CNN_dimest.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "CNN_dimest.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "mount_file_id": "1QlxKSWI0r_1UzZBSPe8cMt64-_jYoULL",
      "authorship_tag": "ABX9TyMGpoOWtr7fZpi3gTXWcrmw",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "TPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/hariharannatesh/Dimension-Estimation-from-Depth-Map-of-Monocular-Image/blob/master/CNN_dimest.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Uo_c3fPTfFr4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "a = []\n",
        "while(1):\n",
        "  a.append(1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AK1gOuErqYUx",
        "colab_type": "code",
        "outputId": "1c51c553-9806-459e-9dcb-16b5c5a85ce4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/gdrive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S2lYtCiotTt-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import cv2\n",
        "import numpy as np\n",
        "from google.colab.patches import cv2_imshow\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.layers import MaxPooling2D\n",
        "from tensorflow.keras.layers import Flatten\n",
        "from tensorflow.keras.layers import Input\n",
        "from tensorflow.keras.layers import Conv2D\n",
        "from tensorflow.keras.layers import Dense\n",
        "from tensorflow.keras.layers import BatchNormalization\n",
        "from tensorflow.keras import models\n",
        "from tensorflow.keras.models import load_model\n",
        "from tensorflow.keras.callbacks import ModelCheckpoint\n",
        "import argparse\n",
        "from sklearn.model_selection import train_test_split\n",
        "import os\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.utils import shuffle\n",
        "from sklearn.decomposition import PCA \n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B3rIPnLyatLJ",
        "colab_type": "code",
        "outputId": "ab36e6ba-1566-42a3-859c-1aea8103b6b2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 234
        }
      },
      "source": [
        "X_pos=np.load('gdrive/My Drive/Dataset/images_depth_positive.npy')\n",
        "X_pos=np.reshape(X_pos,(1276,200,200,1))\n",
        "cv2_imshow(X_pos[1])\n",
        "X_pos=X_pos/255.0\n",
        "Y_pos=np.load('gdrive/My Drive/Dataset/target_depth_positive.npy')\n",
        "print(Y_pos[1])\n",
        "# Y=Y/200"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=RGB size=200x200 at 0x7FE21F4745F8>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "59.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "d9efb89a-fa8b-49b7-bd1d-d1bbe1df9c7c",
        "id": "RazzhgQvgjXy",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 234
        }
      },
      "source": [
        "X=np.load('gdrive/My Drive/Dataset/images_depth_negative.npy')\n",
        "X=np.reshape(X,(1279,200,200,1))\n",
        "cv2_imshow(X[1])\n",
        "X=X/255.0\n",
        "Y=np.load('gdrive/My Drive/Dataset/target_depth_negative.npy')\n",
        "print(Y[1])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=RGB size=200x200 at 0x7FE23ADEC400>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "152.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_6XSVDS3glT_",
        "colab_type": "code",
        "outputId": "d293c481-8723-45bc-f2b0-bfcfbcb12416",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 234
        }
      },
      "source": [
        "X_neg=np.load('gdrive/My Drive/Dataset/images_depth_negative.npy')\n",
        "for i in range(len(X_neg)):\n",
        "  X_neg[i]=cv2.flip(X_neg[i],1)\n",
        "X_neg=np.reshape(X_neg,(1279,200,200,1))\n",
        "cv2_imshow(X_neg[1])\n",
        "X_neg=X_neg/255.0\n",
        "Y_neg=np.load('gdrive/My Drive/Dataset/target_depth_negative.npy')\n",
        "print(Y_neg[1])"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<PIL.Image.Image image mode=RGB size=200x200 at 0x7FE21F1E85C0>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "152.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2KI1JZnr4ByW",
        "colab_type": "code",
        "outputId": "f5ce215a-6fcc-4796-cb40-1ca188777c20",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "X=np.concatenate((X_pos,X_neg),axis=0)\n",
        "Y=np.concatenate((Y_pos,Y_neg),axis=0)\n",
        "print(len(X),len(Y))"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2555 2555\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Nv6vIfKD_Vm-",
        "colab_type": "code",
        "outputId": "53185a4e-c782-4dd4-ce6c-11e38dc06de4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "X,Y=shuffle(X,Y,random_state=100)\n",
        "X_new=X[0:1277]\n",
        "X_rest=X[1277:len(X)]\n",
        "Y_new=Y[0:1277]\n",
        "Y_rest=Y[1277:len(Y)]\n",
        "print(len(X_new))\n",
        "print(len(Y_new))\n",
        "print(len(X_rest))"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1277\n",
            "1277\n",
            "1278\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5mUAaCOke74B",
        "colab_type": "code",
        "outputId": "95284dc5-8024-449d-aa98-ac43916c2415",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 68
        }
      },
      "source": [
        "(trainX,testX,trainY,testY)=train_test_split(X_new,Y_new,test_size=0.2,random_state=73)\n",
        "print(trainX[0].shape)\n",
        "print(len(trainY))\n",
        "print(trainY[0])"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(200, 200, 1)\n",
            "1021\n",
            "105.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qV70zJwal3eK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model=models.Sequential()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AHF2_DIzl8Va",
        "colab_type": "code",
        "outputId": "48c76bfa-30e8-4376-f2c5-5ae446fef6a7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 459
        }
      },
      "source": [
        "model.add(Conv2D(16,(11,11),padding=\"same\",data_format='channels_last',input_shape=(200,200,1),use_bias=True,activation='relu'))\n",
        "model.add(MaxPooling2D(pool_size=(2,2),strides=(2,2)))\n",
