08. Computer Vision - Face Mask Segmentation/Project 8 - CV_Face_Mask_Segmentation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Y0-xd6qKKA2L"
   },
   "source": [
    "## Project Description\n",
    "\n",
    "A hands-on project where the goal is to build a Face Mask Segmentation model which includes building a face detector to locate the position of a face in an image.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "s6E0D4USKAmr"
   },
   "source": [
    "## Data Description\n",
    "\n",
    "**WIDER Face Dataset**\n",
    "\n",
    "WIDER FACE dataset is a Face Mask Segmentation benchmark dataset, in which images are selected from the publicly available WIDER dataset. This data have 32,203 images and 393,703 faces are labeled with a high degree of variability in scale, pose and occlusion as depicted in the sample images.\n",
    "In this project, we are using 409 images and around 1000 faces for ease of computation.\n",
    "\n",
    "We will be using transfer learning on an already trained model to build our segmenter. We will perform **transfer learning** on the **MobileNet** model which is already trained to perform image segmentation. We will need to train the last 6-7 layers and freeze the remaining layers to train the model for face mask segmentation. \n",
    "\n",
    "The output of the model is the face mask segmented data which masks the face in an image. We learn to build a face mask segmentation model using Keras supported by Tensorflow.\n",
    "\n",
    "**Acknowledgment for the datasets**\n",
    "\n",
    "Acknowledgement:  http://mmlab.ie.cuhk.edu.hk/projects/WIDERFace/\n",
    "\n",
    "Mobile Net paper: https://arxiv.org/pdf/1704.04861.pdf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JHYYilH4KAaJ"
   },
   "source": [
    "## Objective\n",
    "\n",
    "To use \"Transfer Learning\" of an Object Detector model to detect any object according to the problem in hand. Here, we are particularly interested in detecting faces in a given image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "id": "KXPH8mrEDTei",
    "outputId": "e8ec7f70-4af3-4f2c-b83a-3d6463e8ad92"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.google.colaboratory.intrinsic+json": {
       "type": "string"
      },
      "text/plain": [
       "'2.3.0'"
      ]
     },
     "execution_count": 1,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%tensorflow_version 2.x\n",
    "import tensorflow\n",
    "tensorflow.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "DvJfgBYjDb0W",
    "outputId": "6a37fb60-daf1-42d8-9872-849080b76e80"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mounted at /content/gdrive\n"
     ]
    }
   ],
   "source": [
    "# Mounting google drive\n",
    "\n",
    "from google.colab import drive\n",
    "drive.mount(\"/content/gdrive\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "nFBKvsTXQyNG"
   },
   "source": [
    "#### Import Packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "5xG1-r2RDWYb"
   },
   "outputs": [],
   "source": [
    "# Imports\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Create features and labels\n",
    "from tensorflow.keras.applications.mobilenet import preprocess_input\n",
    "import cv2\n",
    "\n",
    "# Model\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.models import Model\n",
    "from tensorflow.keras.layers import Concatenate, UpSampling2D, Conv2D, Reshape, Activation, BatchNormalization, SpatialDropout2D\n",
    "from tensorflow.keras.applications.mobilenet import MobileNet\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping, ReduceLROnPlateau\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# To define Loss\n",
    "from tensorflow.keras.losses import binary_crossentropy\n",
    "from tensorflow.keras.backend import log, epsilon"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "fIXBkJDsSj2X"
   },
   "source": [
    "#### Load the 'images.npy' file\n",
    "\n",
    "\n",
    "\n",
    "*   This file contains images with details of bounding boxes\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "id": "yVGZMPG9Diwz"
   },
   "outputs": [],
   "source": [
    "# Reference: https://stackoverflow.com/questions/55890813/how-to-fix-object-arrays-cannot-be-loaded-when-allow-pickle-false-for-imdb-loa\n",
    "np_load_old = np.load\n",
    "\n",
    "# modify the default parameters of np.load\n",
    "np.load = lambda *a,**k: np_load_old(*a, allow_pickle = True, **k)\n",
    "\n",
    "face_data = np.load('/content/gdrive/My Drive/AIML/Projects/images.npy') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ZdFSqR0BHrBu",
    "outputId": "902d50c7-d88a-4e7a-ce5e-612c42cb8a52"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(409, 2)"
      ]
     },
     "execution_count": 5,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "face_data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "PUsC7lpMT7qq"
   },
   "source": [
    "Thus, the dataset contains 409 images."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "H_3s1ZuxUEde"
   },
   "source": [
    "#### Check one sample from the loaded \"images.npy\" file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 422
    },
    "id": "7SI82WM0UQCx",
    "outputId": "e1402ef1-313a-4632-890a-ae577eb0e12f"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x518.4 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize = (15, 7.2))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "plt.axis('off')\n",
    "plt.imshow(face_data[10][0])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "16XaiZG2Uv0h"
   },
   "source": [
    "#### Set image dimensions\n",
    "\n",
    "\n",
    "*   Initialize image height and image width with value: 224\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "184MdvcgE3Z_"
   },
   "outputs": [],
   "source": [
    "ALPHA = 1\n",
    "\n",
    "IMAGE_SIZE = 224\n",
    "IMAGE_HEIGHT = 224\n",
    "IMAGE_WIDTH = 224"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9ysrtqWoVKOo"
   },
   "source": [
    "#### Create features and labels\n",
    "\n",
    "\n",
    "*   Here feature is the image\n",
    "*   The label is the mask\n",
    "*   Images will be stored in \"X\" array\n",
    "*   Masks will be stored in \"masks\" array\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "id": "lYDv2V0EH0bp"
   },
   "outputs": [],
   "source": [
    "masks = np.zeros((int(face_data.shape[0]), IMAGE_HEIGHT, IMAGE_WIDTH))\n",
    "X = np.zeros((int(face_data.shape[0]), IMAGE_HEIGHT, IMAGE_WIDTH, 3))\n",
    "for index in range(face_data.shape[0]):\n",
    "  img = face_data[index][0]\n",
    "  img = cv2.resize(img, dsize = (IMAGE_HEIGHT, IMAGE_WIDTH), interpolation = cv2.INTER_CUBIC)\n",
    "  try:\n",
    "    img = img[:, :, :3]\n",
    "  except:\n",
    "    continue\n",
    "  X[index] = preprocess_input(np.array(img, dtype = np.float32))\n",
    "  for i in face_data[index][1]:\n",
    "    x1 = int(i['points'][0]['x'] * IMAGE_HEIGHT)\n",
    "    x2 = int(i['points'][1]['x'] * IMAGE_WIDTH)\n",
    "    y1 = int(i['points'][0]['y'] * IMAGE_HEIGHT)\n",
    "    y2 = int(i['points'][1]['y'] * IMAGE_WIDTH)\n",
    "    masks[index][y1:y2, x1:x2] = 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Nb8XtXIwHnVV"
   },
   "source": [
    "#### Print the shape of X and mask array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "VOH1F8y0fsBj",
    "outputId": "ebba3a15-1a37-43df-b953-43bb7f4290ad"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(409, 224, 224, 3)"
      ]
     },
     "execution_count": 9,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "VZnlw3iEfsKe",
    "outputId": "55cd474c-1a9b-40ca-f5f9-02e644c27ff1"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(409, 224, 224)"
      ]
     },
     "execution_count": 10,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "masks.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "060MmODoIGvA"
   },
   "source": [
    "#### Print a sample training image, image array and its mask"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wvs8iEyBIlfB"
   },
   "source": [
    "Print the image and image array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 439
    },
    "id": "w14lGhLBfsd6",
    "outputId": "3404fb25-d7a5-4927-fa03-2cdd8807508e"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x518.4 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = 10\n",
    "fig = plt.figure(figsize = (15, 7.2))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "plt.axis('off')\n",
    "_ = plt.imshow(X[n])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 422
    },
    "id": "w-Bwq_KJfsmg",
    "outputId": "e5d4d6c7-d2ce-474b-ded6-154ac49f6dc9"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x518.4 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = 10\n",
    "fig = plt.figure(figsize = (15, 7.2))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "plt.axis('off')\n",
    "_ = plt.imshow(masks[n])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BwA4VyACJVo7"
   },
   "source": [
    "#### Create the model\n",
    "- Add MobileNet as model with below parameter values\n",
    "  - input_shape: IMAGE_HEIGHT, IMAGE_WIDTH, 3\n",
    "  - include_top: False\n",
    "  - alpha: 1.0\n",
    "  - weights: \"imagenet\"\n",
    "- Add UNET architecture layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "96Ot4Wolfsyv"
   },
   "outputs": [],
   "source": [
    "def conv_block_simple(prevlayer, filters, prefix, strides=(1, 1)):\n",
    "    conv = Conv2D(filters, (3, 3), padding = 'same', kernel_initializer = 'he_normal', strides = strides, name = prefix + '_conv')(prevlayer)\n",
    "    conv = BatchNormalization(name = prefix + 'BatchNormalization')(conv)\n",
    "    conv = Activation('relu', name = prefix + 'ActivationLayer')(conv)\n",
    "    return conv\n",
    "\n",
    "def create_model(trainable = True):\n",
    "    model = MobileNet(input_shape = (IMAGE_HEIGHT, IMAGE_WIDTH, 3), include_top = False, alpha = ALPHA, weights = 'imagenet')\n",
    "    for layer in model.layers:\n",
    "        layer.trainable = trainable\n",
    "    \n",
    "    block1 = model.get_layer('conv_pw_13_relu').output\n",
    "    block2 = model.get_layer('conv_pw_11_relu').output\n",
    "    block3 = model.get_layer('conv_pw_5_relu').output\n",
    "    block4 = model.get_layer('conv_pw_3_relu').output\n",
    "    block5 = model.get_layer('conv_pw_1_relu').output\n",
    "    \n",
    "    up1 = Concatenate()([UpSampling2D()(block1), block2])\n",
    "    conv6 = conv_block_simple(up1, 256, 'Conv_6_1')\n",
    "    conv6 = conv_block_simple(conv6, 256, 'Conv_6_2')\n",
    "\n",
    "    up2 = Concatenate()([UpSampling2D()(conv6), block3])\n",
    "    conv7 = conv_block_simple(up2, 256, 'Conv_7_1')\n",
    "    conv7 = conv_block_simple(conv7, 256, 'Conv_7_2')\n",
    "\n",
    "    up3 = Concatenate()([UpSampling2D()(conv7), block4])\n",
    "    conv8 = conv_block_simple(up3, 192, 'Conv_8_1')\n",
    "    conv8 = conv_block_simple(conv8, 128, 'Conv_8_2')\n",
    "\n",
    "    up4 = Concatenate()([UpSampling2D()(conv8), block5])\n",
    "    conv9 = conv_block_simple(up4, 96, 'Conv_9_1')\n",
    "    conv9 = conv_block_simple(conv9, 64, 'Conv_9_2')\n",
    "\n",
    "    up5 = Concatenate()([UpSampling2D()(conv9), model.input])\n",
    "    conv10 = conv_block_simple(up5, 48, 'Conv_10_1')\n",
    "    conv10 = conv_block_simple(conv10, 32, 'Conv_10_2')\n",
    "    conv10 = SpatialDropout2D(0.2)(conv10)\n",
    "    \n",
    "    x = Conv2D(1, (1, 1), activation = 'sigmoid')(conv10)\n",
    "    x = Reshape((IMAGE_SIZE, IMAGE_SIZE))(x)\n",
    "    return Model(inputs = model.input, outputs = x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Dtna1E6of5gM"
   },
   "source": [
    "#### Call the create_model function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "7xIIGtEkfs9T",
    "outputId": "20351701-8a1a-4a42-cf21-7d7f393ab335"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/mobilenet/mobilenet_1_0_224_tf_no_top.h5\n",
      "17227776/17225924 [==============================] - 1s 0us/step\n"
     ]
    }
   ],
   "source": [
    "model = create_model(True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "MlWBXtLggNSA"
   },
   "source": [
    "#### Print model summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "jvNKsvW9fs6I",
    "outputId": "da00dca1-15ee-4c72-98e6-8da3df292ffc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"functional_1\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            [(None, 224, 224, 3) 0                                            \n",
      "__________________________________________________________________________________________________\n",
      "conv1_pad (ZeroPadding2D)       (None, 225, 225, 3)  0           input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv1 (Conv2D)                  (None, 112, 112, 32) 864         conv1_pad[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv1_bn (BatchNormalization)   (None, 112, 112, 32) 128         conv1[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "conv1_relu (ReLU)               (None, 112, 112, 32) 0           conv1_bn[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_1 (DepthwiseConv2D)     (None, 112, 112, 32) 288         conv1_relu[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_1_bn (BatchNormalizatio (None, 112, 112, 32) 128         conv_dw_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_1_relu (ReLU)           (None, 112, 112, 32) 0           conv_dw_1_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_1 (Conv2D)              (None, 112, 112, 64) 2048        conv_dw_1_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_1_bn (BatchNormalizatio (None, 112, 112, 64) 256         conv_pw_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_1_relu (ReLU)           (None, 112, 112, 64) 0           conv_pw_1_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pad_2 (ZeroPadding2D)      (None, 113, 113, 64) 0           conv_pw_1_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_2 (DepthwiseConv2D)     (None, 56, 56, 64)   576         conv_pad_2[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_2_bn (BatchNormalizatio (None, 56, 56, 64)   256         conv_dw_2[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_2_relu (ReLU)           (None, 56, 56, 64)   0           conv_dw_2_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_2 (Conv2D)              (None, 56, 56, 128)  8192        conv_dw_2_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_2_bn (BatchNormalizatio (None, 56, 56, 128)  512         conv_pw_2[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_2_relu (ReLU)           (None, 56, 56, 128)  0           conv_pw_2_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_3 (DepthwiseConv2D)     (None, 56, 56, 128)  1152        conv_pw_2_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_3_bn (BatchNormalizatio (None, 56, 56, 128)  512         conv_dw_3[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_3_relu (ReLU)           (None, 56, 56, 128)  0           conv_dw_3_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_3 (Conv2D)              (None, 56, 56, 128)  16384       conv_dw_3_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_3_bn (BatchNormalizatio (None, 56, 56, 128)  512         conv_pw_3[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_3_relu (ReLU)           (None, 56, 56, 128)  0           conv_pw_3_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pad_4 (ZeroPadding2D)      (None, 57, 57, 128)  0           conv_pw_3_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_4 (DepthwiseConv2D)     (None, 28, 28, 128)  1152        conv_pad_4[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_4_bn (BatchNormalizatio (None, 28, 28, 128)  512         conv_dw_4[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_4_relu (ReLU)           (None, 28, 28, 128)  0           conv_dw_4_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_4 (Conv2D)              (None, 28, 28, 256)  32768       conv_dw_4_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_4_bn (BatchNormalizatio (None, 28, 28, 256)  1024        conv_pw_4[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_4_relu (ReLU)           (None, 28, 28, 256)  0           conv_pw_4_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_5 (DepthwiseConv2D)     (None, 28, 28, 256)  2304        conv_pw_4_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_5_bn (BatchNormalizatio (None, 28, 28, 256)  1024        conv_dw_5[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_5_relu (ReLU)           (None, 28, 28, 256)  0           conv_dw_5_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_5 (Conv2D)              (None, 28, 28, 256)  65536       conv_dw_5_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_5_bn (BatchNormalizatio (None, 28, 28, 256)  1024        conv_pw_5[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_5_relu (ReLU)           (None, 28, 28, 256)  0           conv_pw_5_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pad_6 (ZeroPadding2D)      (None, 29, 29, 256)  0           conv_pw_5_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_6 (DepthwiseConv2D)     (None, 14, 14, 256)  2304        conv_pad_6[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_6_bn (BatchNormalizatio (None, 14, 14, 256)  1024        conv_dw_6[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_6_relu (ReLU)           (None, 14, 14, 256)  0           conv_dw_6_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_6 (Conv2D)              (None, 14, 14, 512)  131072      conv_dw_6_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_6_bn (BatchNormalizatio (None, 14, 14, 512)  2048        conv_pw_6[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_6_relu (ReLU)           (None, 14, 14, 512)  0           conv_pw_6_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_7 (DepthwiseConv2D)     (None, 14, 14, 512)  4608        conv_pw_6_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_7_bn (BatchNormalizatio (None, 14, 14, 512)  2048        conv_dw_7[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_7_relu (ReLU)           (None, 14, 14, 512)  0           conv_dw_7_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_7 (Conv2D)              (None, 14, 14, 512)  262144      conv_dw_7_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_7_bn (BatchNormalizatio (None, 14, 14, 512)  2048        conv_pw_7[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_7_relu (ReLU)           (None, 14, 14, 512)  0           conv_pw_7_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_8 (DepthwiseConv2D)     (None, 14, 14, 512)  4608        conv_pw_7_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_8_bn (BatchNormalizatio (None, 14, 14, 512)  2048        conv_dw_8[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_8_relu (ReLU)           (None, 14, 14, 512)  0           conv_dw_8_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_8 (Conv2D)              (None, 14, 14, 512)  262144      conv_dw_8_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_8_bn (BatchNormalizatio (None, 14, 14, 512)  2048        conv_pw_8[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_8_relu (ReLU)           (None, 14, 14, 512)  0           conv_pw_8_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_9 (DepthwiseConv2D)     (None, 14, 14, 512)  4608        conv_pw_8_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_9_bn (BatchNormalizatio (None, 14, 14, 512)  2048        conv_dw_9[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_9_relu (ReLU)           (None, 14, 14, 512)  0           conv_dw_9_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_9 (Conv2D)              (None, 14, 14, 512)  262144      conv_dw_9_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_9_bn (BatchNormalizatio (None, 14, 14, 512)  2048        