358e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9738/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9739/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9740/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9741/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9742/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9743/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9744/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9745/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9746/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9747/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9748/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9749/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9750/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9751/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9752/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9753/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 71ms/epoch - 4ms/step\n",
            "Epoch 9754/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9755/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9756/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9757/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9758/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9759/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8334e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9760/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9761/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9762/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9763/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8310e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9764/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8286e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9765/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8262e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9766/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9767/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9768/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9769/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9770/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9771/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9772/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9773/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9774/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9775/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9776/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9777/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9778/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9779/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9780/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9781/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9782/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8167e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9783/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9784/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9785/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9786/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9787/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9788/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9789/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9790/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9791/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9792/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9793/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9794/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9795/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9796/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9797/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 9798/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9799/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9800/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 52ms/epoch - 3ms/step\n",
            "Epoch 9801/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9802/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9803/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9804/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9805/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9806/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9807/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9808/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9809/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8238e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9810/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9811/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9812/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9813/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9814/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9815/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9816/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9817/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9818/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9819/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9820/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9821/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9822/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9823/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9824/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8215e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9825/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8167e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9826/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9827/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8167e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9828/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8167e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9829/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8167e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9830/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8191e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9831/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8143e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9832/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8167e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9833/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8167e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9834/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8119e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9835/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8143e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9836/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8119e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9837/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8119e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9838/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8095e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9839/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8119e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9840/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8143e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9841/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8143e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9842/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8119e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9843/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8119e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9844/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8119e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9845/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8119e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9846/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8119e-07 - val_accuracy: 1.0000 - 75ms/epoch - 5ms/step\n",
            "Epoch 9847/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8071e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9848/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8119e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9849/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8071e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9850/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9851/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8071e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9852/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9853/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9854/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9855/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9856/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9857/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9858/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8024e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9859/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9860/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8000e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9861/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7976e-07 - val_accuracy: 1.0000 - 103ms/epoch - 6ms/step\n",
            "Epoch 9862/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8000e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9863/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7976e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9864/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8000e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9865/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9866/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9867/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8024e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9868/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8048e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9869/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8000e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9870/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8024e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9871/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8000e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9872/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8024e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9873/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8000e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9874/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7976e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9875/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7976e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9876/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8000e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9877/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7976e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9878/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7976e