        "model.add(BatchNormalization(axis=2))\n",
        "model.add(Conv2D(32,(7,7),padding=\"same\",data_format='channels_last',use_bias=True,activation='relu'))\n",
        "model.add(MaxPooling2D(pool_size=(2,2),strides=(2,2)))\n",
        "model.add(BatchNormalization(axis=2))\n",
        "model.add(Conv2D(64,(5,5),padding=\"same\",data_format='channels_last',use_bias=True,activation='relu'))\n",
        "model.add(MaxPooling2D(pool_size=(2,2),strides=(2,2)))\n",
        "model.add(BatchNormalization(axis=2))\n",
        "#model.add(MaxPooling2D(pool_size=(2,2),strides=(1,1)))\n",
        "#model.add(Conv2D(128,(3,3),padding=\"same\",data_format='channels_last',use_bias=True,activation='relu'))\n",
        "#model.add(MaxPooling2D(pool_size=(2,2),strides=(2,2)))\n",
        "model.summary()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 200, 200, 16)      1952      \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 100, 100, 16)      0         \n",
            "_________________________________________________________________\n",
            "batch_normalization (BatchNo (None, 100, 100, 16)      400       \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 100, 100, 32)      25120     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 50, 50, 32)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_1 (Batch (None, 50, 50, 32)        200       \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 50, 50, 64)        51264     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 25, 25, 64)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 25, 25, 64)        100       \n",
            "=================================================================\n",
            "Total params: 79,036\n",
            "Trainable params: 78,686\n",
            "Non-trainable params: 350\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "08976492-11ea-4a0d-e27a-0072714ca230",
        "id": "PHslMSdh4ehM",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 595
        }
      },
      "source": [
        "model.add(Flatten())\n",
        "#model.add(Dense(10000,activation='relu',use_bias=True))\n",
        "model.add(Dense(1000,activation='relu',use_bias=True))\n",
        "model.add(Dense(100,activation='relu',use_bias=True))\n",
        "model.add(Dense(1))\n",
        "model.summary()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 200, 200, 16)      1952      \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 100, 100, 16)      0         \n",
            "_________________________________________________________________\n",
            "batch_normalization (BatchNo (None, 100, 100, 16)      400       \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 100, 100, 32)      25120     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 50, 50, 32)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_1 (Batch (None, 50, 50, 32)        200       \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 50, 50, 64)        51264     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 25, 25, 64)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 25, 25, 64)        100       \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 40000)             0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 1000)              40001000  \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 100)               100100    \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 1)                 101       \n",
            "=================================================================\n",
            "Total params: 40,180,237\n",
            "Trainable params: 40,179,887\n",
            "Non-trainable params: 350\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZWJi5v53j9fb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "checkpoint_path = \"gdrive/My Drive/training/cpbn_mix_pos500.ckpt\"\n",
        "checkpoint_dir = os.path.dirname(checkpoint_path)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-3jcBn53k3Ja",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "cp_callback = ModelCheckpoint(filepath=checkpoint_path,verbose=1,save_weights_only=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9qTCsdMxojIb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.save_weights(checkpoint_path.format(epoch=0))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mXRMEl6qfq1s",
        "colab_type": "code",
        "outputId": "90b3c1d6-c34b-46ce-ad31-d825eb9d1160",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "model.load_weights(checkpoint_path)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.training.tracking.util.CheckpointLoadStatus at 0x7f1f41ea57b8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1BBMf6f7s_VC",
        "colab_type": "code",
        "outputId": "71d938ba-ac1a-4b07-e657-3cc14c94cce2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "model.compile(optimizer='rmsprop',loss=tf.keras.losses.MeanSquaredError(),metrics=['mean_squared_error'])\n",
        "history=model.fit(trainX,trainY,initial_epoch=412,epochs=500,shuffle=1,validation_split=0.2,callbacks=[cp_callback])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 413/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 256.4176 - mean_squared_error: 256.0977\n",
            "Epoch 00413: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 950ms/step - loss: 256.4176 - mean_squared_error: 256.0977 - val_loss: 376.6798 - val_mean_squared_error: 395.0449\n",
            "Epoch 414/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 211.6191 - mean_squared_error: 213.3567\n",
            "Epoch 00414: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 933ms/step - loss: 211.6191 - mean_squared_error: 213.3567 - val_loss: 405.4696 - val_mean_squared_error: 404.1830\n",
            "Epoch 415/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 229.4209 - mean_squared_error: 231.7070\n",
            "Epoch 00415: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 940ms/step - loss: 229.4209 - mean_squared_error: 231.7070 - val_loss: 317.1484 - val_mean_squared_error: 320.2318\n",
            "Epoch 416/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 204.0536 - mean_squared_error: 199.0047\n",