conv_pw_9[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_9_relu (ReLU)           (None, 14, 14, 512)  0           conv_pw_9_bn[0][0]               \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_10 (DepthwiseConv2D)    (None, 14, 14, 512)  4608        conv_pw_9_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_10_bn (BatchNormalizati (None, 14, 14, 512)  2048        conv_dw_10[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_10_relu (ReLU)          (None, 14, 14, 512)  0           conv_dw_10_bn[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_10 (Conv2D)             (None, 14, 14, 512)  262144      conv_dw_10_relu[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_10_bn (BatchNormalizati (None, 14, 14, 512)  2048        conv_pw_10[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_10_relu (ReLU)          (None, 14, 14, 512)  0           conv_pw_10_bn[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_11 (DepthwiseConv2D)    (None, 14, 14, 512)  4608        conv_pw_10_relu[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_11_bn (BatchNormalizati (None, 14, 14, 512)  2048        conv_dw_11[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_11_relu (ReLU)          (None, 14, 14, 512)  0           conv_dw_11_bn[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_11 (Conv2D)             (None, 14, 14, 512)  262144      conv_dw_11_relu[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_11_bn (BatchNormalizati (None, 14, 14, 512)  2048        conv_pw_11[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_11_relu (ReLU)          (None, 14, 14, 512)  0           conv_pw_11_bn[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv_pad_12 (ZeroPadding2D)     (None, 15, 15, 512)  0           conv_pw_11_relu[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_12 (DepthwiseConv2D)    (None, 7, 7, 512)    4608        conv_pad_12[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_12_bn (BatchNormalizati (None, 7, 7, 512)    2048        conv_dw_12[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_12_relu (ReLU)          (None, 7, 7, 512)    0           conv_dw_12_bn[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_12 (Conv2D)             (None, 7, 7, 1024)   524288      conv_dw_12_relu[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_12_bn (BatchNormalizati (None, 7, 7, 1024)   4096        conv_pw_12[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_12_relu (ReLU)          (None, 7, 7, 1024)   0           conv_pw_12_bn[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_13 (DepthwiseConv2D)    (None, 7, 7, 1024)   9216        conv_pw_12_relu[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_13_bn (BatchNormalizati (None, 7, 7, 1024)   4096        conv_dw_13[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_dw_13_relu (ReLU)          (None, 7, 7, 1024)   0           conv_dw_13_bn[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_13 (Conv2D)             (None, 7, 7, 1024)   1048576     conv_dw_13_relu[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_13_bn (BatchNormalizati (None, 7, 7, 1024)   4096        conv_pw_13[0][0]                 \n",
      "__________________________________________________________________________________________________\n",
      "conv_pw_13_relu (ReLU)          (None, 7, 7, 1024)   0           conv_pw_13_bn[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d (UpSampling2D)    (None, 14, 14, 1024) 0           conv_pw_13_relu[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "concatenate (Concatenate)       (None, 14, 14, 1536) 0           up_sampling2d[0][0]              \n",
      "                                                                 conv_pw_11_relu[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "Conv_6_1_conv (Conv2D)          (None, 14, 14, 256)  3539200     concatenate[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "Conv_6_1BatchNormalization (Bat (None, 14, 14, 256)  1024        Conv_6_1_conv[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_6_1ActivationLayer (Activa (None, 14, 14, 256)  0           Conv_6_1BatchNormalization[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "Conv_6_2_conv (Conv2D)          (None, 14, 14, 256)  590080      Conv_6_1ActivationLayer[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "Conv_6_2BatchNormalization (Bat (None, 14, 14, 256)  1024        Conv_6_2_conv[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_6_2ActivationLayer (Activa (None, 14, 14, 256)  0           Conv_6_2BatchNormalization[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d_1 (UpSampling2D)  (None, 28, 28, 256)  0           Conv_6_2ActivationLayer[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, 28, 28, 512)  0           up_sampling2d_1[0][0]            \n",
      "                                                                 conv_pw_5_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "Conv_7_1_conv (Conv2D)          (None, 28, 28, 256)  1179904     concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_7_1BatchNormalization (Bat (None, 28, 28, 256)  1024        Conv_7_1_conv[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_7_1ActivationLayer (Activa (None, 28, 28, 256)  0           Conv_7_1BatchNormalization[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "Conv_7_2_conv (Conv2D)          (None, 28, 28, 256)  590080      Conv_7_1ActivationLayer[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "Conv_7_2BatchNormalization (Bat (None, 28, 28, 256)  1024        Conv_7_2_conv[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_7_2ActivationLayer (Activa (None, 28, 28, 256)  0           Conv_7_2BatchNormalization[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d_2 (UpSampling2D)  (None, 56, 56, 256)  0           Conv_7_2ActivationLayer[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_2 (Concatenate)     (None, 56, 56, 384)  0           up_sampling2d_2[0][0]            \n",
      "                                                                 conv_pw_3_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "Conv_8_1_conv (Conv2D)          (None, 56, 56, 192)  663744      concatenate_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_8_1BatchNormalization (Bat (None, 56, 56, 192)  768         Conv_8_1_conv[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_8_1ActivationLayer (Activa (None, 56, 56, 192)  0           Conv_8_1BatchNormalization[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "Conv_8_2_conv (Conv2D)          (None, 56, 56, 128)  221312      Conv_8_1ActivationLayer[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "Conv_8_2BatchNormalization (Bat (None, 56, 56, 128)  512         Conv_8_2_conv[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_8_2ActivationLayer (Activa (None, 56, 56, 128)  0           Conv_8_2BatchNormalization[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d_3 (UpSampling2D)  (None, 112, 112, 128 0           Conv_8_2ActivationLayer[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_3 (Concatenate)     (None, 112, 112, 192 0           up_sampling2d_3[0][0]            \n",
      "                                                                 conv_pw_1_relu[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "Conv_9_1_conv (Conv2D)          (None, 112, 112, 96) 165984      concatenate_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_9_1BatchNormalization (Bat (None, 112, 112, 96) 384         Conv_9_1_conv[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_9_1ActivationLayer (Activa (None, 112, 112, 96) 0           Conv_9_1BatchNormalization[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "Conv_9_2_conv (Conv2D)          (None, 112, 112, 64) 55360       Conv_9_1ActivationLayer[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "Conv_9_2BatchNormalization (Bat (None, 112, 112, 64) 256         Conv_9_2_conv[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_9_2ActivationLayer (Activa (None, 112, 112, 64) 0           Conv_9_2BatchNormalization[0][0] \n",
      "__________________________________________________________________________________________________\n",
      "up_sampling2d_4 (UpSampling2D)  (None, 224, 224, 64) 0           Conv_9_2ActivationLayer[0][0]    \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_4 (Concatenate)     (None, 224, 224, 67) 0           up_sampling2d_4[0][0]            \n",
      "                                                                 input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "Conv_10_1_conv (Conv2D)         (None, 224, 224, 48) 28992       concatenate_4[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "Conv_10_1BatchNormalization (Ba (None, 224, 224, 48) 192         Conv_10_1_conv[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "Conv_10_1ActivationLayer (Activ (None, 224, 224, 48) 0           Conv_10_1BatchNormalization[0][0]\n",
      "__________________________________________________________________________________________________\n",
      "Conv_10_2_conv (Conv2D)         (None, 224, 224, 32) 13856       Conv_10_1ActivationLayer[0][0]   \n",
      "__________________________________________________________________________________________________\n",
      "Conv_10_2BatchNormalization (Ba (None, 224, 224, 32) 128         Conv_10_2_conv[0][0]             \n",
      "__________________________________________________________________________________________________\n",
      "Conv_10_2ActivationLayer (Activ (None, 224, 224, 32) 0           Conv_10_2BatchNormalization[0][0]\n",
      "__________________________________________________________________________________________________\n",
      "spatial_dropout2d (SpatialDropo (None, 224, 224, 32) 0           Conv_10_2ActivationLayer[0][0]   \n",
      "__________________________________________________________________________________________________\n",
      "conv2d (Conv2D)                 (None, 224, 224, 1)  33          spatial_dropout2d[0][0]          \n",
      "__________________________________________________________________________________________________\n",
      "reshape (Reshape)               (None, 224, 224)     0           conv2d[0][0]                     \n",
      "==================================================================================================\n",
      "Total params: 10,283,745\n",
      "Trainable params: 10,258,689\n",
      "Non-trainable params: 25,056\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Yx3PrjdZgrzT"
   },
   "source": [
    "#### Define dice coefficient function\n",
    "\n",
    "\n",
    "*   Create a function to calculate dice coefficient\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "id": "HMFHD2PEfsvL"
   },
   "outputs": [],
   "source": [
    "def dice_coefficient(y_true, y_pred):\n",
    "    numerator = 2 * tf.reduce_sum(y_true * y_pred)\n",
    "    denominator = tf.reduce_sum(y_true + y_pred)\n",
    "\n",
    "    return numerator / (denominator + tf.keras.backend.epsilon())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "19sEX_7_hhmI"
   },
   "source": [
    "#### Define loss function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "id": "uVSD-uiZfssd"
   },
   "outputs": [],
   "source": [
    "def loss(y_true, y_pred):\n",
    "    return binary_crossentropy(y_true, y_pred) - log(dice_coefficient(y_true, y_pred) + epsilon())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "MkWeRzQSiB-v"
   },
   "source": [
    "#### Compile the model\n",
    "- Complie the model using below parameters\n",
    "  - loss: use the loss function defined above\n",
    "  - optimizers: use Adam optimizer\n",
    "  - metrics: use dice_coefficient function defined above"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "id": "2GwcCTM5fsjo"
   },
   "outputs": [],
   "source": [
    "adam = Adam(lr = 1e-4, beta_1 = 0.9, beta_2 = 0.999, epsilon = None, decay = 0.0, amsgrad = False)\n",
    "model.compile(loss = loss, optimizer = adam, metrics = [dice_coefficient])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Nq__LX4YilGd"
   },
   "source": [
    "#### Define callbacks\n",
    "- Use ModelCheckpoint\n",
    "- Use EarlyStopping\n",
    "- Use ReduceLROnPlateau"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "id": "4-HPKiq_fsbO"
   },
   "outputs": [],
   "source": [
    "checkpoint = ModelCheckpoint('model_{loss:.2f}.h5', monitor = 'loss', verbose = 1, save_best_only = True, save_weights_only = True, mode = 'min', save_freq = 1)\n",
    "stop = EarlyStopping(monitor = 'loss', patience = 5, mode = 'min')\n",
    "reduce_lr = ReduceLROnPlateau(monitor = 'loss', factor = 0.2, patience = 5, min_lr = 1e-6, verbose = 1, mode = 'min')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "_ehJUiThkuTk"
   },
   "source": [
    "#### Split the data into training and testing\n",
    "- 400 images in training\n",
    "- 9 images in testing data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "id": "iszT-D0Tku7E"
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, masks, test_size = 0.022, shuffle = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "wRXWSsYrlMAK",
    "outputId": "13cf9810-096b-4cbd-da3c-fd0ca478fb0c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of images in training data =  400\n"
     ]
    }
   ],
   "source": [
    "print('Number of images in training data = ', X_train.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "k6o5JvpvlMbT",
    "outputId": "6be848bd-122e-4d9f-da90-953ffe3ed7ca"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of images in testing data =  9\n"
     ]
    }
   ],
   "source": [
    "print('Number of images in testing data = ', X_test.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "cWOeUv2ElMnR",
    "outputId": "0bda30d1-3c43-4023-9010-5918e1e59ae9"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Numner of labels in training data =  400\n"
     ]
    }
   ],
   "source": [
    "print('Numner of labels in training data = ', y_train.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "F1yBq3PtlMQC",
    "outputId": "0d87a917-30b5-420f-9567-bbcb0c29abd3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Numner of labels in testing data =  9\n"
     ]
    }
   ],
   "source": [
    "print('Numner of labels in testing data = ', y_test.shape[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kb2qSRCvkluO"
   },
   "source": [
    "#### Fit the model\n",
    "- Fit the model using below parameters\n",
    "  - epochs: you can decide\n",
    "  - batch_size: 1\n",
    "  - callbacks: use the callbacks defined above"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ltjEFuOqfsHH",
    "outputId": "b46576e7-aabc-463d-9561-dbc2b6319ee6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1;30;43mStreaming output truncated to the last 5000 lines.\u001b[0m\n",
      "307/400 [======================>.......] - ETA: 2:40 - loss: 0.1373 - dice_coefficient: 0.9405\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "308/400 [======================>.......] - ETA: 2:39 - loss: 0.1369 - dice_coefficient: 0.9406\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "309/400 [======================>.......] - ETA: 2:37 - loss: 0.1367 - dice_coefficient: 0.9407\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "310/400 [======================>.......] - ETA: 2:35 - loss: 0.1365 - dice_coefficient: 0.9407\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "311/400 [======================>.......] - ETA: 2:33 - loss: 0.1362 - dice_coefficient: 0.9408\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "312/400 [======================>.......] - ETA: 2:32 - loss: 0.1359 - dice_coefficient: 0.9409\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "313/400 [======================>.......] - ETA: 2:30 - loss: 0.1357 - dice_coefficient: 0.9410\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "314/400 [======================>.......] - ETA: 2:28 - loss: 0.1354 - dice_coefficient: 0.9411\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "315/400 [======================>.......] - ETA: 2:27 - loss: 0.1352 - dice_coefficient: 0.9412\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "316/400 [======================>.......] - ETA: 2:25 - loss: 0.1350 - dice_coefficient: 0.9412\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "317/400 [======================>.......] - ETA: 2:23 - loss: 0.1347 - dice_coefficient: 0.9413\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "318/400 [======================>.......] - ETA: 2:21 - loss: 0.1344 - dice_coefficient: 0.9413\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "319/400 [======================>.......] - ETA: 2:20 - loss: 0.1343 - dice_coefficient: 0.9414\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "320/400 [=======================>......] - ETA: 2:18 - loss: 0.1340 - dice_coefficient: 0.9414\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "321/400 [=======================>......] - ETA: 2:16 - loss: 0.1338 - dice_coefficient: 0.9415\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "322/400 [=======================>......] - ETA: 2:14 - loss: 0.1337 - dice_coefficient: 0.9414\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "323/400 [=======================>......] - ETA: 2:13 - loss: 0.1335 - dice_coefficient: 0.9415\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "324/400 [=======================>......] - ETA: 2:11 - loss: 0.1333 - dice_coefficient: 0.9415\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "325/400 [=======================>......] - ETA: 2:09 - loss: 0.1330 - dice_coefficient: 0.9416\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "326/400 [=======================>......] - ETA: 2:08 - loss: 0.1327 - dice_coefficient: 0.9417\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "327/400 [=======================>......] - ETA: 2:06 - loss: 0.1324 - dice_coefficient: 0.9418\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "328/400 [=======================>......] - ETA: 2:04 - loss: 0.1321 - dice_coefficient: 0.9419\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "329/400 [=======================>......] - ETA: 2:02 - loss: 0.1319 - dice_coefficient: 0.9419\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "330/400 [=======================>......] - ETA: 2:01 - loss: 0.1316 - dice_coefficient: 0.9420\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "331/400 [=======================>......] - ETA: 1:59 - loss: 0.1313 - dice_coefficient: 0.9421\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "332/400 [=======================>......] - ETA: 1:57 - loss: 0.1311 - dice_coefficient: 0.9421\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "333/400 [=======================>......] - ETA: 1:55 - loss: 0.1309 - dice_coefficient: 0.9421\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "334/400 [========================>.....] - ETA: 1:54 - loss: 0.1306 - dice_coefficient: 0.9422\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "335/400 [========================>.....] - ETA: 1:52 - loss: 0.1303 - dice_coefficient: 0.9423\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "336/400 [========================>.....] - ETA: 1:50 - loss: 0.1300 - dice_coefficient: 0.9424\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "337/400 [========================>.....] - ETA: 1:49 - loss: 0.1298 - dice_coefficient: 0.9425\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "338/400 [========================>.....] - ETA: 1:47 - loss: 0.1296 - dice_coefficient: 0.9425\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "339/400 [========================>.....] - ETA: 1:45 - loss: 0.1298 - dice_coefficient: 0.9424\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "340/400 [========================>.....] - ETA: 1:43 - loss: 0.1295 - dice_coefficient: 0.9425\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "341/400 [========================>.....] - ETA: 