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9879/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.8000e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9880/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7976e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9881/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7952e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9882/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7952e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9883/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9884/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7905e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9885/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7905e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9886/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7952e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9887/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9888/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7952e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9889/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7905e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9890/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7881e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9891/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7905e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9892/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9893/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7905e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9894/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9895/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9896/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7881e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9897/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9898/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9899/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7881e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9900/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7905e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9901/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9902/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7905e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9903/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9904/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7952e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9905/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9906/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7952e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9907/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9908/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7928e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9909/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7905e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9910/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7905e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9911/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7881e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9912/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7881e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9913/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7857e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9914/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7857e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9915/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7881e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9916/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 70ms/epoch - 4ms/step\n",
            "Epoch 9917/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9918/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9919/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9920/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7857e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9921/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9922/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9923/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9924/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9925/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7857e-07 - val_accuracy: 1.0000 - 51ms/epoch - 3ms/step\n",
            "Epoch 9926/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9927/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9928/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9929/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9930/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7881e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9931/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9932/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7857e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9933/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7833e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9934/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7809e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9935/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7785e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9936/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9937/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9938/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7809e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9939/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7785e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9940/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9941/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9942/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9943/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7785e-07 - val_accuracy: 1.0000 - 56ms/epoch - 3ms/step\n",
            "Epoch 9944/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9945/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7738e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9946/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 68ms/epoch - 4ms/step\n",
            "Epoch 9947/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9948/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7738e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9949/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7738e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9950/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9951/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9952/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9953/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9954/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9955/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9956/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7738e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9957/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7762e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9958/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7714e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9959/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7714e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9960/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7714e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9961/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9962/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9963/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9964/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9965/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9966/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7714e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9967/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9968/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9969/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9970/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9971/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9972/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9973/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9974/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9975/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9976/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 69ms/epoch - 4ms/step\n",
            "Epoch 9977/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 54ms/epoch - 3ms/step\n",
            "Epoch 9978/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9979/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9980/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9981/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 9982/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9983/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 60ms/epoch - 4ms/step\n",
            "Epoch 9984/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n",
            "Epoch 9985/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 63ms/epoch - 4ms/step\n",
            "Epoch 9986/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 55ms/epoch - 3ms/step\n",
            "Epoch 9987/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7690e-07 - val_accuracy: 1.0000 - 67ms/epoch - 4ms/step\n",
            "Epoch 9988/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9989/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 59ms/epoch - 4ms/step\n",
            "Epoch 9990/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 62ms/epoch - 4ms/step\n",
            "Epoch 9991/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9992/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9993/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 58ms/epoch - 4ms/step\n",
            "Epoch 9994/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 66ms/epoch - 4ms/step\n",
            "Epoch 9995/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7666e-07 - val_accuracy: 1.0000 - 61ms/epoch - 4ms/step\n",
            "Epoch 9996/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7642e-07 - val_accuracy: 1.0000 - 56ms/epoch - 4ms/step\n",
            "Epoch 9997/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7619e-07 - val_accuracy: 1.0000 - 65ms/epoch - 4ms/step\n",
            "Epoch 9998/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7619e-07 - val_accuracy: 1.0000 - 57ms/epoch - 4ms/step\n",
            "Epoch 9999/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7619e-07 - val_accuracy: 1.0000 - 64ms/epoch - 4ms/step\n",
            "Epoch 10000/10000\n",
            "16/16 - 0s - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 1.7619e-07 - val_accuracy: 1.0000 - 53ms/epoch - 3ms/step\n"
          ]
        }
      ],
      "source": [
        "#设置参数\n",
        "\n",
        "epochs=10000 #迭代次数\n",
        "batch_size=32 #batchsize\n",
        "number_of_res_blocks=1 #残差块个数\n",
        "number_dense_of_res_block=4 #残差块的隐层数\n",
        "number_neurons_of_dense=2 #隐层的神经元数\n",
        "res_model,history=ResMLP_Model(number_of_res_blocks,\n",