            "Epoch 00416: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 952ms/step - loss: 204.0536 - mean_squared_error: 199.0047 - val_loss: 662.4074 - val_mean_squared_error: 675.8621\n",
            "Epoch 417/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 217.4214 - mean_squared_error: 215.2827\n",
            "Epoch 00417: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 958ms/step - loss: 217.4214 - mean_squared_error: 215.2827 - val_loss: 364.9100 - val_mean_squared_error: 379.4298\n",
            "Epoch 418/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 190.3419 - mean_squared_error: 193.1256\n",
            "Epoch 00418: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 953ms/step - loss: 190.3419 - mean_squared_error: 193.1256 - val_loss: 242.3470 - val_mean_squared_error: 258.9716\n",
            "Epoch 419/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 203.0818 - mean_squared_error: 204.6648\n",
            "Epoch 00419: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 948ms/step - loss: 203.0818 - mean_squared_error: 204.6648 - val_loss: 296.0779 - val_mean_squared_error: 310.3387\n",
            "Epoch 420/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 191.6440 - mean_squared_error: 191.5038\n",
            "Epoch 00420: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 944ms/step - loss: 191.6440 - mean_squared_error: 191.5038 - val_loss: 320.3452 - val_mean_squared_error: 341.6626\n",
            "Epoch 421/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 190.4860 - mean_squared_error: 191.9524\n",
            "Epoch 00421: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 971ms/step - loss: 190.4860 - mean_squared_error: 191.9524 - val_loss: 226.1558 - val_mean_squared_error: 239.4919\n",
            "Epoch 422/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 167.5387 - mean_squared_error: 168.6683\n",
            "Epoch 00422: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 933ms/step - loss: 167.5387 - mean_squared_error: 168.6683 - val_loss: 274.9185 - val_mean_squared_error: 285.2516\n",
            "Epoch 423/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 169.3380 - mean_squared_error: 167.5641\n",
            "Epoch 00423: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 951ms/step - loss: 169.3380 - mean_squared_error: 167.5641 - val_loss: 286.1454 - val_mean_squared_error: 301.9529\n",
            "Epoch 424/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 174.2358 - mean_squared_error: 174.0135\n",
            "Epoch 00424: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 948ms/step - loss: 174.2358 - mean_squared_error: 174.0135 - val_loss: 249.0827 - val_mean_squared_error: 259.4663\n",
            "Epoch 425/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 188.9256 - mean_squared_error: 190.6809\n",
            "Epoch 00425: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 947ms/step - loss: 188.9256 - mean_squared_error: 190.6809 - val_loss: 322.9665 - val_mean_squared_error: 339.4360\n",
            "Epoch 426/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 187.9033 - mean_squared_error: 190.6100\n",
            "Epoch 00426: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 950ms/step - loss: 187.9033 - mean_squared_error: 190.6100 - val_loss: 324.2539 - val_mean_squared_error: 336.1344\n",
            "Epoch 427/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 176.4861 - mean_squared_error: 175.6442\n",
            "Epoch 00427: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 944ms/step - loss: 176.4861 - mean_squared_error: 175.6442 - val_loss: 243.1481 - val_mean_squared_error: 259.4371\n",
            "Epoch 428/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 156.4136 - mean_squared_error: 157.1185\n",
            "Epoch 00428: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 935ms/step - loss: 156.4136 - mean_squared_error: 157.1185 - val_loss: 227.5838 - val_mean_squared_error: 243.0780\n",
            "Epoch 429/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 173.5792 - mean_squared_error: 173.4711\n",
            "Epoch 00429: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 939ms/step - loss: 173.5792 - mean_squared_error: 173.4711 - val_loss: 253.5483 - val_mean_squared_error: 271.9006\n",
            "Epoch 430/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 171.9514 - mean_squared_error: 166.0075\n",
            "Epoch 00430: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 949ms/step - loss: 171.9514 - mean_squared_error: 166.0075 - val_loss: 461.3202 - val_mean_squared_error: 481.7596\n",
            "Epoch 431/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 167.3478 - mean_squared_error: 167.0814\n",
            "Epoch 00431: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 958ms/step - loss: 167.3478 - mean_squared_error: 167.0814 - val_loss: 255.6932 - val_mean_squared_error: 270.8242\n",
            "Epoch 432/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 183.9273 - mean_squared_error: 183.0770\n",
            "Epoch 00432: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 953ms/step - loss: 183.9273 - mean_squared_error: 183.0770 - val_loss: 273.7730 - val_mean_squared_error: 276.0275\n",
            "Epoch 433/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 163.0396 - mean_squared_error: 161.6566\n",
            "Epoch 00433: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 973ms/step - loss: 163.0396 - mean_squared_error: 161.6566 - val_loss: 272.6223 - val_mean_squared_error: 276.3504\n",
            "Epoch 434/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 176.7337 - mean_squared_error: 176.5571\n",
            "Epoch 00434: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 954ms/step - loss: 176.7337 - mean_squared_error: 176.5571 - val_loss: 234.8636 - val_mean_squared_error: 246.1943\n",
            "Epoch 435/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 153.0094 - mean_squared_error: 153.0700\n",