1:42 - loss: 0.1293 - dice_coefficient: 0.9425\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "342/400 [========================>.....] - ETA: 1:40 - loss: 0.1290 - dice_coefficient: 0.9427\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "343/400 [========================>.....] - ETA: 1:38 - loss: 0.1288 - dice_coefficient: 0.9428\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "344/400 [========================>.....] - ETA: 1:36 - loss: 0.1285 - dice_coefficient: 0.9428\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "345/400 [========================>.....] - ETA: 1:35 - loss: 0.1287 - dice_coefficient: 0.9426\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "346/400 [========================>.....] - ETA: 1:33 - loss: 0.1284 - dice_coefficient: 0.9427\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "347/400 [=========================>....] - ETA: 1:31 - loss: 0.1283 - dice_coefficient: 0.9426\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "348/400 [=========================>....] - ETA: 1:30 - loss: 0.1282 - dice_coefficient: 0.9426\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "349/400 [=========================>....] - ETA: 1:28 - loss: 0.1280 - dice_coefficient: 0.9427\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "350/400 [=========================>....] - ETA: 1:26 - loss: 0.1277 - dice_coefficient: 0.9428\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "351/400 [=========================>....] - ETA: 1:24 - loss: 0.1278 - dice_coefficient: 0.9428\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "352/400 [=========================>....] - ETA: 1:23 - loss: 0.1276 - dice_coefficient: 0.9428\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "353/400 [=========================>....] - ETA: 1:21 - loss: 0.1273 - dice_coefficient: 0.9429\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "354/400 [=========================>....] - ETA: 1:19 - loss: 0.1270 - dice_coefficient: 0.9430\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "355/400 [=========================>....] - ETA: 1:17 - loss: 0.1270 - dice_coefficient: 0.9429\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "356/400 [=========================>....] - ETA: 1:16 - loss: 0.1268 - dice_coefficient: 0.9429\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "357/400 [=========================>....] - ETA: 1:14 - loss: 0.1266 - dice_coefficient: 0.9430\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "358/400 [=========================>....] - ETA: 1:12 - loss: 0.1264 - dice_coefficient: 0.9431\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "359/400 [=========================>....] - ETA: 1:10 - loss: 0.1262 - dice_coefficient: 0.9431\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "360/400 [==========================>...] - ETA: 1:09 - loss: 0.1259 - dice_coefficient: 0.9432\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "361/400 [==========================>...] - ETA: 1:07 - loss: 0.1258 - dice_coefficient: 0.9432\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "362/400 [==========================>...] - ETA: 1:05 - loss: 0.1257 - dice_coefficient: 0.9431\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "363/400 [==========================>...] - ETA: 1:04 - loss: 0.1255 - dice_coefficient: 0.9432\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "364/400 [==========================>...] - ETA: 1:02 - loss: 0.1253 - dice_coefficient: 0.9432\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "365/400 [==========================>...] - ETA: 1:00 - loss: 0.1252 - dice_coefficient: 0.9432\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "366/400 [==========================>...] - ETA: 58s - loss: 0.1252 - dice_coefficient: 0.9431 \n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "367/400 [==========================>...] - ETA: 57s - loss: 0.1250 - dice_coefficient: 0.9432\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "368/400 [==========================>...] - ETA: 55s - loss: 0.1248 - dice_coefficient: 0.9432\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "369/400 [==========================>...] - ETA: 53s - loss: 0.1247 - dice_coefficient: 0.9431\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "370/400 [==========================>...] - ETA: 51s - loss: 0.1246 - dice_coefficient: 0.9432\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "371/400 [==========================>...] - ETA: 50s - loss: 0.1247 - dice_coefficient: 0.9430\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "372/400 [==========================>...] - ETA: 48s - loss: 0.1245 - dice_coefficient: 0.9430\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "373/400 [==========================>...] - ETA: 46s - loss: 0.1244 - dice_coefficient: 0.9430\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "374/400 [===========================>..] - ETA: 45s - loss: 0.1243 - dice_coefficient: 0.9430\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "375/400 [===========================>..] - ETA: 43s - loss: 0.1240 - dice_coefficient: 0.9431\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "376/400 [===========================>..] - ETA: 41s - loss: 0.1240 - dice_coefficient: 0.9431\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "377/400 [===========================>..] - ETA: 39s - loss: 0.1237 - dice_coefficient: 0.9432\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "378/400 [===========================>..] - ETA: 38s - loss: 0.1235 - dice_coefficient: 0.9433\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "379/400 [===========================>..] - ETA: 36s - loss: 0.1234 - dice_coefficient: 0.9433\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "380/400 [===========================>..] - ETA: 34s - loss: 0.1232 - dice_coefficient: 0.9434\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "381/400 [===========================>..] - ETA: 32s - loss: 0.1231 - dice_coefficient: 0.9434\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "382/400 [===========================>..] - ETA: 31s - loss: 0.1230 - dice_coefficient: 0.9433\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "383/400 [===========================>..] - ETA: 29s - loss: 0.1228 - dice_coefficient: 0.9434\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "384/400 [===========================>..] - ETA: 27s - loss: 0.1226 - dice_coefficient: 0.9434\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "385/400 [===========================>..] - ETA: 26s - loss: 0.1224 - dice_coefficient: 0.9435\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "386/400 [===========================>..] - ETA: 24s - loss: 0.1221 - dice_coefficient: 0.9436\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "387/400 [============================>.] - ETA: 22s - loss: 0.1220 - dice_coefficient: 0.9436\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "388/400 [============================>.] - ETA: 20s - loss: 0.1219 - dice_coefficient: 0.9437\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "389/400 [============================>.] - ETA: 19s - loss: 0.1217 - dice_coefficient: 0.9437\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "390/400 [============================>.] - ETA: 17s - loss: 0.1215 - dice_coefficient: 0.9437\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "391/400 [============================>.] - ETA: 15s - loss: 0.1214 - dice_coefficient: 0.9438\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "392/400 [============================>.] - ETA: 13s - loss: 0.1213 - dice_coefficient: 0.9438\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "393/400 [============================>.] - ETA: 12s - loss: 0.1211 - dice_coefficient: 0.9439\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "394/400 [============================>.] - ETA: 10s - loss: 0.1210 - dice_coefficient: 0.9438\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "395/400 [============================>.] - ETA: 8s - loss: 0.1207 - dice_coefficient: 0.9439 \n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "396/400 [============================>.] - ETA: 6s - loss: 0.1206 - dice_coefficient: 0.9439\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "397/400 [============================>.] - ETA: 5s - loss: 0.1203 - dice_coefficient: 0.9440\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "398/400 [============================>.] - ETA: 3s - loss: 0.1202 - dice_coefficient: 0.9440\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "399/400 [============================>.] - ETA: 1s - loss: 0.1200 - dice_coefficient: 0.9441\n",
      "Epoch 00024: loss did not improve from 0.04222\n",
      "400/400 [==============================] - 698s 2s/step - loss: 0.1198 - dice_coefficient: 0.9442 - val_loss: 0.9661 - val_dice_coefficient: 0.5656\n",
      "Epoch 25/30\n",
      "\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "  1/400 [..............................] - ETA: 0s - loss: 0.1258 - dice_coefficient: 0.9545\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "  2/400 [..............................] - ETA: 5:44 - loss: 0.1070 - dice_coefficient: 0.9536\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "  3/400 [..............................] - ETA: 7:38 - loss: 0.0818 - dice_coefficient: 0.9643\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "  4/400 [..............................] - ETA: 8:34 - loss: 0.0729 - dice_coefficient: 0.9630\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "  5/400 [..............................] - ETA: 9:08 - loss: 0.0726 - dice_coefficient: 0.9640\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "  6/400 [..............................] - ETA: 9:31 - loss: 0.0700 - dice_coefficient: 0.9649\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "  7/400 [..............................] - ETA: 9:47 - loss: 0.0667 - dice_coefficient: 0.9650\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "  8/400 [..............................] - ETA: 9:58 - loss: 0.0638 - dice_coefficient: 0.9650\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "  9/400 [..............................] - ETA: 10:06 - loss: 0.0611 - dice_coefficient: 0.9664\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 10/400 [..............................] - ETA: 10:11 - loss: 0.0593 - dice_coefficient: 0.9669\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 11/400 [..............................] - ETA: 10:16 - loss: 0.0586 - dice_coefficient: 0.9664\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 12/400 [..............................] - ETA: 10:19 - loss: 0.0579 - dice_coefficient: 0.9660\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 13/400 [..............................] - ETA: 10:22 - loss: 0.0572 - dice_coefficient: 0.9651\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 14/400 [>.............................] - ETA: 10:24 - loss: 0.0631 - dice_coefficient: 0.9600\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 15/400 [>.............................] - ETA: 10:25 - loss: 0.0627 - dice_coefficient: 0.9596\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 16/400 [>.............................] - ETA: 10:26 - loss: 0.0627 - dice_coefficient: 0.9602\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 17/400 [>.............................] - ETA: 10:26 - loss: 0.0622 - dice_coefficient: 0.9609\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 18/400 [>.............................] - ETA: 10:27 - loss: 0.0620 - dice_coefficient: 0.9614\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 19/400 [>.............................] - ETA: 10:28 - loss: 0.0629 - dice_coefficient: 0.9616\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 20/400 [>.............................] - ETA: 10:28 - loss: 0.0619 - dice_coefficient: 0.9620\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 21/400 [>.............................] - ETA: 10:28 - loss: 0.0623 - dice_coefficient: 0.9617\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 22/400 [>.............................] - ETA: 10:27 - loss: 0.0627 - dice_coefficient: 0.9612\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 23/400 [>.............................] - ETA: 10:28 - loss: 0.0622 - dice_coefficient: 0.9614\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 24/400 [>.............................] - ETA: 10:27 - loss: 0.0622 - dice_coefficient: 0.9610\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 25/400 [>.............................] - ETA: 10:26 - loss: 0.0635 - dice_coefficient: 0.9599\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 26/400 [>.............................] - ETA: 10:26 - loss: 0.0632 - dice_coefficient: 0.9596\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 27/400 [=>............................] - ETA: 10:25 - loss: 0.0622 - dice_coefficient: 0.9601\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 28/400 [=>............................] - ETA: 10:24 - loss: 0.0609 - dice_coefficient: 0.9610\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 29/400 [=>............................] - ETA: 10:23 - loss: 0.0603 - dice_coefficient: 0.9614\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 30/400 [=>............................] - ETA: 10:22 - loss: 0.0612 - dice_coefficient: 0.9602\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 31/400 [=>............................] - ETA: 10:21 - loss: 0.0624 - dice_coefficient: 0.9588\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 32/400 [=>............................] - ETA: 10:20 - loss: 0.0632 - dice_coefficient: 0.9580\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 33/400 [=>............................] - ETA: 10:19 - loss: 0.0642 - dice_coefficient: 0.9576\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 34/400 [=>............................] - ETA: 10:18 - loss: 0.0632 - dice_coefficient: 0.9583\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 35/400 [=>............................] - ETA: 10:17 - loss: 0.0627 - dice_coefficient: 0.9588\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 36/400 [=>............................] - ETA: 10:16 - loss: 0.0621 - dice_coefficient: 0.9592\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 37/400 [=>............................] - ETA: 10:14 - loss: 0.0627 - dice_coefficient: 0.9584\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 38/400 [=>............................] - ETA: 10:13 - loss: 0.0623 - dice_coefficient: 0.9586\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 39/400 [=>............................] - ETA: 10:12 - loss: 0.0619 - dice_coefficient: 0.9589\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 40/400 [==>...........................] - ETA: 10:11 - loss: 0.0611 - dice_coefficient: 0.9596\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 41/400 [==>...........................] - ETA: 10:09 - loss: 0.0606 - dice_coefficient: 0.9601\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 42/400 [==>...........................] - ETA: 10:08 - loss: 0.0609 - dice_coefficient: 0.9599\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 43/400 [==>...........................] - ETA: 10:06 - loss: 0.0604 - dice_coefficient: 0.9602\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 44/400 [==>...........................] - ETA: 10:05 - loss: 0.0606 - dice_coefficient: 0.9600\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 45/400 [==>...........................] - ETA: 10:04 - loss: 0.0595 - dice_coefficient: 0.9607\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 46/400 [==>...........................] - ETA: 10:03 - loss: 0.0595 - dice_coefficient: 0.9606\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 47/400 [==>...........................] - ETA: 10:01 - loss: 0.0605 - dice_coefficient: 0.9593\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 48/400 [==>...........................] - ETA: 10:00 - loss: 0.0603 - dice_coefficient: 0.9592\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 49/400 [==>...........................] - ETA: 9:59 - loss: 0.0610 - dice_coefficient: 0.9587 \n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 50/400 [==>...........................] - ETA: 9:57 - loss: 0.0614 - dice_coefficient: 0.9583\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 51/400 [==>...........................] - ETA: 9:56 - loss: 0.0617 - dice_coefficient: 0.9584\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 52/400 [==>...........................] - ETA: 9:55 - loss: 0.0630 - dice_coefficient: 0.9582\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 53/400 [==>...........................] - ETA: 9:53 - loss: 0.0641 - dice_coefficient: 0.9577\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 54/400 [===>..........................] - ETA: 9:52 - loss: 0.0641 - dice_coefficient: 0.9580\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 55/400 [===>..........................] - ETA: 9:51 - loss: 0.0644 - dice_coefficient: 0.9579\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 56/400 [===>..........................] - ETA: 9:49 - loss: 0.0643 - dice_coefficient: 0.9580\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 57/400 [===>..........................] - ETA: 9:48 - loss: 0.0644 - dice_coefficient: 0.9578\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 58/400 [===>..........................] - ETA: 9:46 - loss: 0.0646 - dice_coefficient: 0.9577\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 59/400 [===>..........................] - ETA: 9:45 - loss: 0.0640 - dice_coefficient: 0.9581\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 60/400 [===>..........................] - ETA: 9:43 - loss: 0.0639 - dice_coefficient: 0.9579\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 61/400 [===>..........................] - ETA: 9:42 - loss: 0.0653 - dice_coefficient: 0.9574\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 62/400 [===>..........................] - ETA: 9:40 - loss: 0.0656 - dice_coefficient: 0.9570\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 63/400 [===>..........................] - ETA: 9:39 - loss: 0.0656 - dice_coefficient: 0.9571\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 64/400 [===>..........................] - ETA: 9:37 - loss: 0.0653 - dice_coefficient: 0.9573\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 65/400 [===>..........................] - ETA: 9:35 - loss: 0.0656 - dice_coefficient: 0.9572\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 66/400 [===>..........................] - ETA: 9:34 - loss: 0.0660 - dice_coefficient: 0.9571\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 67/400 [====>.........................] - ETA: 9:32 - loss: 0.0654 - dice_coefficient: 0.9575\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 68/400 [====>.........................] - ETA: 9:30 - loss: 0.0650 - dice_coefficient: 0.9577\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 69/400 [====>.........................] - ETA: 9:29 - loss: 0.0648 - dice_coefficient: 0.9579\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 70/400 [====>.........................] - ETA: 9:27 - loss: 0.0650 - dice_coefficient: 0.9579\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 71/400 [====>.........................] - ETA: 9:25 - loss: 0.0653 - dice_coefficient: 0.9577\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 72/400 [====>.........................] - ETA: 9:24 - loss: 0.0653 - dice_coefficient: 0.9576\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 73/400 [====>.........................] - ETA: 9:22 - loss: 0.0653 - dice_coefficient: 0.9578\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 74/400 [====>.........................] - ETA: 9:20 - loss: 0.0659 - dice_coefficient: 0.9572\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 