        "                               number_dense_of_res_block,\n",
        "                               number_neurons_of_dense)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XObIjPPVJrRI"
      },
      "source": [
        "### **可视化结果**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2Vp2CmNNJsfg"
      },
      "source": [
        "#### **绘制训练曲线**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 308
        },
        "id": "Y4DPuAKw-sjv",
        "outputId": "5902e349-126d-4cdb-bf50-fa4093617ba1"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "#绘制训练曲线\n",
        "plot_loss_accuracy(history,\n",
        "                    f'Training curve',\n",
        "                    f'Training curve')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hcT0ySlkJuJ6"
      },
      "source": [
        "#### **输出模型结构信息**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SYnKqSh4-xT9",
        "outputId": "3244fbba-1110-4a28-d185-e00628d0f910"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"model\"\n",
            "__________________________________________________________________________________________________\n",
            " Layer (type)                   Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            " input_1 (InputLayer)           [(None, 2)]          0           []                               \n",
            "                                                                                                  \n",
            " 0th-resBlock_0th-hidden (Dense  (None, 2)           6           ['input_1[0][0]']                \n",
            " )                                                                                                \n",
            "                                                                                                  \n",
            " 0th-resBlock_linear (Dense)    (None, 2)            6           ['0th-resBlock_0th-hidden[0][0]']\n",
            "                                                                                                  \n",
            " 0th-resBlock_Add (Add)         (None, 2)            0           ['0th-resBlock_linear[0][0]',    \n",
            "                                                                  'input_1[0][0]']                \n",
            "                                                                                                  \n",
            " 1th-resBlock_0th-hidden (Dense  (None, 2)           6           ['0th-resBlock_Add[0][0]']       \n",
            " )                                                                                                \n",
            "                                                                                                  \n",
            " 1th-resBlock_linear (Dense)    (None, 2)            6           ['1th-resBlock_0th-hidden[0][0]']\n",
            "                                                                                                  \n",
            " 1th-resBlock_Add (Add)         (None, 2)            0           ['1th-resBlock_linear[0][0]',    \n",
            "                                                                  '0th-resBlock_Add[0][0]']       \n",
            "                                                                                                  \n",
            " 2th-resBlock_0th-hidden (Dense  (None, 2)           6           ['1th-resBlock_Add[0][0]']       \n",
            " )                                                                                                \n",
            "                                                                                                  \n",
            " 2th-resBlock_linear (Dense)    (None, 2)            6           ['2th-resBlock_0th-hidden[0][0]']\n",
            "                                                                                                  \n",
            " 2th-resBlock_Add (Add)         (None, 2)            0           ['2th-resBlock_linear[0][0]',    \n",
            "                                                                  '1th-resBlock_Add[0][0]']       \n",
            "                                                                                                  \n",
            " 3th-resBlock_0th-hidden (Dense  (None, 2)           6           ['2th-resBlock_Add[0][0]']       \n",
            " )                                                                                                \n",
            "                                                                                                  \n",
            " 3th-resBlock_linear (Dense)    (None, 2)            6           ['3th-resBlock_0th-hidden[0][0]']\n",
            "                                                                                                  \n",
            " 3th-resBlock_Add (Add)         (None, 2)            0           ['3th-resBlock_linear[0][0]',    \n",
            "                                                                  '2th-resBlock_Add[0][0]']       \n",
            "                                                                                                  \n",
            " 4th-resBlock_0th-hidden (Dense  (None, 2)           6           ['3th-resBlock_Add[0][0]']       \n",
            " )                                                                                                \n",
            "                                                                                                  \n",
            " 4th-resBlock_linear (Dense)    (None, 2)            6           ['4th-resBlock_0th-hidden[0][0]']\n",
            "                                                                                                  \n",
            " 4th-resBlock_Add (Add)         (None, 2)            0           ['4th-resBlock_linear[0][0]',    \n",
            "                                                                  '3th-resBlock_Add[0][0]']       \n",
            "                                                                                                  \n",
            " dense (Dense)                  (None, 2)            6           ['4th-resBlock_Add[0][0]']       \n",
            "                                                                                                  \n",
            "==================================================================================================\n",
            "Total params: 66\n",
            "Trainable params: 66\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "#打印模型结构和参数信息\n",
        "res_model.summary()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1WpDSBAmJvRm"
      },
      "source": [
        "#### **对原始特征空间剖分的可视化**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 408
        },
        "id": "fTXOiBaF-xxt",
        "outputId": "acc86d31-6aa3-4144-e848-9b5a4e2924cb"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "79/79 [==============================] - 0s 1ms/step\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 576x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "#对原始特征空间剖分的可视化\n",
        "\n",
        "fig = plt.figure(figsize = (8,6))\n",
        "prob = res_model.predict(p)[:,1]\n",
        "plt.scatter(*p.T,c = prob,cmap = cm_bright)\n",
        "plt.scatter(*X.T,c = y,cmap = cm_bright,edgecolors='white',s = 30,linewidths = 0.5)\n",
        "plt.colorbar()\n",
        "plt.title(f'Outputs of ResNet')\n",
        "plt.savefig(f'空间剖分结果.png')\n",
        "plt.savefig(f'空间剖分结果.pdf')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YaeVCfO4JwUl"
      },
      "source": [
        "#### **更换背景显示原始样本**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 391
        },
        "id": "s955LTQW-zMI",
        "outputId": "e604dadf-57b9-4156-fd4c-bae74c4ba122"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 576x432 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "#更换背景显示原始样本\n",
        "\n",
        "mpl.style.use('ggplot')\n",
        "\n",
        "fig = plt.figure(figsize = (8,6))\n",
        "top = cm.get_cmap('Oranges_r', 512)\n",
        "bottom = cm.get_cmap('Blues', 512)\n",
        "newcolors = np.vstack((top(np.linspace(0.55, 1, 512)), bottom(np.linspace(0, 0.75, 512))))\n",
        "cm_bright = ListedColormap(newcolors, name='OrangeBlue')\n",
        "\n",
        "plt.scatter(*X.T, c=y, cmap=cm_bright, edgecolors='white', s = 30, linewidths = .5)\n",
        "\n",
        "ax = fig.get_axes()\n",
        "plt.axis('equal')\n",
        "plt.colorbar(ax = ax)\n",
        "plt.title(f'Raw data ({n_samples} points)')\n",
        "plt.savefig(f'Raw data ({n_samples} points).png')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ObA8iLAkJyQ4"
      },
      "source": [
        "#### **特征变换过程可视化**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "Kl5QDwgt-0wO",
        "outputId": "ea247b93-9403-41b3-b3db-85bff158210a"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 576x432 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 576x432 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 576x432 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 576x432 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 576x432 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "#对样本点特征变换的可视化\n",
        "\n",
        "add_layers=[]\n",
        "for item in res_model.layers:\n",
        "    if isinstance(item,tf.keras.layers.Add):\n",
        "        add_layers.append(item)\n",
        "        \n",
        "inp = res_model.input\n",
        "outputs = [layer.output for layer in add_layers]\n",
        "functors = [K.function([inp], [out]) for out in outputs]\n",
        "layer_outs = [func([X]) for func in functors]\n",
        "\n",
        "#可视化经过每一个残差块特征变换后样本点的状态\n",
        "for idx in range(len(layer_outs)):\n",
        "    fig = plt.figure(figsize = (8,6))\n",
        "    plt.scatter(layer_outs[idx][0][:,0],layer_outs[idx][0][:,1],\n",
        "            c = y,cmap = cm_bright,edgecolors='white',s = 30,linewidths = 0.1)\n",
        "    plt.axis('equal')\n",
        "    plt.title(f'Activation of residual block {idx+1}')\n",
        "    plt.colorbar()\n",
        "    plt.savefig(f'Activation of residual block {idx+1}.png')\n",
        "    plt.savefig(f'Activation of residual block {idx+1}.pdf')\n",
        "    plt.show()\n",
        "    \n",
        "#生成动图\n",
        "def create_gif(image_list, gif_name, duration=1):\n",
        "    frames = []\n",
        "    for image_name in image_list:\n",
        "        frames.append(imageio.imread(image_name))\n",
        "    imageio.mimsave(gif_name, frames, 'GIF', duration=duration)\n",
        "    return\n",
        "\n",
        "def main():\n",
        "    image_list = [f'Raw data ({n_samples} points).png']\n",
        "    for i in range(len(layer_outs)):\n",
        "        image_list.append(f'Activation of residual block {i+1}.png')\n",
        "    gif_name = '特征变换动图.gif'\n",
        "    duration = 0.8\n",
        "    create_gif(image_list, gif_name, duration)\n",
        "\n",
        "main()"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}