            "Epoch 00435: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 953ms/step - loss: 153.0094 - mean_squared_error: 153.0700 - val_loss: 330.1701 - val_mean_squared_error: 348.4969\n",
            "Epoch 436/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 181.0547 - mean_squared_error: 181.3937\n",
            "Epoch 00436: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 979ms/step - loss: 181.0547 - mean_squared_error: 181.3937 - val_loss: 463.1734 - val_mean_squared_error: 480.0306\n",
            "Epoch 437/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 158.7148 - mean_squared_error: 156.9602\n",
            "Epoch 00437: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 953ms/step - loss: 158.7148 - mean_squared_error: 156.9602 - val_loss: 235.4185 - val_mean_squared_error: 242.1861\n",
            "Epoch 438/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 148.3363 - mean_squared_error: 149.3889\n",
            "Epoch 00438: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 935ms/step - loss: 148.3363 - mean_squared_error: 149.3889 - val_loss: 249.2270 - val_mean_squared_error: 266.3096\n",
            "Epoch 439/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 184.0392 - mean_squared_error: 176.2139\n",
            "Epoch 00439: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 945ms/step - loss: 184.0392 - mean_squared_error: 176.2139 - val_loss: 334.0654 - val_mean_squared_error: 328.1938\n",
            "Epoch 440/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 165.7593 - mean_squared_error: 166.1310\n",
            "Epoch 00440: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 941ms/step - loss: 165.7593 - mean_squared_error: 166.1310 - val_loss: 276.8310 - val_mean_squared_error: 297.4874\n",
            "Epoch 441/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 159.2711 - mean_squared_error: 159.3372\n",
            "Epoch 00441: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 940ms/step - loss: 159.2711 - mean_squared_error: 159.3372 - val_loss: 226.6829 - val_mean_squared_error: 238.8387\n",
            "Epoch 442/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 153.5248 - mean_squared_error: 149.1375\n",
            "Epoch 00442: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 953ms/step - loss: 153.5248 - mean_squared_error: 149.1375 - val_loss: 328.5606 - val_mean_squared_error: 347.0507\n",
            "Epoch 443/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 195.5818 - mean_squared_error: 198.3753\n",
            "Epoch 00443: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 941ms/step - loss: 195.5818 - mean_squared_error: 198.3753 - val_loss: 288.1063 - val_mean_squared_error: 306.7277\n",
            "Epoch 444/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 167.1296 - mean_squared_error: 165.8484\n",
            "Epoch 00444: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 948ms/step - loss: 167.1296 - mean_squared_error: 165.8484 - val_loss: 263.1187 - val_mean_squared_error: 269.4920\n",
            "Epoch 445/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 167.5071 - mean_squared_error: 168.8873\n",
            "Epoch 00445: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 972ms/step - loss: 167.5071 - mean_squared_error: 168.8873 - val_loss: 272.0921 - val_mean_squared_error: 287.7353\n",
            "Epoch 446/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 163.7446 - mean_squared_error: 163.4294\n",
            "Epoch 00446: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 958ms/step - loss: 163.7446 - mean_squared_error: 163.4294 - val_loss: 561.8094 - val_mean_squared_error: 580.9878\n",
            "Epoch 447/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 160.4778 - mean_squared_error: 160.5702\n",
            "Epoch 00447: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 970ms/step - loss: 160.4778 - mean_squared_error: 160.5702 - val_loss: 252.3568 - val_mean_squared_error: 266.7471\n",
            "Epoch 448/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 185.5319 - mean_squared_error: 185.7950\n",
            "Epoch 00448: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 950ms/step - loss: 185.5319 - mean_squared_error: 185.7950 - val_loss: 243.8586 - val_mean_squared_error: 257.4103\n",
            "Epoch 449/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 156.4543 - mean_squared_error: 158.7283\n",
            "Epoch 00449: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 953ms/step - loss: 156.4543 - mean_squared_error: 158.7283 - val_loss: 208.8122 - val_mean_squared_error: 221.5694\n",
            "Epoch 450/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 158.7623 - mean_squared_error: 160.9012\n",
            "Epoch 00450: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 950ms/step - loss: 158.7623 - mean_squared_error: 160.9012 - val_loss: 249.3465 - val_mean_squared_error: 260.7078\n",
            "Epoch 451/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 164.2091 - mean_squared_error: 164.5974\n",
            "Epoch 00451: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 946ms/step - loss: 164.2091 - mean_squared_error: 164.5974 - val_loss: 270.5662 - val_mean_squared_error: 282.8685\n",
            "Epoch 452/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 174.0086 - mean_squared_error: 163.3890\n",
            "Epoch 00452: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 949ms/step - loss: 174.0086 - mean_squared_error: 163.3890 - val_loss: 307.7018 - val_mean_squared_error: 328.0051\n",
            "Epoch 453/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 147.4791 - mean_squared_error: 145.8742\n",
            "Epoch 00453: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 950ms/step - loss: 147.4791 - mean_squared_error: 145.8742 - val_loss: 322.0950 - val_mean_squared_error: 329.1659\n",
            "Epoch 454/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 153.1275 - mean_squared_error: 154.4428\n",
            "Epoch 00454: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 942ms/step - loss: 153.1275 - mean_squared_error: 154.4428 - val_loss: 260.4854 - val_mean_squared_error: 273.3042\n",