75/400 [====>.........................] - ETA: 9:19 - loss: 0.0658 - dice_coefficient: 0.9571\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 76/400 [====>.........................] - ETA: 9:17 - loss: 0.0655 - dice_coefficient: 0.9573\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 77/400 [====>.........................] - ETA: 9:15 - loss: 0.0651 - dice_coefficient: 0.9576\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 78/400 [====>.........................] - ETA: 9:13 - loss: 0.0649 - dice_coefficient: 0.9578\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 79/400 [====>.........................] - ETA: 9:12 - loss: 0.0651 - dice_coefficient: 0.9579\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 80/400 [=====>........................] - ETA: 9:10 - loss: 0.0652 - dice_coefficient: 0.9577\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 81/400 [=====>........................] - ETA: 9:08 - loss: 0.0654 - dice_coefficient: 0.9577\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 82/400 [=====>........................] - ETA: 9:07 - loss: 0.0653 - dice_coefficient: 0.9577\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 83/400 [=====>........................] - ETA: 9:05 - loss: 0.0658 - dice_coefficient: 0.9570\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 84/400 [=====>........................] - ETA: 9:03 - loss: 0.0661 - dice_coefficient: 0.9570\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 85/400 [=====>........................] - ETA: 9:02 - loss: 0.0659 - dice_coefficient: 0.9572\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 86/400 [=====>........................] - ETA: 9:00 - loss: 0.0667 - dice_coefficient: 0.9567\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 87/400 [=====>........................] - ETA: 8:58 - loss: 0.0665 - dice_coefficient: 0.9567\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 88/400 [=====>........................] - ETA: 8:57 - loss: 0.0660 - dice_coefficient: 0.9570\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 89/400 [=====>........................] - ETA: 8:55 - loss: 0.0659 - dice_coefficient: 0.9573\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 90/400 [=====>........................] - ETA: 8:53 - loss: 0.0660 - dice_coefficient: 0.9570\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 91/400 [=====>........................] - ETA: 8:52 - loss: 0.0659 - dice_coefficient: 0.9572\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 92/400 [=====>........................] - ETA: 8:50 - loss: 0.0658 - dice_coefficient: 0.9573\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 93/400 [=====>........................] - ETA: 8:48 - loss: 0.0660 - dice_coefficient: 0.9574\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 94/400 [======>.......................] - ETA: 8:47 - loss: 0.0662 - dice_coefficient: 0.9571\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 95/400 [======>.......................] - ETA: 8:45 - loss: 0.0659 - dice_coefficient: 0.9573\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 96/400 [======>.......................] - ETA: 8:44 - loss: 0.0658 - dice_coefficient: 0.9574\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 97/400 [======>.......................] - ETA: 8:42 - loss: 0.0674 - dice_coefficient: 0.9559\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 98/400 [======>.......................] - ETA: 8:40 - loss: 0.0673 - dice_coefficient: 0.9560\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      " 99/400 [======>.......................] - ETA: 8:39 - loss: 0.0697 - dice_coefficient: 0.9541\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "100/400 [======>.......................] - ETA: 8:37 - loss: 0.0702 - dice_coefficient: 0.9540\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "101/400 [======>.......................] - ETA: 8:35 - loss: 0.0698 - dice_coefficient: 0.9543\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "102/400 [======>.......................] - ETA: 8:34 - loss: 0.0696 - dice_coefficient: 0.9543\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "103/400 [======>.......................] - ETA: 8:32 - loss: 0.0695 - dice_coefficient: 0.9543\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "104/400 [======>.......................] - ETA: 8:30 - loss: 0.0696 - dice_coefficient: 0.9542\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "105/400 [======>.......................] - ETA: 8:28 - loss: 0.0696 - dice_coefficient: 0.9542\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "106/400 [======>.......................] - ETA: 8:27 - loss: 0.0697 - dice_coefficient: 0.9543\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "107/400 [=======>......................] - ETA: 8:25 - loss: 0.0698 - dice_coefficient: 0.9542\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "108/400 [=======>......................] - ETA: 8:23 - loss: 0.0694 - dice_coefficient: 0.9545\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "109/400 [=======>......................] - ETA: 8:22 - loss: 0.0690 - dice_coefficient: 0.9547\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "110/400 [=======>......................] - ETA: 8:20 - loss: 0.0690 - dice_coefficient: 0.9547\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "111/400 [=======>......................] - ETA: 8:18 - loss: 0.0689 - dice_coefficient: 0.9548\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "112/400 [=======>......................] - ETA: 8:16 - loss: 0.0690 - dice_coefficient: 0.9546\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "113/400 [=======>......................] - ETA: 8:15 - loss: 0.0693 - dice_coefficient: 0.9545\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "114/400 [=======>......................] - ETA: 8:13 - loss: 0.0690 - dice_coefficient: 0.9547\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "115/400 [=======>......................] - ETA: 8:11 - loss: 0.0690 - dice_coefficient: 0.9547\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "116/400 [=======>......................] - ETA: 8:10 - loss: 0.0692 - dice_coefficient: 0.9545\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "117/400 [=======>......................] - ETA: 8:08 - loss: 0.0695 - dice_coefficient: 0.9544\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "118/400 [=======>......................] - ETA: 8:06 - loss: 0.0692 - dice_coefficient: 0.9545\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "119/400 [=======>......................] - ETA: 8:05 - loss: 0.0697 - dice_coefficient: 0.9540\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "120/400 [========>.....................] - ETA: 8:03 - loss: 0.0699 - dice_coefficient: 0.9538\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "121/400 [========>.....................] - ETA: 8:01 - loss: 0.0697 - dice_coefficient: 0.9539\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "122/400 [========>.....................] - ETA: 7:59 - loss: 0.0694 - dice_coefficient: 0.9540\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "123/400 [========>.....................] - ETA: 7:58 - loss: 0.0693 - dice_coefficient: 0.9541\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "124/400 [========>.....................] - ETA: 7:56 - loss: 0.0693 - dice_coefficient: 0.9541\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "125/400 [========>.....................] - ETA: 7:54 - loss: 0.0693 - dice_coefficient: 0.9542\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "126/400 [========>.....................] - ETA: 7:53 - loss: 0.0697 - dice_coefficient: 0.9541\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "127/400 [========>.....................] - ETA: 7:51 - loss: 0.0720 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "128/400 [========>.....................] - ETA: 7:49 - loss: 0.0719 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "129/400 [========>.....................] - ETA: 7:47 - loss: 0.0719 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "130/400 [========>.....................] - ETA: 7:46 - loss: 0.0723 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "131/400 [========>.....................] - ETA: 7:44 - loss: 0.0725 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "132/400 [========>.....................] - ETA: 7:42 - loss: 0.0733 - dice_coefficient: 0.9511\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "133/400 [========>.....................] - ETA: 7:41 - loss: 0.0732 - dice_coefficient: 0.9511\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "134/400 [=========>....................] - ETA: 7:39 - loss: 0.0731 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "135/400 [=========>....................] - ETA: 7:37 - loss: 0.0733 - dice_coefficient: 0.9510\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "136/400 [=========>....................] - ETA: 7:36 - loss: 0.0731 - dice_coefficient: 0.9511\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "137/400 [=========>....................] - ETA: 7:34 - loss: 0.0730 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "138/400 [=========>....................] - ETA: 7:32 - loss: 0.0730 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "139/400 [=========>....................] - ETA: 7:30 - loss: 0.0726 - dice_coefficient: 0.9514\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "140/400 [=========>....................] - ETA: 7:29 - loss: 0.0727 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "141/400 [=========>....................] - ETA: 7:27 - loss: 0.0726 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "142/400 [=========>....................] - ETA: 7:25 - loss: 0.0727 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "143/400 [=========>....................] - ETA: 7:24 - loss: 0.0729 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "144/400 [=========>....................] - ETA: 7:22 - loss: 0.0729 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "145/400 [=========>....................] - ETA: 7:20 - loss: 0.0729 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "146/400 [=========>....................] - ETA: 7:18 - loss: 0.0731 - dice_coefficient: 0.9510\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "147/400 [==========>...................] - ETA: 7:17 - loss: 0.0728 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "148/400 [==========>...................] - ETA: 7:15 - loss: 0.0729 - dice_coefficient: 0.9511\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "149/400 [==========>...................] - ETA: 7:13 - loss: 0.0729 - dice_coefficient: 0.9511\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "150/400 [==========>...................] - ETA: 7:11 - loss: 0.0727 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "151/400 [==========>...................] - ETA: 7:10 - loss: 0.0727 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "152/400 [==========>...................] - ETA: 7:08 - loss: 0.0726 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "153/400 [==========>...................] - ETA: 7:06 - loss: 0.0723 - dice_coefficient: 0.9514\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "154/400 [==========>...................] - ETA: 7:05 - loss: 0.0722 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "155/400 [==========>...................] - ETA: 7:03 - loss: 0.0723 - dice_coefficient: 0.9514\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "156/400 [==========>...................] - ETA: 7:01 - loss: 0.0723 - dice_coefficient: 0.9514\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "157/400 [==========>...................] - ETA: 7:00 - loss: 0.0722 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "158/400 [==========>...................] - ETA: 6:58 - loss: 0.0721 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "159/400 [==========>...................] - ETA: 6:56 - loss: 0.0723 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "160/400 [===========>..................] - ETA: 6:55 - loss: 0.0721 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "161/400 [===========>..................] - ETA: 6:53 - loss: 0.0719 - dice_coefficient: 0.9518\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "162/400 [===========>..................] - ETA: 6:51 - loss: 0.0719 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "163/400 [===========>..................] - ETA: 6:50 - loss: 0.0719 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "164/400 [===========>..................] - ETA: 6:48 - loss: 0.0718 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "165/400 [===========>..................] - ETA: 6:46 - loss: 0.0720 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "166/400 [===========>..................] - ETA: 6:44 - loss: 0.0717 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "167/400 [===========>..................] - ETA: 6:43 - loss: 0.0723 - dice_coefficient: 0.9511\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "168/400 [===========>..................] - ETA: 6:41 - loss: 0.0722 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "169/400 [===========>..................] - ETA: 6:39 - loss: 0.0721 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "170/400 [===========>..................] - ETA: 6:38 - loss: 0.0721 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "171/400 [===========>..................] - ETA: 6:36 - loss: 0.0724 - dice_coefficient: 0.9510\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "172/400 [===========>..................] - ETA: 6:34 - loss: 0.0723 - dice_coefficient: 0.9510\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "173/400 [===========>..................] - ETA: 6:33 - loss: 0.0724 - dice_coefficient: 0.9509\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "174/400 [============>.................] - ETA: 6:31 - loss: 0.0722 - dice_coefficient: 0.9510\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "175/400 [============>.................] - ETA: 6:29 - loss: 0.0722 - dice_coefficient: 0.9511\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "176/400 [============>.................] - ETA: 6:27 - loss: 0.0719 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "177/400 [============>.................] - ETA: 6:26 - loss: 0.0719 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "178/400 [============>.................] - ETA: 6:24 - loss: 0.0719 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "179/400 [============>.................] - ETA: 6:22 - loss: 0.0718 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "180/400 [============>.................] - ETA: 6:21 - loss: 0.0718 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "181/400 [============>.................] - ETA: 6:19 - loss: 0.0717 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "182/400 [============>.................] - ETA: 6:17 - loss: 0.0715 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "183/400 [============>.................] - ETA: 6:15 - loss: 0.0714 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "184/400 [============>.................] - ETA: 6:14 - loss: 0.0713 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "185/400 [============>.................] - ETA: 6:12 - loss: 0.0713 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "186/400 [============>.................] - ETA: 6:10 - loss: 0.0717 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "187/400 [=============>................] - ETA: 6:09 - loss: 0.0717 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "188/400 [=============>................] - ETA: 6:07 - loss: 0.0716 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "189/400 [=============>................] - ETA: 6:05 - loss: 0.0719 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "190/400 [=============>................] - ETA: 6:03 - loss: 0.0724 - dice_coefficient: 0.9508\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "191/400 [=============>................] - ETA: 6:02 - loss: 0.0724 - dice_coefficient: 0.9507\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "192/400 [=============>................] - ETA: 6:00 - loss: 0.0722 - dice_coefficient: 0.9508\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "193/400 [=============>................] - ETA: 5:58 - loss: 0.0724 - dice_coefficient: 0.9508\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "194/400 [=============>................] - ETA: 5:57 - loss: 0.0726 - dice_coefficient: 0.9506\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "195/400 [=============>................] - ETA: 5:55 - loss: 0.0725 - dice_coefficient: 0.9507\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "196/400 [=============>................] - ETA: 5:53 - loss: 0.0725 - dice_coefficient: 0.9508\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "197/400 [=============>................] - ETA: 5:51 - loss: 0.0724 - dice_coefficient: 0.9509\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "198/400 [=============>................] - ETA: 5:50 - loss: 0.0723 - dice_coefficient: 0.9510\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "199/400 [=============>................] - ETA: 5:48 - loss: 0.0721 - dice_coefficient: 0.9512\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "200/400 [==============>...............] - ETA: 5:46 - loss: 0.0718 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "201/400 [==============>...............] - ETA: 5:45 - loss: 0.0718 - dice_coefficient: 0.9514\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "202/400 [==============>...............] - ETA: 5:43 - loss: 0.0717 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "203/400 [==============>...............] - ETA: 5:41 - loss: 0.0715 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "204/400 [==============>...............] - ETA: 5:39 - loss: 0.0716 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "205/400 [==============>...............] - ETA: 5:38 - loss: 0.0715 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "206/400 [==============>...............] - ETA: 5:36 - loss: 0.0712 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "207/400 [==============>...............] - ETA: 5:34 - loss: 0.0713 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "208/400 [==============>...............] - ETA: 5:33 - loss: 0.0712 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "209/400 [==============>...............] - ETA: 5:31 - loss: 0.0711 - dice_coefficient: 0.9518\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "210/400 [==============>...............] - ETA: 5:29 - loss: 0.0709 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "211/400 [==============>...............] - ETA: 5:27 - loss: 0.0713 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "212/400 [==============>...............] - ETA: 5:26 - loss: 0.0711 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "213/400 [==============>...............] - ETA: 5:24 - loss: 0.0709 - dice_coefficient: 0.9518\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "214/400 [===============>..............] - ETA: 5:22 - loss: 0.0707 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "215/400 [===============>..............] - ETA: 5:20 - loss: 0.0710 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "216/400 [===============>..............] - ETA: 5:19 - loss: 0.0710 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "217/400 [===============>..............] - ETA: 5:17 - loss: 0.0709 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "218/400 [===============>..............] - ETA: 5:15 - loss: 0.0709 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "219/400 [===============>..............] - ETA: 5:14 - loss: 0.0713 - dice_coefficient: 0.9513\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "220/400 [===============>..............] - ETA: 5:12 - loss: 0.0711 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "221/400 [===============>..............] - ETA: 5:10 - loss: 0.0711 - dice_coefficient: 0.9514\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "222/400 [===============>..............] - ETA: 5:08 - loss: 0.0710 - dice_coefficient: 0.9515\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "223/400 [===============>..............] - ETA: 5:07 - loss: 0.0709 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "224/400 [===============>..............] - ETA: 5:05 - loss: 0.0708 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "225/400 [===============>..............] - ETA: 5:03 - loss: 0.0707 - dice_coefficient: 0.9516\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "226/400 [===============>..............] - ETA: 5:02 - loss: 0.0706 - dice_coefficient: 0.9517\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "227/400 [================>.............] - ETA: 5:00 - loss: 0.0706 - dice_coefficient: 0.9518\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "228/400 [================>.............] - ETA: 4:58 - loss: 0.0704 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "229/400 [================>.............] - ETA: 4:56 - loss: 0.0703 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "230/400 [================>.............] - ETA: 4:55 - loss: 0.0704 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "231/400 [================>.............] - ETA: 4:53 - loss: 0.0703 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "232/400 [================>.............] - ETA: 4:51 - loss: 0.0703 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "233/400 [================>.............] - ETA: 4:50 - loss: 0.0702 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "234/400 [================>.............] - ETA: 4:48 - loss: 0.0703 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "235/400 [================>.............] - ETA: 4:46 - loss: 0.0705 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "236/400 [================>.............] - ETA: 4:44 - loss: 0.0705 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "237/400 [================>.............] - ETA: 4:43 - loss: 0.0704 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "238/400 [================>.............] - ETA: 4:41 - loss: 0.0702 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "239/400 [================>.............] - ETA: 4:39 - loss: 0.0703 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "240/400 [=================>............] - ETA: 4:38 - loss: 0.0704 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "241/400 [=================>............] - ETA: 4:36 - loss: 0.0706 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "242/400 [=================>............] - ETA: 4:34 - loss: 0.0704 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "243/400 [=================>............] - ETA: 4:32 - loss: 0.0705 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "244/400 [=================>............] - ETA: 4:31 - loss: 0.0704 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "245/400 [=================>............] - ETA: 4:29 - loss: 0.0703 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "246/400 [=================>............] - ETA: 4:27 - loss: 0.0702 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "247/400 [=================>............] - ETA: 4:25 - loss: 0.0701 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "248/400 [=================>............] - ETA: 4:24 - loss: 0.0700 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "249/400 [=================>............] - ETA: 4:22 - loss: 0.0699 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "250/400 [=================>............] - ETA: 4:20 - loss: 0.0698 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "251/400 [=================>............] - ETA: 4:19 - loss: 0.0699 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "252/400 [=================>............] - ETA: 4:17 - loss: 0.0700 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "253/400 [=================>............] - ETA: 4:15 - loss: 0.0702 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "254/400 [==================>...........] - ETA: 4:13 - loss: 0.0701 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "255/400 [==================>...........] - ETA: 4:12 - loss: 0.0700 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "256/400 [==================>...........] - ETA: 4:10 - loss: 0.0701 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "257/400 [==================>...........] - ETA: 4:08 - loss: 0.0701 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "258/400 [==================>...........] - ETA: 4:06 - loss: 0.0700 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "259/400 [==================>...........] - ETA: 4:05 - loss: 0.0698 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "260/400 [==================>...........] - ETA: 4:03 - loss: 0.0698 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "261/400 [==================>...........] - ETA: 4:01 - loss: 0.0696 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "262/400 [==================>...........] - ETA: 3:59 - loss: 0.0696 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "263/400 [==================>...........] - ETA: 3:58 - loss: 0.0696 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "264/400 [==================>...........] - ETA: 3:56 - loss: 0.0695 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "265/400 [==================>...........] - ETA: 3:54 - loss: 0.0694 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "266/400 [==================>...........] - ETA: 3:53 - loss: 0.0694 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "267/400 [===================>..........] - ETA: 3:51 - loss: 0.0701 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "268/400 [===================>..........] - ETA: 3:49 - loss: 0.0699 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "269/400 [===================>..........] - ETA: 3:47 - loss: 0.0698 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "270/400 [===================>..........] - ETA: 3:46 - loss: 0.0696 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "271/400 [===================>..........] - ETA: 3:44 - loss: 0.0697 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "272/400 [===================>..........] - ETA: 3:42 - loss: 0.0696 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "273/400 [===================>..........] - ETA: 3:41 - loss: 0.0695 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "274/400 [===================>..........] - ETA: 3:39 - loss: 0.0698 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "275/400 [===================>..........] - ETA: 3:37 - loss: 0.0698 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "276/400 [===================>..........] - ETA: 3:35 - loss: 0.0697 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "277/400 [===================>..........] - ETA: 3:34 - loss: 0.0696 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "278/400 [===================>..........] - ETA: 3:32 - loss: 0.0696 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "279/400 [===================>..........] - ETA: 3:30 - loss: 0.0695 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "280/400 [====================>.........] - ETA: 3:28 - loss: 0.0694 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "281/400 [====================>.........] - ETA: 3:27 - loss: 0.0693 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "282/400 [====================>.........] - ETA: 3:25 - loss: 0.0693 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "283/400 [====================>.........] - ETA: 3:23 - loss: 0.0693 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "284/400 [====================>.........] - ETA: 3:21 - loss: 0.0692 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "285/400 [====================>.........] - ETA: 3:20 - loss: 0.0692 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "286/400 [====================>.........] - ETA: 3:18 - loss: 0.0692 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "287/400 [====================>.........] - ETA: 3:16 - loss: 0.0693 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "288/400 [====================>.........] - ETA: 3:15 - loss: 0.0692 - dice_coefficient: 0.9528\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "289/400 [====================>.........] - ETA: 3:13 - loss: 0.0693 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "290/400 [====================>.........] - ETA: 3:11 - loss: 0.0692 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "291/400 [====================>.........] - ETA: 3:09 - loss: 0.0696 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "292/400 [====================>.........] - ETA: 3:08 - loss: 0.0695 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "293/400 [====================>.........] - ETA: 3:06 - loss: 0.0696 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "294/400 [=====================>........] - ETA: 3:04 - loss: 0.0695 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "295/400 [=====================>........] - ETA: 3:02 - loss: 0.0694 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "296/400 [=====================>........] - ETA: 3:01 - loss: 0.0694 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "297/400 [=====================>........] - ETA: 2:59 - loss: 0.0693 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "298/400 [=====================>........] - ETA: 2:57 - loss: 0.0694 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "299/400 [=====================>........] - ETA: 2:55 - loss: 0.0695 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "300/400 [=====================>........] - ETA: 2:54 - loss: 0.0694 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "301/400 [=====================>........] - ETA: 2:52 - loss: 0.0693 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "302/400 [=====================>........] - ETA: 2:50 - loss: 0.0694 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "303/400 [=====================>........] - ETA: 2:48 - loss: 0.0693 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "304/400 [=====================>........] - ETA: 2:47 - loss: 0.0695 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "305/400 [=====================>........] - ETA: 2:45 - loss: 0.0694 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "306/400 [=====================>........] - ETA: 2:43 - loss: 0.0694 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "307/400 [======================>.......] - ETA: 2:42 - loss: 0.0693 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "308/400 [======================>.......] - ETA: 2:40 - loss: 0.0693 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "309/400 [======================>.......] - ETA: 2:38 - loss: 0.0692 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "310/400 [======================>.......] - ETA: 2:36 - loss: 0.0692 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "311/400 [======================>.......] - ETA: 2:35 - loss: 0.0691 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "312/400 [======================>.......] - ETA: 2:33 - loss: 0.0690 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "313/400 [======================>.......] - ETA: 2:31 - loss: 0.0691 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "314/400 [======================>.......] - ETA: 2:29 - loss: 0.0692 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "315/400 [======================>.......] - ETA: 2:28 - loss: 0.0691 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "316/400 [======================>.......] - ETA: 2:26 - loss: 0.0690 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "317/400 [======================>.......] - ETA: 2:24 - loss: 0.0690 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "318/400 [======================>.......] - ETA: 2:22 - loss: 0.0689 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "319/400 [======================>.......] - ETA: 2:21 - loss: 0.0688 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "320/400 [=======================>......] - ETA: 2:19 - loss: 0.0688 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "321/400 [=======================>......] - ETA: 2:17 - loss: 0.0689 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "322/400 [=======================>......] - ETA: 2:16 - loss: 0.0688 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "323/400 [=======================>......] - ETA: 2:14 - loss: 0.0688 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "324/400 [=======================>......] - ETA: 2:12 - loss: 0.0691 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "325/400 [=======================>......] - ETA: 2:10 - loss: 0.0695 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "326/400 [=======================>......] - ETA: 2:09 - loss: 0.0697 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "327/400 [=======================>......] - ETA: 2:07 - loss: 0.0696 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "328/400 [=======================>......] - ETA: 2:05 - loss: 0.0697 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "329/400 [=======================>......] - ETA: 2:03 - loss: 0.0696 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "330/400 [=======================>......] - ETA: 2:02 - loss: 0.0696 - dice_coefficient: 0.9519\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "331/400 [=======================>......] - ETA: 2:00 - loss: 0.0696 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "332/400 [=======================>......] - ETA: 1:58 - loss: 0.0696 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "333/400 [=======================>......] - ETA: 1:56 - loss: 0.0694 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "334/400 [========================>.....] - ETA: 1:55 - loss: 0.0693 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "335/400 [========================>.....] - ETA: 1:53 - loss: 0.0696 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "336/400 [========================>.....] - ETA: 1:51 - loss: 0.0695 - dice_coefficient: 0.9520\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "337/400 [========================>.....] - ETA: 1:49 - loss: 0.0694 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "338/400 [========================>.....] - ETA: 1:48 - loss: 0.0694 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "339/400 [========================>.....] - ETA: 1:46 - loss: 0.0693 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "340/400 [========================>.....] - ETA: 1:44 - loss: 0.0694 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "341/400 [========================>.....] - ETA: 1:42 - loss: 0.0693 - dice_coefficient: 0.9521\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "342/400 [========================>.....] - ETA: 1:41 - loss: 0.0692 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "343/400 [========================>.....] - ETA: 1:39 - loss: 0.0692 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "344/400 [========================>.....] - ETA: 1:37 - loss: 0.0692 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "345/400 [========================>.....] - ETA: 1:35 - loss: 0.0692 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "346/400 [========================>.....] - ETA: 1:34 - loss: 0.0692 - dice_coefficient: 0.9522\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "347/400 [=========================>....] - ETA: 1:32 - loss: 0.0691 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "348/400 [=========================>....] - ETA: 1:30 - loss: 0.0690 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "349/400 [=========================>....] - ETA: 1:28 - loss: 0.0689 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "350/400 [=========================>....] - ETA: 1:27 - loss: 0.0689 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "351/400 [=========================>....] - ETA: 1:25 - loss: 0.0690 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "352/400 [=========================>....] - ETA: 1:23 - loss: 0.0690 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "353/400 [=========================>....] - ETA: 1:21 - loss: 0.0690 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "354/400 [=========================>....] - ETA: 1:20 - loss: 0.0690 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "355/400 [=========================>....] - ETA: 1:18 - loss: 0.0690 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "356/400 [=========================>....] - ETA: 1:16 - loss: 0.0690 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "357/400 [=========================>....] - ETA: 1:15 - loss: 0.0691 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "358/400 [=========================>....] - ETA: 1:13 - loss: 0.0691 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "359/400 [=========================>....] - ETA: 1:11 - loss: 0.0690 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "360/400 [==========================>...] - ETA: 1:09 - loss: 0.0690 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "361/400 [==========================>...] - ETA: 1:08 - loss: 0.0689 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "362/400 [==========================>...] - ETA: 1:06 - loss: 0.0688 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "363/400 [==========================>...] - ETA: 1:04 - loss: 0.0688 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "364/400 [==========================>...] - ETA: 1:02 - loss: 0.0687 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "365/400 [==========================>...] - ETA: 1:01 - loss: 0.0686 - dice_coefficient: 0.9528\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "366/400 [==========================>...] - ETA: 59s - loss: 0.0686 - dice_coefficient: 0.9528 \n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "367/400 [==========================>...] - ETA: 57s - loss: 0.0687 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "368/400 [==========================>...] - ETA: 55s - loss: 0.0687 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "369/400 [==========================>...] - ETA: 54s - loss: 0.0687 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "370/400 [==========================>...] - ETA: 52s - loss: 0.0686 - dice_coefficient: 0.9528\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "371/400 [==========================>...] - ETA: 50s - loss: 0.0686 - dice_coefficient: 0.9528\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "372/400 [==========================>...] - ETA: 48s - loss: 0.0686 - dice_coefficient: 0.9528\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "373/400 [==========================>...] - ETA: 47s - loss: 0.0687 - dice_coefficient: 0.9528\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "374/400 [===========================>..] - ETA: 45s - loss: 0.0687 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "375/400 [===========================>..] - ETA: 43s - loss: 0.0687 - dice_coefficient: 0.9528\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "376/400 [===========================>..] - ETA: 41s - loss: 0.0687 - dice_coefficient: 0.9528\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "377/400 [===========================>..] - ETA: 40s - loss: 0.0688 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "378/400 [===========================>..] - ETA: 38s - loss: 0.0688 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "379/400 [===========================>..] - ETA: 36s - loss: 0.0689 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "380/400 [===========================>..] - ETA: 34s - loss: 0.0692 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "381/400 [===========================>..] - ETA: 33s - loss: 0.0692 - dice_coefficient: 0.9524\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "382/400 [===========================>..] - ETA: 31s - loss: 0.0691 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "383/400 [===========================>..] - ETA: 29s - loss: 0.0690 - dice_coefficient: 0.9525\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "384/400 [===========================>..] - ETA: 27s - loss: 0.0689 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "385/400 [===========================>..] - ETA: 26s - loss: 0.0689 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "386/400 [===========================>..] - ETA: 24s - loss: 0.0688 - dice_coefficient: 0.9526\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "387/400 [============================>.] - ETA: 22s - loss: 0.0687 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "388/400 [============================>.] - ETA: 20s - loss: 0.0687 - dice_coefficient: 0.9527\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "389/400 [============================>.] - ETA: 19s - loss: 0.0691 - dice_coefficient: 0.9523\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "390/400 [============================>.] - ETA: 17s - loss: 0.1105 - dice_coefficient: 0.9499\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "391/400 [============================>.] - ETA: 15s - loss: 0.1103 - dice_coefficient: 0.9500\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "392/400 [============================>.] - ETA: 13s - loss: 0.1101 - dice_coefficient: 0.9500\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "393/400 [============================>.] - ETA: 12s - loss: 0.1100 - dice_coefficient: 0.9500\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "394/400 [============================>.] - ETA: 10s - loss: 0.1100 - dice_coefficient: 0.9500\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "395/400 [============================>.] - ETA: 8s - loss: 0.1098 - dice_coefficient: 0.9500 \n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "396/400 [============================>.] - ETA: 6s - loss: 0.1097 - dice_coefficient: 0.9500\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "397/400 [============================>.] - ETA: 5s - loss: 0.1096 - dice_coefficient: 0.9500\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "398/400 [============================>.] - ETA: 3s - loss: 0.1094 - dice_coefficient: 0.9501\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "399/400 [============================>.] - ETA: 1s - loss: 0.1093 - dice_coefficient: 0.9502\n",