            "Epoch 455/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 160.8642 - mean_squared_error: 158.9069\n",
            "Epoch 00455: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 956ms/step - loss: 160.8642 - mean_squared_error: 158.9069 - val_loss: 283.0504 - val_mean_squared_error: 297.1667\n",
            "Epoch 456/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 167.7433 - mean_squared_error: 170.1669\n",
            "Epoch 00456: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 944ms/step - loss: 167.7433 - mean_squared_error: 170.1669 - val_loss: 297.0312 - val_mean_squared_error: 310.3626\n",
            "Epoch 457/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 158.0487 - mean_squared_error: 156.6338\n",
            "Epoch 00457: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 951ms/step - loss: 158.0487 - mean_squared_error: 156.6338 - val_loss: 235.9747 - val_mean_squared_error: 251.3463\n",
            "Epoch 458/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 152.5524 - mean_squared_error: 152.7168\n",
            "Epoch 00458: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 941ms/step - loss: 152.5524 - mean_squared_error: 152.7168 - val_loss: 243.3458 - val_mean_squared_error: 258.8690\n",
            "Epoch 459/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 139.1725 - mean_squared_error: 139.0876\n",
            "Epoch 00459: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 950ms/step - loss: 139.1725 - mean_squared_error: 139.0876 - val_loss: 299.1914 - val_mean_squared_error: 304.6737\n",
            "Epoch 460/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 173.0771 - mean_squared_error: 172.9686\n",
            "Epoch 00460: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 943ms/step - loss: 173.0771 - mean_squared_error: 172.9686 - val_loss: 267.7280 - val_mean_squared_error: 280.0616\n",
            "Epoch 461/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 150.1517 - mean_squared_error: 151.0673\n",
            "Epoch 00461: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 954ms/step - loss: 150.1517 - mean_squared_error: 151.0673 - val_loss: 297.3402 - val_mean_squared_error: 309.9610\n",
            "Epoch 462/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 163.0949 - mean_squared_error: 157.8964\n",
            "Epoch 00462: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 971ms/step - loss: 163.0949 - mean_squared_error: 157.8964 - val_loss: 395.2864 - val_mean_squared_error: 418.4110\n",
            "Epoch 463/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 157.2594 - mean_squared_error: 157.5196\n",
            "Epoch 00463: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 946ms/step - loss: 157.2594 - mean_squared_error: 157.5196 - val_loss: 275.6288 - val_mean_squared_error: 293.7052\n",
            "Epoch 464/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 150.8231 - mean_squared_error: 151.1884\n",
            "Epoch 00464: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 957ms/step - loss: 150.8231 - mean_squared_error: 151.1884 - val_loss: 359.5471 - val_mean_squared_error: 354.1350\n",
            "Epoch 465/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 165.1986 - mean_squared_error: 165.2094\n",
            "Epoch 00465: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 956ms/step - loss: 165.1986 - mean_squared_error: 165.2094 - val_loss: 335.6079 - val_mean_squared_error: 353.1530\n",
            "Epoch 466/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 137.7846 - mean_squared_error: 138.0218\n",
            "Epoch 00466: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 954ms/step - loss: 137.7846 - mean_squared_error: 138.0218 - val_loss: 267.6291 - val_mean_squared_error: 282.2177\n",
            "Epoch 467/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 138.6009 - mean_squared_error: 136.1721\n",
            "Epoch 00467: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 957ms/step - loss: 138.6009 - mean_squared_error: 136.1721 - val_loss: 378.0066 - val_mean_squared_error: 399.2529\n",
            "Epoch 468/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 133.0662 - mean_squared_error: 129.7483\n",
            "Epoch 00468: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 954ms/step - loss: 133.0662 - mean_squared_error: 129.7483 - val_loss: 634.1104 - val_mean_squared_error: 652.8398\n",
            "Epoch 469/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 143.7743 - mean_squared_error: 141.7001\n",
            "Epoch 00469: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 953ms/step - loss: 143.7743 - mean_squared_error: 141.7001 - val_loss: 346.2969 - val_mean_squared_error: 352.6180\n",
            "Epoch 470/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 143.6857 - mean_squared_error: 142.3093\n",
            "Epoch 00470: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 936ms/step - loss: 143.6857 - mean_squared_error: 142.3093 - val_loss: 294.2446 - val_mean_squared_error: 296.3331\n",
            "Epoch 471/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 154.0419 - mean_squared_error: 154.3881\n",
            "Epoch 00471: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 931ms/step - loss: 154.0419 - mean_squared_error: 154.3881 - val_loss: 319.6007 - val_mean_squared_error: 337.4927\n",
            "Epoch 472/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 154.7228 - mean_squared_error: 155.8409\n",
            "Epoch 00472: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 934ms/step - loss: 154.7228 - mean_squared_error: 155.8409 - val_loss: 352.4016 - val_mean_squared_error: 371.9601\n",
            "Epoch 473/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 136.6995 - mean_squared_error: 137.5984\n",
            "Epoch 00473: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 941ms/step - loss: 136.6995 - mean_squared_error: 137.5984 - val_loss: 488.2791 - val_mean_squared_error: 510.6227\n",
            "Epoch 474/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 153.8053 - mean_squared_error: 156.5692\n",