      "Epoch 00025: loss did not improve from 0.04222\n",
      "400/400 [==============================] - 702s 2s/step - loss: 0.1091 - dice_coefficient: 0.9502 - val_loss: 1.0725 - val_dice_coefficient: 0.5393\n",
      "Epoch 26/30\n",
      "\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "  1/400 [..............................] - ETA: 0s - loss: 0.0564 - dice_coefficient: 0.9710\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "  2/400 [..............................] - ETA: 5:46 - loss: 0.0551 - dice_coefficient: 0.9722\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "  3/400 [..............................] - ETA: 7:37 - loss: 0.0508 - dice_coefficient: 0.9706\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "  4/400 [..............................] - ETA: 8:35 - loss: 0.0552 - dice_coefficient: 0.9622\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "  5/400 [..............................] - ETA: 9:09 - loss: 0.0524 - dice_coefficient: 0.9629\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "  6/400 [..............................] - ETA: 9:32 - loss: 0.0527 - dice_coefficient: 0.9612\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "  7/400 [..............................] - ETA: 9:46 - loss: 0.0530 - dice_coefficient: 0.9628\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "  8/400 [..............................] - ETA: 9:57 - loss: 0.0515 - dice_coefficient: 0.9642\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "  9/400 [..............................] - ETA: 10:06 - loss: 0.0512 - dice_coefficient: 0.9640\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 10/400 [..............................] - ETA: 10:12 - loss: 0.0487 - dice_coefficient: 0.9662\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 11/400 [..............................] - ETA: 10:17 - loss: 0.0537 - dice_coefficient: 0.9635\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 12/400 [..............................] - ETA: 10:22 - loss: 0.0553 - dice_coefficient: 0.9630\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 13/400 [..............................] - ETA: 10:24 - loss: 0.0577 - dice_coefficient: 0.9626\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 14/400 [>.............................] - ETA: 10:27 - loss: 0.0631 - dice_coefficient: 0.9579\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 15/400 [>.............................] - ETA: 10:28 - loss: 0.0680 - dice_coefficient: 0.9531\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 16/400 [>.............................] - ETA: 10:29 - loss: 0.0678 - dice_coefficient: 0.9536\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 17/400 [>.............................] - ETA: 10:29 - loss: 0.0678 - dice_coefficient: 0.9533\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 18/400 [>.............................] - ETA: 10:30 - loss: 0.0665 - dice_coefficient: 0.9542\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 19/400 [>.............................] - ETA: 10:30 - loss: 0.0665 - dice_coefficient: 0.9544\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 20/400 [>.............................] - ETA: 10:29 - loss: 0.0718 - dice_coefficient: 0.9492\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 21/400 [>.............................] - ETA: 10:29 - loss: 0.0721 - dice_coefficient: 0.9488\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 22/400 [>.............................] - ETA: 10:28 - loss: 0.0724 - dice_coefficient: 0.9481\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 23/400 [>.............................] - ETA: 10:27 - loss: 0.0734 - dice_coefficient: 0.9468\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 24/400 [>.............................] - ETA: 10:27 - loss: 0.0710 - dice_coefficient: 0.9485\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 25/400 [>.............................] - ETA: 10:26 - loss: 0.0706 - dice_coefficient: 0.9494\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 26/400 [>.............................] - ETA: 10:25 - loss: 0.0699 - dice_coefficient: 0.9504\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 27/400 [=>............................] - ETA: 10:24 - loss: 0.0701 - dice_coefficient: 0.9512\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 28/400 [=>............................] - ETA: 10:23 - loss: 0.0740 - dice_coefficient: 0.9472\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 29/400 [=>............................] - ETA: 10:22 - loss: 0.0731 - dice_coefficient: 0.9485\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 30/400 [=>............................] - ETA: 10:21 - loss: 0.0718 - dice_coefficient: 0.9491\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 31/400 [=>............................] - ETA: 10:20 - loss: 0.0724 - dice_coefficient: 0.9484\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 32/400 [=>............................] - ETA: 10:19 - loss: 0.0725 - dice_coefficient: 0.9480\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 33/400 [=>............................] - ETA: 10:18 - loss: 0.0727 - dice_coefficient: 0.9476\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 34/400 [=>............................] - ETA: 10:16 - loss: 0.0720 - dice_coefficient: 0.9482\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 35/400 [=>............................] - ETA: 10:15 - loss: 0.0730 - dice_coefficient: 0.9486\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 36/400 [=>............................] - ETA: 10:14 - loss: 0.0727 - dice_coefficient: 0.9491\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 37/400 [=>............................] - ETA: 10:13 - loss: 0.0721 - dice_coefficient: 0.9498\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 38/400 [=>............................] - ETA: 10:11 - loss: 0.0715 - dice_coefficient: 0.9505\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 39/400 [=>............................] - ETA: 10:10 - loss: 0.0723 - dice_coefficient: 0.9495\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 40/400 [==>...........................] - ETA: 10:08 - loss: 0.0725 - dice_coefficient: 0.9495\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 41/400 [==>...........................] - ETA: 10:07 - loss: 0.0737 - dice_coefficient: 0.9494\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 42/400 [==>...........................] - ETA: 10:05 - loss: 0.0737 - dice_coefficient: 0.9493\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 43/400 [==>...........................] - ETA: 10:04 - loss: 0.0737 - dice_coefficient: 0.9496\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 44/400 [==>...........................] - ETA: 10:03 - loss: 0.0742 - dice_coefficient: 0.9490\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 45/400 [==>...........................] - ETA: 10:01 - loss: 0.0732 - dice_coefficient: 0.9497\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 46/400 [==>...........................] - ETA: 10:00 - loss: 0.0726 - dice_coefficient: 0.9500\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 47/400 [==>...........................] - ETA: 9:58 - loss: 0.0724 - dice_coefficient: 0.9499 \n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 48/400 [==>...........................] - ETA: 9:57 - loss: 0.0726 - dice_coefficient: 0.9502\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 49/400 [==>...........................] - ETA: 9:55 - loss: 0.0733 - dice_coefficient: 0.9495\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 50/400 [==>...........................] - ETA: 9:54 - loss: 0.0736 - dice_coefficient: 0.9497\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 51/400 [==>...........................] - ETA: 9:52 - loss: 0.0735 - dice_coefficient: 0.9500\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 52/400 [==>...........................] - ETA: 9:51 - loss: 0.0742 - dice_coefficient: 0.9492\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 53/400 [==>...........................] - ETA: 9:49 - loss: 0.0742 - dice_coefficient: 0.9496\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 54/400 [===>..........................] - ETA: 9:48 - loss: 0.0737 - dice_coefficient: 0.9498\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 55/400 [===>..........................] - ETA: 9:46 - loss: 0.0733 - dice_coefficient: 0.9500\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 56/400 [===>..........................] - ETA: 9:45 - loss: 0.0727 - dice_coefficient: 0.9503\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 57/400 [===>..........................] - ETA: 9:43 - loss: 0.0725 - dice_coefficient: 0.9506\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 58/400 [===>..........................] - ETA: 9:42 - loss: 0.0720 - dice_coefficient: 0.9510\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 59/400 [===>..........................] - ETA: 9:40 - loss: 0.0716 - dice_coefficient: 0.9512\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 60/400 [===>..........................] - ETA: 9:39 - loss: 0.0712 - dice_coefficient: 0.9513\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 61/400 [===>..........................] - ETA: 9:37 - loss: 0.0713 - dice_coefficient: 0.9511\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 62/400 [===>..........................] - ETA: 9:35 - loss: 0.0711 - dice_coefficient: 0.9510\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 63/400 [===>..........................] - ETA: 9:34 - loss: 0.0708 - dice_coefficient: 0.9514\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 64/400 [===>..........................] - ETA: 9:32 - loss: 0.0712 - dice_coefficient: 0.9511\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 65/400 [===>..........................] - ETA: 9:31 - loss: 0.0704 - dice_coefficient: 0.9517\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 66/400 [===>..........................] - ETA: 9:29 - loss: 0.0710 - dice_coefficient: 0.9516\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 67/400 [====>.........................] - ETA: 9:27 - loss: 0.0710 - dice_coefficient: 0.9513\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 68/400 [====>.........................] - ETA: 9:26 - loss: 0.0719 - dice_coefficient: 0.9510\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 69/400 [====>.........................] - ETA: 9:24 - loss: 0.0719 - dice_coefficient: 0.9511\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 70/400 [====>.........................] - ETA: 9:22 - loss: 0.0723 - dice_coefficient: 0.9506\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 71/400 [====>.........................] - ETA: 9:21 - loss: 0.0723 - dice_coefficient: 0.9504\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 72/400 [====>.........................] - ETA: 9:19 - loss: 0.0719 - dice_coefficient: 0.9507\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 73/400 [====>.........................] - ETA: 9:17 - loss: 0.0722 - dice_coefficient: 0.9505\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 74/400 [====>.........................] - ETA: 9:16 - loss: 0.0730 - dice_coefficient: 0.9498\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 75/400 [====>.........................] - ETA: 9:14 - loss: 0.0730 - dice_coefficient: 0.9500\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 76/400 [====>.........................] - ETA: 9:13 - loss: 0.0727 - dice_coefficient: 0.9503\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 77/400 [====>.........................] - ETA: 9:11 - loss: 0.0723 - dice_coefficient: 0.9506\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 78/400 [====>.........................] - ETA: 9:09 - loss: 0.0724 - dice_coefficient: 0.9506\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 79/400 [====>.........................] - ETA: 9:07 - loss: 0.0731 - dice_coefficient: 0.9501\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 80/400 [=====>........................] - ETA: 9:06 - loss: 0.0731 - dice_coefficient: 0.9502\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 81/400 [=====>........................] - ETA: 9:04 - loss: 0.0727 - dice_coefficient: 0.9504\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 82/400 [=====>........................] - ETA: 9:02 - loss: 0.0728 - dice_coefficient: 0.9501\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 83/400 [=====>........................] - ETA: 9:01 - loss: 0.0726 - dice_coefficient: 0.9503\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 84/400 [=====>........................] - ETA: 8:59 - loss: 0.0726 - dice_coefficient: 0.9501\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 85/400 [=====>........................] - ETA: 8:58 - loss: 0.0724 - dice_coefficient: 0.9501\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 86/400 [=====>........................] - ETA: 8:56 - loss: 0.2597 - dice_coefficient: 0.9391\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 87/400 [=====>........................] - ETA: 8:54 - loss: 0.2576 - dice_coefficient: 0.9392\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 88/400 [=====>........................] - ETA: 8:53 - loss: 0.2552 - dice_coefficient: 0.9395\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 89/400 [=====>........................] - ETA: 8:51 - loss: 0.2529 - dice_coefficient: 0.9396\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 90/400 [=====>........................] - ETA: 8:49 - loss: 0.2504 - dice_coefficient: 0.9401\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 91/400 [=====>........................] - ETA: 8:48 - loss: 0.2484 - dice_coefficient: 0.9403\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 92/400 [=====>........................] - ETA: 8:46 - loss: 0.2461 - dice_coefficient: 0.9407\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 93/400 [=====>........................] - ETA: 8:44 - loss: 0.2442 - dice_coefficient: 0.9407\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 94/400 [======>.......................] - ETA: 8:43 - loss: 0.2422 - dice_coefficient: 0.9411\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 95/400 [======>.......................] - ETA: 8:41 - loss: 0.2404 - dice_coefficient: 0.9412\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 96/400 [======>.......................] - ETA: 8:39 - loss: 0.2383 - dice_coefficient: 0.9415\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 97/400 [======>.......................] - ETA: 8:38 - loss: 0.2364 - dice_coefficient: 0.9417\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 98/400 [======>.......................] - ETA: 8:36 - loss: 0.2346 - dice_coefficient: 0.9418\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      " 99/400 [======>.......................] - ETA: 8:35 - loss: 0.2329 - dice_coefficient: 0.9419\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "100/400 [======>.......................] - ETA: 8:33 - loss: 0.2312 - dice_coefficient: 0.9422\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "101/400 [======>.......................] - ETA: 8:32 - loss: 0.2297 - dice_coefficient: 0.9422\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "102/400 [======>.......................] - ETA: 8:30 - loss: 0.2284 - dice_coefficient: 0.9422\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "103/400 [======>.......................] - ETA: 8:28 - loss: 0.2270 - dice_coefficient: 0.9420\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "104/400 [======>.......................] - ETA: 8:27 - loss: 0.2251 - dice_coefficient: 0.9424\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "105/400 [======>.......................] - ETA: 8:25 - loss: 0.2236 - dice_coefficient: 0.9427\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "106/400 [======>.......................] - ETA: 8:23 - loss: 0.2221 - dice_coefficient: 0.9427\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "107/400 [=======>......................] - ETA: 8:22 - loss: 0.2204 - dice_coefficient: 0.9430\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "108/400 [=======>......................] - ETA: 8:20 - loss: 0.2187 - dice_coefficient: 0.9433\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "109/400 [=======>......................] - ETA: 8:19 - loss: 0.2175 - dice_coefficient: 0.9434\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "110/400 [=======>......................] - ETA: 8:17 - loss: 0.2162 - dice_coefficient: 0.9435\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "111/400 [=======>......................] - ETA: 8:15 - loss: 0.2148 - dice_coefficient: 0.9436\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "112/400 [=======>......................] - ETA: 8:14 - loss: 0.2136 - dice_coefficient: 0.9435\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "113/400 [=======>......................] - ETA: 8:12 - loss: 0.2120 - dice_coefficient: 0.9439\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "114/400 [=======>......................] - ETA: 8:10 - loss: 0.2106 - dice_coefficient: 0.9441\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "115/400 [=======>......................] - ETA: 8:09 - loss: 0.2091 - dice_coefficient: 0.9443\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "116/400 [=======>......................] - ETA: 8:07 - loss: 0.2077 - dice_coefficient: 0.9446\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "117/400 [=======>......................] - ETA: 8:05 - loss: 0.2066 - dice_coefficient: 0.9448\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "118/400 [=======>......................] - ETA: 8:04 - loss: 0.2053 - dice_coefficient: 0.9450\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "119/400 [=======>......................] - ETA: 8:02 - loss: 0.2040 - dice_coefficient: 0.9452\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "120/400 [========>.....................] - ETA: 8:00 - loss: 0.2026 - dice_coefficient: 0.9454\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "121/400 [========>.....................] - ETA: 7:58 - loss: 0.2014 - dice_coefficient: 0.9456\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "122/400 [========>.....................] - ETA: 7:57 - loss: 0.2004 - dice_coefficient: 0.9456\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "123/400 [========>.....................] - ETA: 7:55 - loss: 0.1995 - dice_coefficient: 0.9456\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "124/400 [========>.....................] - ETA: 7:53 - loss: 0.1985 - dice_coefficient: 0.9456\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "125/400 [========>.....................] - ETA: 7:52 - loss: 0.1976 - dice_coefficient: 0.9455\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "126/400 [========>.....................] - ETA: 7:50 - loss: 0.1966 - dice_coefficient: 0.9456\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "127/400 [========>.....................] - ETA: 7:48 - loss: 0.1958 - dice_coefficient: 0.9454\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "128/400 [========>.....................] - ETA: 7:47 - loss: 0.1947 - dice_coefficient: 0.9455\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "129/400 [========>.....................] - ETA: 7:45 - loss: 0.1938 - dice_coefficient: 0.9455\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "130/400 [========>.....................] - ETA: 7:43 - loss: 0.1927 - dice_coefficient: 0.9457\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "131/400 [========>.....................] - ETA: 7:41 - loss: 0.1919 - dice_coefficient: 0.9455\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "132/400 [========>.....................] - ETA: 7:40 - loss: 0.1906 - dice_coefficient: 0.9458\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "133/400 [========>.....................] - ETA: 7:38 - loss: 0.1904 - dice_coefficient: 0.9451\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "134/400 [=========>....................] - ETA: 7:36 - loss: 0.1892 - dice_coefficient: 0.9454\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "135/400 [=========>....................] - ETA: 7:35 - loss: 0.1883 - dice_coefficient: 0.9455\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "136/400 [=========>....................] - ETA: 7:33 - loss: 0.1874 - dice_coefficient: 0.9456\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "137/400 [=========>....................] - ETA: 7:31 - loss: 0.1873 - dice_coefficient: 0.9450\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "138/400 [=========>....................] - ETA: 7:29 - loss: 0.1862 - dice_coefficient: 0.9453\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "139/400 [=========>....................] - ETA: 7:28 - loss: 0.1852 - dice_coefficient: 0.9454\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "140/400 [=========>....................] - ETA: 7:26 - loss: 0.1844 - dice_coefficient: 0.9455\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "141/400 [=========>....................] - ETA: 7:25 - loss: 0.1832 - dice_coefficient: 0.9458\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "142/400 [=========>....................] - ETA: 7:23 - loss: 0.1840 - dice_coefficient: 0.9445\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "143/400 [=========>....................] - ETA: 7:21 - loss: 0.1831 - dice_coefficient: 0.9446\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "144/400 [=========>....................] - ETA: 7:19 - loss: 0.1820 - dice_coefficient: 0.9449\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "145/400 [=========>....................] - ETA: 7:18 - loss: 0.1811 - dice_coefficient: 0.9450\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "146/400 [=========>....................] - ETA: 7:16 - loss: 0.1800 - dice_coefficient: 0.9453\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "147/400 [==========>...................] - ETA: 7:14 - loss: 0.1798 - dice_coefficient: 0.9448\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "148/400 [==========>...................] - ETA: 7:13 - loss: 0.1788 - dice_coefficient: 0.9450\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "149/400 [==========>...................] - ETA: 7:11 - loss: 0.1779 - dice_coefficient: 0.9451\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "150/400 [==========>...................] - ETA: 7:09 - loss: 0.1771 - dice_coefficient: 0.9454\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "151/400 [==========>...................] - ETA: 7:07 - loss: 0.1765 - dice_coefficient: 0.9454\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "152/400 [==========>...................] - ETA: 7:06 - loss: 0.1756 - dice_coefficient: 0.9455\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "153/400 [==========>...................] - ETA: 7:04 - loss: 0.1748 - dice_coefficient: 0.9456\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "154/400 [==========>...................] - ETA: 7:02 - loss: 0.1741 - dice_coefficient: 0.9455\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "155/400 [==========>...................] - ETA: 7:01 - loss: 0.1734 - dice_coefficient: 0.9456\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "156/400 [==========>...................] - ETA: 6:59 - loss: 0.1728 - dice_coefficient: 0.9456\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "157/400 [==========>...................] - ETA: 6:57 - loss: 0.1720 - dice_coefficient: 0.9458\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "158/400 [==========>...................] - ETA: 6:56 - loss: 0.1711 - dice_coefficient: 0.9460\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "159/400 [==========>...................] - ETA: 6:54 - loss: 0.1703 - dice_coefficient: 0.9461\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "160/400 [===========>..................] - ETA: 6:52 - loss: 0.1697 - dice_coefficient: 0.9462\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "161/400 [===========>..................] - ETA: 6:51 - loss: 0.1690 - dice_coefficient: 0.9464\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "162/400 [===========>..................] - ETA: 6:49 - loss: 0.1684 - dice_coefficient: 0.9463\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "163/400 [===========>..................] - ETA: 6:47 - loss: 0.1679 - dice_coefficient: 0.9463\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "164/400 [===========>..................] - ETA: 6:46 - loss: 0.1674 - dice_coefficient: 0.9463\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "165/400 [===========>..................] - ETA: 6:44 - loss: 0.1668 - dice_coefficient: 0.9463\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "166/400 [===========>..................] - ETA: 6:43 - loss: 0.1662 - dice_coefficient: 0.9464\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "167/400 [===========>..................] - ETA: 6:41 - loss: 0.1656 - dice_coefficient: 0.9465\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "168/400 [===========>..................] - ETA: 6:39 - loss: 0.1648 - dice_coefficient: 0.9467\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "169/400 [===========>..................] - ETA: 6:38 - loss: 0.1642 - dice_coefficient: 0.9468\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "170/400 [===========>..................] - ETA: 6:36 - loss: 0.1637 - dice_coefficient: 0.9468\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "171/400 [===========>..................] - ETA: 6:34 - loss: 0.1632 - dice_coefficient: 0.9469\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "172/400 [===========>..................] - ETA: 6:33 - loss: 0.1624 - dice_coefficient: 0.9471\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "173/400 [===========>..................] - ETA: 6:31 - loss: 0.1618 - dice_coefficient: 0.9472\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "174/400 [============>.................] - ETA: 6:29 - loss: 0.1616 - dice_coefficient: 0.9471\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "175/400 [============>.................] - ETA: 6:28 - loss: 0.1610 - dice_coefficient: 0.9471\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "176/400 [============>.................] - ETA: 6:26 - loss: 0.1608 - dice_coefficient: 0.9469\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "177/400 [============>.................] - ETA: 6:24 - loss: 0.1602 - dice_coefficient: 0.9470\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "178/400 [============>.................] - ETA: 6:23 - loss: 0.1596 - dice_coefficient: 0.9472\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "179/400 [============>.................] - ETA: 6:21 - loss: 0.1592 - dice_coefficient: 0.9474\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "180/400 [============>.................] - ETA: 6:19 - loss: 0.1587 - dice_coefficient: 0.9473\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "181/400 [============>.................] - ETA: 6:18 - loss: 0.1582 - dice_coefficient: 0.9474\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "182/400 [============>.................] - ETA: 6:16 - loss: 0.1576 - dice_coefficient: 0.9474\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "183/400 [============>.................] - ETA: 6:14 - loss: 0.1572 - dice_coefficient: 0.9475\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "184/400 [============>.................] - ETA: 6:13 - loss: 0.1567 - dice_coefficient: 0.9474\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "185/400 [============>.................] - ETA: 6:11 - loss: 0.1561 - dice_coefficient: 0.9475\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "186/400 [============>.................] - ETA: 6:09 - loss: 0.1557 - dice_coefficient: 0.9475\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "187/400 [=============>................] - ETA: 6:07 - loss: 0.1552 - dice_coefficient: 0.9476\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "188/400 [=============>................] - ETA: 6:06 - loss: 0.1545 - dice_coefficient: 0.9477\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "189/400 [=============>................] - ETA: 6:04 - loss: 0.1540 - dice_coefficient: 0.9477\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "190/400 [=============>................] - ETA: 6:02 - loss: 0.1535 - dice_coefficient: 0.9477\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "191/400 [=============>................] - ETA: 6:01 - loss: 0.1529 - dice_coefficient: 0.9479\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "192/400 [=============>................] - ETA: 5:59 - loss: 0.1524 - dice_coefficient: 0.9478\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "193/400 [=============>................] - ETA: 5:57 - loss: 0.1519 - dice_coefficient: 0.9479\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "194/400 [=============>................] - ETA: 5:56 - loss: 0.1513 - dice_coefficient: 0.9481\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "195/400 [=============>................] - ETA: 5:54 - loss: 0.1506 - dice_coefficient: 0.9483\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "196/400 [=============>................] - ETA: 5:52 - loss: 0.1504 - dice_coefficient: 0.9482\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "197/400 [=============>................] - ETA: 5:50 - loss: 0.1500 - dice_coefficient: 0.9482\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "198/400 [=============>................] - ETA: 5:49 - loss: 0.1496 - dice_coefficient: 0.9482\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "199/400 [=============>................] - ETA: 5:47 - loss: 0.1494 - dice_coefficient: 0.9481\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "200/400 [==============>...............] - ETA: 5:45 - loss: 0.1488 - dice_coefficient: 0.9482\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "201/400 [==============>...............] - ETA: 5:44 - loss: 0.1483 - dice_coefficient: 0.9483\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "202/400 [==============>...............] - ETA: 5:42 - loss: 0.1480 - dice_coefficient: 0.9483\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "203/400 [==============>...............] - ETA: 5:40 - loss: 0.1474 - dice_coefficient: 0.9484\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "204/400 [==============>...............] - ETA: 5:38 - loss: 0.1469 - dice_coefficient: 0.9485\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "205/400 [==============>...............] - ETA: 5:37 - loss: 0.1463 - dice_coefficient: 0.9487\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "206/400 [==============>...............] - ETA: 5:35 - loss: 0.1458 - dice_coefficient: 0.9488\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "207/400 [==============>...............] - ETA: 5:33 - loss: 0.1454 - dice_coefficient: 0.9489\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "208/400 [==============>...............] - ETA: 5:32 - loss: 0.1451 - dice_coefficient: 0.9489\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "209/400 [==============>...............] - ETA: 5:30 - loss: 0.1447 - dice_coefficient: 0.9490\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "210/400 [==============>...............] - ETA: 5:28 - loss: 0.1442 - dice_coefficient: 0.9491\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "211/400 [==============>...............] - ETA: 5:26 - loss: 0.1439 - dice_coefficient: 0.9491\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "212/400 [==============>...............] - ETA: 5:25 - loss: 0.1437 - dice_coefficient: 0.9491\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "213/400 [==============>...............] - ETA: 5:23 - loss: 0.1432 - dice_coefficient: 0.9492\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "214/400 [===============>..............] - ETA: 5:21 - loss: 0.1429 - dice_coefficient: 0.9492\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "215/400 [===============>..............] - ETA: 5:19 - loss: 0.1423 - dice_coefficient: 0.9493\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "216/400 [===============>..............] - ETA: 5:18 - loss: 0.1419 - dice_coefficient: 0.9495\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "217/400 [===============>..............] - ETA: 5:16 - loss: 0.1416 - dice_coefficient: 0.9495\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "218/400 [===============>..............] - ETA: 5:14 - loss: 0.1411 - dice_coefficient: 0.9495\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "219/400 [===============>..............] - ETA: 5:13 - loss: 0.1406 - dice_coefficient: 0.9496\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "220/400 [===============>..............] - ETA: 5:11 - loss: 0.1404 - dice_coefficient: 0.9495\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "221/400 [===============>..............] - ETA: 5:09 - loss: 0.1401 - dice_coefficient: 0.9495\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "222/400 [===============>..............] - ETA: 5:07 - loss: 0.1398 - dice_coefficient: 0.9496\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "223/400 [===============>..............] - ETA: 5:06 - loss: 0.1393 - dice_coefficient: 0.9498\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "224/400 [===============>..............] - ETA: 5:04 - loss: 0.1390 - dice_coefficient: 0.9497\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "225/400 [===============>..............] - ETA: 5:02 - loss: 0.1385 - dice_coefficient: 0.9499\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "226/400 [===============>..............] - ETA: 5:01 - loss: 0.1382 - dice_coefficient: 0.9499\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "227/400 [================>.............] - ETA: 4:59 - loss: 0.1378 - dice_coefficient: 0.9500\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "228/400 [================>.............] - ETA: 4:57 - loss: 0.1374 - dice_coefficient: 0.9501\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "229/400 [================>.............] - ETA: 4:55 - loss: 0.1370 - dice_coefficient: 0.9502\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "230/400 [================>.............] - ETA: 4:54 - loss: 0.1367 - dice_coefficient: 0.9502\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "231/400 [================>.............] - ETA: 4:52 - loss: 0.1364 - dice_coefficient: 0.9502\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "232/400 [================>.............] - ETA: 4:50 - loss: 0.1359 - dice_coefficient: 0.9503\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "233/400 [================>.............] - ETA: 4:49 - loss: 0.1356 - dice_coefficient: 0.9504\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "234/400 [================>.............] - ETA: 4:47 - loss: 0.1354 - dice_coefficient: 0.9504\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "235/400 [================>.............] - ETA: 4:45 - loss: 0.1349 - dice_coefficient: 0.9506\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "236/400 [================>.............] - ETA: 4:44 - loss: 0.1345 - dice_coefficient: 0.9507\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "237/400 [================>.............] - ETA: 4:42 - loss: 0.1340 - dice_coefficient: 0.9508\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "238/400 [================>.............] - ETA: 4:40 - loss: 0.1337 - dice_coefficient: 0.9509\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "239/400 [================>.............] - ETA: 4:39 - loss: 0.1334 - dice_coefficient: 0.9510\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "240/400 [=================>............] - ETA: 4:37 - loss: 0.1331 - dice_coefficient: 0.9510\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "241/400 [=================>............] - ETA: 4:35 - loss: 0.1330 - dice_coefficient: 0.9508\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "242/400 [=================>............] - ETA: 4:33 - loss: 0.1327 - dice_coefficient: 0.9508\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "243/400 [=================>............] - ETA: 4:32 - loss: 0.1322 - dice_coefficient: 0.9510\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "244/400 [=================>............] - ETA: 4:30 - loss: 0.1319 - dice_coefficient: 0.9511\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "245/400 [=================>............] - ETA: 4:28 - loss: 0.1315 - dice_coefficient: 0.9512\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "246/400 [=================>............] - ETA: 4:27 - loss: 0.1311 - dice_coefficient: 0.9513\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "247/400 [=================>............] - ETA: 4:25 - loss: 0.1307 - dice_coefficient: 0.9514\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "248/400 [=================>............] - ETA: 4:24 - loss: 