            "Epoch 00474: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 941ms/step - loss: 153.8053 - mean_squared_error: 156.5692 - val_loss: 243.6594 - val_mean_squared_error: 258.4261\n",
            "Epoch 475/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 162.8869 - mean_squared_error: 161.3376\n",
            "Epoch 00475: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 933ms/step - loss: 162.8869 - mean_squared_error: 161.3376 - val_loss: 286.9118 - val_mean_squared_error: 304.9199\n",
            "Epoch 476/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 141.6057 - mean_squared_error: 143.1694\n",
            "Epoch 00476: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 938ms/step - loss: 141.6057 - mean_squared_error: 143.1694 - val_loss: 271.5753 - val_mean_squared_error: 286.1103\n",
            "Epoch 477/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 132.4678 - mean_squared_error: 133.6797\n",
            "Epoch 00477: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 937ms/step - loss: 132.4678 - mean_squared_error: 133.6797 - val_loss: 334.1638 - val_mean_squared_error: 358.5871\n",
            "Epoch 478/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 153.0698 - mean_squared_error: 154.2334\n",
            "Epoch 00478: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 935ms/step - loss: 153.0698 - mean_squared_error: 154.2334 - val_loss: 273.2580 - val_mean_squared_error: 290.5739\n",
            "Epoch 479/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 144.2370 - mean_squared_error: 137.5331\n",
            "Epoch 00479: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 951ms/step - loss: 144.2370 - mean_squared_error: 137.5331 - val_loss: 292.2021 - val_mean_squared_error: 311.3748\n",
            "Epoch 480/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 125.8041 - mean_squared_error: 126.5320\n",
            "Epoch 00480: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 943ms/step - loss: 125.8041 - mean_squared_error: 126.5320 - val_loss: 273.6353 - val_mean_squared_error: 293.1053\n",
            "Epoch 481/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 128.9278 - mean_squared_error: 129.1080\n",
            "Epoch 00481: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 947ms/step - loss: 128.9278 - mean_squared_error: 129.1080 - val_loss: 312.2341 - val_mean_squared_error: 332.3187\n",
            "Epoch 482/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 138.5482 - mean_squared_error: 139.6729\n",
            "Epoch 00482: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 948ms/step - loss: 138.5482 - mean_squared_error: 139.6729 - val_loss: 388.3124 - val_mean_squared_error: 409.8266\n",
            "Epoch 483/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 130.0225 - mean_squared_error: 129.2282\n",
            "Epoch 00483: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 944ms/step - loss: 130.0225 - mean_squared_error: 129.2282 - val_loss: 361.3040 - val_mean_squared_error: 379.4296\n",
            "Epoch 484/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 137.7548 - mean_squared_error: 139.0692\n",
            "Epoch 00484: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 935ms/step - loss: 137.7548 - mean_squared_error: 139.0692 - val_loss: 225.1458 - val_mean_squared_error: 238.4815\n",
            "Epoch 485/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 138.8027 - mean_squared_error: 133.7336\n",
            "Epoch 00485: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 960ms/step - loss: 138.8027 - mean_squared_error: 133.7336 - val_loss: 399.8389 - val_mean_squared_error: 415.7512\n",
            "Epoch 486/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 128.3664 - mean_squared_error: 127.6811\n",
            "Epoch 00486: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 942ms/step - loss: 128.3664 - mean_squared_error: 127.6811 - val_loss: 620.9330 - val_mean_squared_error: 616.2828\n",
            "Epoch 487/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 175.8382 - mean_squared_error: 162.7218\n",
            "Epoch 00487: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 951ms/step - loss: 175.8382 - mean_squared_error: 162.7218 - val_loss: 351.4202 - val_mean_squared_error: 375.7088\n",
            "Epoch 488/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 112.8262 - mean_squared_error: 112.3948\n",
            "Epoch 00488: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 951ms/step - loss: 112.8262 - mean_squared_error: 112.3948 - val_loss: 277.3753 - val_mean_squared_error: 299.4504\n",
            "Epoch 489/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 140.6629 - mean_squared_error: 142.0439\n",
            "Epoch 00489: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 943ms/step - loss: 140.6629 - mean_squared_error: 142.0439 - val_loss: 277.1626 - val_mean_squared_error: 289.7689\n",
            "Epoch 490/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 132.6839 - mean_squared_error: 131.1644\n",
            "Epoch 00490: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 948ms/step - loss: 132.6839 - mean_squared_error: 131.1644 - val_loss: 301.1300 - val_mean_squared_error: 319.9153\n",
            "Epoch 491/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 122.5341 - mean_squared_error: 119.7260\n",
            "Epoch 00491: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 945ms/step - loss: 122.5341 - mean_squared_error: 119.7260 - val_loss: 274.4301 - val_mean_squared_error: 285.4377\n",
            "Epoch 492/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 141.4998 - mean_squared_error: 138.9016\n",
            "Epoch 00492: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 935ms/step - loss: 141.4998 - mean_squared_error: 138.9016 - val_loss: 406.5369 - val_mean_squared_error: 402.3699\n",
            "Epoch 493/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 148.4946 - mean_squared_error: 146.4575\n",
            "Epoch 00493: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 24s 941ms/step - loss: 148.4946 - mean_squared_error: 146.4575 - val_loss: 425.1089 - val_mean_squared_error: 420.8681\n",
            "Epoch 494/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 140.6753 - mean_squared_error: 142.0039\n",
            "Epoch 00494: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 961ms/step - loss: 140.6753 - mean_squared_error: 142.0039 - val_loss: 289.7095 - val_mean_squared_error: 309.7576\n",
            "Epoch 495/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 122.6265 - mean_squared_error: 123.7443\n",
            "Epoch 00495: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 946ms/step - loss: 122.6265 - mean_squared_error: 123.7443 - val_loss: 257.3329 - val_mean_squared_error: 271.8021\n",
            "Epoch 496/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 143.6557 - mean_squared_error: 145.6908\n",
            "Epoch 00496: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 966ms/step - loss: 143.6557 - mean_squared_error: 145.6908 - val_loss: 238.7389 - val_mean_squared_error: 254.3334\n",
            "Epoch 497/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 145.8327 - mean_squared_error: 146.1854\n",
            "Epoch 00497: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 962ms/step - loss: 145.8327 - mean_squared_error: 146.1854 - val_loss: 341.5936 - val_mean_squared_error: 349.4346\n",
            "Epoch 498/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 114.7505 - mean_squared_error: 111.0373\n",
            "Epoch 00498: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 953ms/step - loss: 114.7505 - mean_squared_error: 111.0373 - val_loss: 424.9518 - val_mean_squared_error: 447.4769\n",
            "Epoch 499/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 140.7661 - mean_squared_error: 140.0040\n",
            "Epoch 00499: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 954ms/step - loss: 140.7661 - mean_squared_error: 140.0040 - val_loss: 260.6055 - val_mean_squared_error: 279.5025\n",
            "Epoch 500/500\n",
            "26/26 [==============================] - ETA: 0s - loss: 126.6075 - mean_squared_error: 128.6577\n",
            "Epoch 00500: saving model to gdrive/My Drive/training/cpbn_mix_pos500.ckpt\n",
            "26/26 [==============================] - 25s 951ms/step - loss: 126.6075 - mean_squared_error: 128.6577 - val_loss: 230.2127 - val_mean_squared_error: 245.2628\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-vY-Y2LBytH2",
        "colab_type": "code",
        "outputId": "e0f09a29-1b1f-43a0-fe4b-43aa052c5d13",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        }
      },
      "source": [
        "\n",
        "plt.plot(history.history['loss'],label='mse_train')\n",
        "plt.plot(history.history['val_loss'],label='mse_val')\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Loss')\n",
        "# plt.ylim(0,15000)\n",
        "plt.legend(loc='upper right')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f1f02b569b0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Sz3S_tgF8weY",
        "colab_type": "code",
        "outputId": "375b3692-5303-494d-fe52-18e12af1c1bb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print(testY[2])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "94.0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Drr2bgO31ajg",
        "colab_type": "code",
        "outputId": "46a3a2fb-4029-4807-9529-9c610814e6a1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_pred=model.predict(testX[2].reshape(1,200,200,1))\n",
        "print(y_pred)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[96.36484]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "clGNuX2nmg9C",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!mkdir -p save_model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kdH5ALJdDpe-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.save('gdrive/My Drive/save_model/posbnmix_model_500.h5')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oCOLWBaum_9g",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "new_model=load_model('gdrive/My Drive/save_model/posbnmix_model_500.h5')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7cECixLynUQA",
        "colab_type": "code",
        "outputId": "fe4b6d1f-49f4-4e6c-fc49-1368851600d0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 595
        }
      },
      "source": [
        "new_model.summary()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 200, 200, 16)      1952      \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 100, 100, 16)      0         \n",
            "_________________________________________________________________\n",
            "batch_normalization (BatchNo (None, 100, 100, 16)      400       \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 100, 100, 32)      25120     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 50, 50, 32)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_1 (Batch (None, 50, 50, 32)        200       \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 50, 50, 64)        51264     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 25, 25, 64)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 25, 25, 64)        100       \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 40000)             0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 1000)              40001000  \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 100)               100100    \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 1)                 101       \n",
            "=================================================================\n",
            "Total params: 40,180,237\n",
            "Trainable params: 40,179,887\n",
            "Non-trainable params: 350\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IFSY-msk_GJu",
        "colab_type": "code",
        "outputId": "9f84c340-675c-4b7d-9791-a8660c24894e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "new_model.evaluate(x=X_rest,y=Y_rest)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "40/40 [==============================] - 2s 60ms/step - loss: 220.2052 - mean_squared_error: 220.4419\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[220.2052459716797, 220.44189453125]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "abwXdN0kiKpq",
        "colab_type": "code",
        "outputId": "2539b5a4-24e3-49ea-fa2e-c2afb4604b1e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "new_model.evaluate(x=testX,y=testY)\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "8/8 [==============================] - 0s 54ms/step - loss: 285.7157 - mean_squared_error: 285.7157\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[285.7156982421875, 285.7156982421875]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EfWHlJLFfJW1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predicted= new_model.predict(X_rest)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dvbb236PiaJh",
        "colab_type": "code",
        "outputId": "24e8e20c-b656-4dbf-f264-fc881c55cf28",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        }
      },
      "source": [
        "plt.scatter(range(len(X_rest)),Y_rest,label='target_value')\n",
        "plt.scatter(range(len(X_rest)),predicted,label='predicted_value')\n",
        "plt.xlabel('Sample number')\n",
        "plt.ylabel('Dimension in cm')\n",
        "plt.legend(loc='upper right')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f9aa7913a20>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dv8Fbh_clZNQ",
        "colab_type": "code",
        "outputId": "324e6001-a764-441a-8a93-4536c24427a1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 187
        }
      },
      "source": [
        "y_pred_10=new_model.predict(X_rest[10].reshape(1,200,200,1))\n",
        "y_pred_97=new_model.predict(X_rest[97].reshape(1,200,200,1))\n",
        "y_pred_1000=new_model.predict(X_rest[1000].reshape(1,200,200,1))\n",
        "y_pred_579=new_model.predict(X_rest[579].reshape(1,200,200,1))\n",
        "y_pred_477=new_model.predict(X_rest[477].reshape(1,200,200,1))\n",
        "y_pred_919=new_model.predict(X_rest[919].reshape(1,200,200,1))\n",
        "y_pred_176=new_model.predict(X_rest[176].reshape(1,200,200,1))\n",
        "y_pred_213=new_model.predict(X_rest[213].reshape(1,200,200,1))\n",
        "y_pred_845=new_model.predict(X_rest[845].reshape(1,200,200,1))\n",
        "y_pred_623=new_model.predict(X_rest[623].reshape(1,200,200,1))\n",
        "print(Y_rest[10],y_pred_10)\n",
        "print(Y_rest[97],y_pred_97)\n",
        "print(Y_rest[1000],y_pred_1000)\n",
        "print(Y_rest[579],y_pred_579)\n",
        "print(Y_rest[477],y_pred_477)\n",
        "print(Y_rest[919],y_pred_919)\n",
        "print(Y_rest[176],y_pred_176)\n",
        "print(Y_rest[213],y_pred_213)\n",
        "print(Y_rest[845],y_pred_845)\n",
        "print(Y_rest[623],y_pred_623)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "160.0 [[170.33775]]\n",
            "111.0 [[118.13747]]\n",
            "160.0 [[154.977]]\n",
            "66.0 [[62.442207]]\n",
            "124.0 [[106.062256]]\n",
            "73.0 [[92.72249]]\n",
            "72.0 [[77.17134]]\n",
            "51.0 [[51.96319]]\n",
            "61.0 [[63.658054]]\n",
            "95.0 [[115.52521]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XxyIJPyiiSZA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "predicted=new_model.predict(testX)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ac0G8YpxfXt_",
        "colab_type": "code",
        "outputId": "4bbe663f-0837-4a94-8228-25995f890f9e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        }
      },
      "source": [
        "plt.scatter(range(len(testX)),testY,label='target_value')\n",
        "plt.scatter(range(len(testX)),predicted,label='predicted_value')\n",
        "plt.xlabel('Sample number')\n",
        "plt.ylabel('Dimension in cm')\n",
        "plt.legend(loc='upper right')"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7fe1dd7454a8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dpf23fLXr0y9",
        "colab_type": "code",
        "outputId": "50eca2f1-c969-4c36-bc3d-193572b06bea",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_pred=new_model.predict(testX[2].reshape(1,200,200,1))\n",
        "print(y_pred)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[120.06555]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Iv5VrKFI__Nb",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        },
        "outputId": "3d55b4e6-28e0-43af-8228-d26b7dec7daa"
      },
      "source": [
        "new_model2=load_model('gdrive/My Drive/save_model/positive_model_500.h5')\n",
        "new_model2.summary()"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 200, 200, 16)      1952      \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 100, 100, 16)      0         \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 100, 100, 32)      25120     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 50, 50, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 50, 50, 64)        51264     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_2 (MaxPooling2 (None, 25, 25, 64)        0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 40000)             0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 1000)              40001000  \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 100)               100100    \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 1)                 101       \n",
            "=================================================================\n",
            "Total params: 40,179,537\n",
            "Trainable params: 40,179,537\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZdeiLlG1AVo3",
        "colab_type": "code",
        "outputId": "c71c302c-2042-4a52-99f9-b4c53dc4b73e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "new_model2.evaluate(x=testX,y=testY)"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "8/8 [==============================] - 0s 52ms/step - loss: 392.2873 - mean_squared_error: 392.2873\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[392.28729248046875, 392.28729248046875]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    }
  ]
}