0.1306 - dice_coefficient: 0.9513\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "249/400 [=================>............] - ETA: 4:22 - loss: 0.1301 - dice_coefficient: 0.9514\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "250/400 [=================>............] - ETA: 4:20 - loss: 0.1297 - dice_coefficient: 0.9515\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "251/400 [=================>............] - ETA: 4:18 - loss: 0.1294 - dice_coefficient: 0.9516\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "252/400 [=================>............] - ETA: 4:17 - loss: 0.1291 - dice_coefficient: 0.9517\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "253/400 [=================>............] - ETA: 4:15 - loss: 0.1286 - dice_coefficient: 0.9518\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "254/400 [==================>...........] - ETA: 4:13 - loss: 0.1283 - dice_coefficient: 0.9519\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "255/400 [==================>...........] - ETA: 4:12 - loss: 0.1280 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "256/400 [==================>...........] - ETA: 4:10 - loss: 0.1278 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "257/400 [==================>...........] - ETA: 4:08 - loss: 0.1275 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "258/400 [==================>...........] - ETA: 4:07 - loss: 0.1271 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "259/400 [==================>...........] - ETA: 4:05 - loss: 0.1268 - dice_coefficient: 0.9522\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "260/400 [==================>...........] - ETA: 4:03 - loss: 0.1266 - dice_coefficient: 0.9522\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "261/400 [==================>...........] - ETA: 4:01 - loss: 0.1264 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "262/400 [==================>...........] - ETA: 4:00 - loss: 0.1264 - dice_coefficient: 0.9519\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "263/400 [==================>...........] - ETA: 3:58 - loss: 0.1262 - dice_coefficient: 0.9519\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "264/400 [==================>...........] - ETA: 3:56 - loss: 0.1259 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "265/400 [==================>...........] - ETA: 3:55 - loss: 0.1257 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "266/400 [==================>...........] - ETA: 3:53 - loss: 0.1254 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "267/400 [===================>..........] - ETA: 3:51 - loss: 0.1254 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "268/400 [===================>..........] - ETA: 3:49 - loss: 0.1252 - dice_coefficient: 0.9519\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "269/400 [===================>..........] - ETA: 3:48 - loss: 0.1249 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "270/400 [===================>..........] - ETA: 3:46 - loss: 0.1248 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "271/400 [===================>..........] - ETA: 3:44 - loss: 0.1246 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "272/400 [===================>..........] - ETA: 3:43 - loss: 0.1245 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "273/400 [===================>..........] - ETA: 3:41 - loss: 0.1241 - dice_coefficient: 0.9522\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "274/400 [===================>..........] - ETA: 3:39 - loss: 0.1242 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "275/400 [===================>..........] - ETA: 3:37 - loss: 0.1241 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "276/400 [===================>..........] - ETA: 3:36 - loss: 0.1239 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "277/400 [===================>..........] - ETA: 3:34 - loss: 0.1238 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "278/400 [===================>..........] - ETA: 3:32 - loss: 0.1235 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "279/400 [===================>..........] - ETA: 3:31 - loss: 0.1233 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "280/400 [====================>.........] - ETA: 3:29 - loss: 0.1231 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "281/400 [====================>.........] - ETA: 3:27 - loss: 0.1229 - dice_coefficient: 0.9519\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "282/400 [====================>.........] - ETA: 3:25 - loss: 0.1226 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "283/400 [====================>.........] - ETA: 3:24 - loss: 0.1225 - dice_coefficient: 0.9520\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "284/400 [====================>.........] - ETA: 3:22 - loss: 0.1222 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "285/400 [====================>.........] - ETA: 3:20 - loss: 0.1220 - dice_coefficient: 0.9521\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "286/400 [====================>.........] - ETA: 3:18 - loss: 0.1218 - dice_coefficient: 0.9522\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "287/400 [====================>.........] - ETA: 3:17 - loss: 0.1217 - dice_coefficient: 0.9522\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "288/400 [====================>.........] - ETA: 3:15 - loss: 0.1214 - dice_coefficient: 0.9523\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "289/400 [====================>.........] - ETA: 3:13 - loss: 0.1210 - dice_coefficient: 0.9524\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "290/400 [====================>.........] - ETA: 3:11 - loss: 0.1209 - dice_coefficient: 0.9523\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "291/400 [====================>.........] - ETA: 3:10 - loss: 0.1207 - dice_coefficient: 0.9523\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "292/400 [====================>.........] - ETA: 3:08 - loss: 0.1204 - dice_coefficient: 0.9523\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "293/400 [====================>.........] - ETA: 3:06 - loss: 0.1201 - dice_coefficient: 0.9524\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "294/400 [=====================>........] - ETA: 3:04 - loss: 0.1199 - dice_coefficient: 0.9524\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "295/400 [=====================>........] - ETA: 3:03 - loss: 0.1196 - dice_coefficient: 0.9525\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "296/400 [=====================>........] - ETA: 3:01 - loss: 0.1194 - dice_coefficient: 0.9525\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "297/400 [=====================>........] - ETA: 2:59 - loss: 0.1192 - dice_coefficient: 0.9526\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "298/400 [=====================>........] - ETA: 2:58 - loss: 0.1188 - dice_coefficient: 0.9527\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "299/400 [=====================>........] - ETA: 2:56 - loss: 0.1188 - dice_coefficient: 0.9525\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "300/400 [=====================>........] - ETA: 2:54 - loss: 0.1186 - dice_coefficient: 0.9526\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "301/400 [=====================>........] - ETA: 2:52 - loss: 0.1184 - dice_coefficient: 0.9526\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "302/400 [=====================>........] - ETA: 2:51 - loss: 0.1182 - dice_coefficient: 0.9527\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "303/400 [=====================>........] - ETA: 2:49 - loss: 0.1180 - dice_coefficient: 0.9527\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "304/400 [=====================>........] - ETA: 2:47 - loss: 0.1178 - dice_coefficient: 0.9528\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "305/400 [=====================>........] - ETA: 2:45 - loss: 0.1175 - dice_coefficient: 0.9528\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "306/400 [=====================>........] - ETA: 2:44 - loss: 0.1173 - dice_coefficient: 0.9529\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "307/400 [======================>.......] - ETA: 2:42 - loss: 0.1171 - dice_coefficient: 0.9529\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "308/400 [======================>.......] - ETA: 2:40 - loss: 0.1168 - dice_coefficient: 0.9531\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "309/400 [======================>.......] - ETA: 2:38 - loss: 0.1168 - dice_coefficient: 0.9529\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "310/400 [======================>.......] - ETA: 2:37 - loss: 0.1165 - dice_coefficient: 0.9530\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "311/400 [======================>.......] - ETA: 2:35 - loss: 0.1164 - dice_coefficient: 0.9529\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "312/400 [======================>.......] - ETA: 2:33 - loss: 0.1161 - dice_coefficient: 0.9530\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "313/400 [======================>.......] - ETA: 2:31 - loss: 0.1163 - dice_coefficient: 0.9527\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "314/400 [======================>.......] - ETA: 2:30 - loss: 0.1160 - dice_coefficient: 0.9528\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "315/400 [======================>.......] - ETA: 2:28 - loss: 0.1158 - dice_coefficient: 0.9529\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "316/400 [======================>.......] - ETA: 2:26 - loss: 0.1155 - dice_coefficient: 0.9529\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "317/400 [======================>.......] - ETA: 2:24 - loss: 0.1153 - dice_coefficient: 0.9529\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "318/400 [======================>.......] - ETA: 2:23 - loss: 0.1152 - dice_coefficient: 0.9529\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "319/400 [======================>.......] - ETA: 2:21 - loss: 0.1150 - dice_coefficient: 0.9530\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "320/400 [=======================>......] - ETA: 2:19 - loss: 0.1149 - dice_coefficient: 0.9529\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "321/400 [=======================>......] - ETA: 2:18 - loss: 0.1147 - dice_coefficient: 0.9529\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "322/400 [=======================>......] - ETA: 2:16 - loss: 0.1144 - dice_coefficient: 0.9530\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "323/400 [=======================>......] - ETA: 2:14 - loss: 0.1142 - dice_coefficient: 0.9530\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "324/400 [=======================>......] - ETA: 2:12 - loss: 0.1141 - dice_coefficient: 0.9531\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "325/400 [=======================>......] - ETA: 2:11 - loss: 0.1141 - dice_coefficient: 0.9530\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "326/400 [=======================>......] - ETA: 2:09 - loss: 0.1139 - dice_coefficient: 0.9530\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "327/400 [=======================>......] - ETA: 2:07 - loss: 0.1137 - dice_coefficient: 0.9530\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "328/400 [=======================>......] - ETA: 2:05 - loss: 0.1135 - dice_coefficient: 0.9531\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "329/400 [=======================>......] - ETA: 2:04 - loss: 0.1135 - dice_coefficient: 0.9530\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "330/400 [=======================>......] - ETA: 2:02 - loss: 0.1132 - dice_coefficient: 0.9531\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "331/400 [=======================>......] - ETA: 2:00 - loss: 0.1130 - dice_coefficient: 0.9531\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "332/400 [=======================>......] - ETA: 1:58 - loss: 0.1128 - dice_coefficient: 0.9532\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "333/400 [=======================>......] - ETA: 1:57 - loss: 0.1126 - dice_coefficient: 0.9532\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "334/400 [========================>.....] - ETA: 1:55 - loss: 0.1124 - dice_coefficient: 0.9533\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "335/400 [========================>.....] - ETA: 1:53 - loss: 0.1123 - dice_coefficient: 0.9533\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "336/400 [========================>.....] - ETA: 1:51 - loss: 0.1121 - dice_coefficient: 0.9533\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "337/400 [========================>.....] - ETA: 1:50 - loss: 0.1120 - dice_coefficient: 0.9533\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "338/400 [========================>.....] - ETA: 1:48 - loss: 0.1119 - dice_coefficient: 0.9533\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "339/400 [========================>.....] - ETA: 1:46 - loss: 0.1117 - dice_coefficient: 0.9533\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "340/400 [========================>.....] - ETA: 1:44 - loss: 0.1116 - dice_coefficient: 0.9533\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "341/400 [========================>.....] - ETA: 1:43 - loss: 0.1114 - dice_coefficient: 0.9534\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "342/400 [========================>.....] - ETA: 1:41 - loss: 0.1112 - dice_coefficient: 0.9534\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "343/400 [========================>.....] - ETA: 1:39 - loss: 0.1111 - dice_coefficient: 0.9534\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "344/400 [========================>.....] - ETA: 1:37 - loss: 0.1110 - dice_coefficient: 0.9534\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "345/400 [========================>.....] - ETA: 1:36 - loss: 0.1108 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "346/400 [========================>.....] - ETA: 1:34 - loss: 0.1105 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "347/400 [=========================>....] - ETA: 1:32 - loss: 0.1103 - dice_coefficient: 0.9536\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "348/400 [=========================>....] - ETA: 1:30 - loss: 0.1103 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "349/400 [=========================>....] - ETA: 1:29 - loss: 0.1102 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "350/400 [=========================>....] - ETA: 1:27 - loss: 0.1100 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "351/400 [=========================>....] - ETA: 1:25 - loss: 0.1098 - dice_coefficient: 0.9536\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "352/400 [=========================>....] - ETA: 1:23 - loss: 0.1099 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "353/400 [=========================>....] - ETA: 1:22 - loss: 0.1097 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "354/400 [=========================>....] - ETA: 1:20 - loss: 0.1095 - dice_coefficient: 0.9536\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "355/400 [=========================>....] - ETA: 1:18 - loss: 0.1097 - dice_coefficient: 0.9533\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "356/400 [=========================>....] - ETA: 1:16 - loss: 0.1095 - dice_coefficient: 0.9534\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "357/400 [=========================>....] - ETA: 1:15 - loss: 0.1093 - dice_coefficient: 0.9534\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "358/400 [=========================>....] - ETA: 1:13 - loss: 0.1092 - dice_coefficient: 0.9534\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "359/400 [=========================>....] - ETA: 1:11 - loss: 0.1090 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "360/400 [==========================>...] - ETA: 1:09 - loss: 0.1088 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "361/400 [==========================>...] - ETA: 1:08 - loss: 0.1087 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "362/400 [==========================>...] - ETA: 1:06 - loss: 0.1085 - dice_coefficient: 0.9536\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "363/400 [==========================>...] - ETA: 1:04 - loss: 0.1085 - dice_coefficient: 0.9535\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "364/400 [==========================>...] - ETA: 1:02 - loss: 0.1084 - dice_coefficient: 0.9536\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "365/400 [==========================>...] - ETA: 1:01 - loss: 0.1082 - dice_coefficient: 0.9536\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "366/400 [==========================>...] - ETA: 59s - loss: 0.1080 - dice_coefficient: 0.9536 \n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "367/400 [==========================>...] - ETA: 57s - loss: 0.1079 - dice_coefficient: 0.9537\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "368/400 [==========================>...] - ETA: 55s - loss: 0.1077 - dice_coefficient: 0.9537\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "369/400 [==========================>...] - ETA: 54s - loss: 0.1075 - dice_coefficient: 0.9538\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "370/400 [==========================>...] - ETA: 52s - loss: 0.1073 - dice_coefficient: 0.9539\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "371/400 [==========================>...] - ETA: 50s - loss: 0.1073 - dice_coefficient: 0.9538\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "372/400 [==========================>...] - ETA: 48s - loss: 0.1071 - dice_coefficient: 0.9538\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "373/400 [==========================>...] - ETA: 47s - loss: 0.1070 - dice_coefficient: 0.9538\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "374/400 [===========================>..] - ETA: 45s - loss: 0.1068 - dice_coefficient: 0.9539\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "375/400 [===========================>..] - ETA: 43s - loss: 0.1067 - dice_coefficient: 0.9539\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "376/400 [===========================>..] - ETA: 41s - loss: 0.1065 - dice_coefficient: 0.9539\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "377/400 [===========================>..] - ETA: 40s - loss: 0.1064 - dice_coefficient: 0.9540\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "378/400 [===========================>..] - ETA: 38s - loss: 0.1062 - dice_coefficient: 0.9540\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "379/400 [===========================>..] - ETA: 36s - loss: 0.1060 - dice_coefficient: 0.9540\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "380/400 [===========================>..] - ETA: 34s - loss: 0.1059 - dice_coefficient: 0.9540\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "381/400 [===========================>..] - ETA: 33s - loss: 0.1058 - dice_coefficient: 0.9541\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "382/400 [===========================>..] - ETA: 31s - loss: 0.1056 - dice_coefficient: 0.9541\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "383/400 [===========================>..] - ETA: 29s - loss: 0.1055 - dice_coefficient: 0.9542\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "384/400 [===========================>..] - ETA: 27s - loss: 0.1054 - dice_coefficient: 0.9542\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "385/400 [===========================>..] - ETA: 26s - loss: 0.1053 - dice_coefficient: 0.9542\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "386/400 [===========================>..] - ETA: 24s - loss: 0.1051 - dice_coefficient: 0.9543\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "387/400 [============================>.] - ETA: 22s - loss: 0.1050 - dice_coefficient: 0.9543\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "388/400 [============================>.] - ETA: 20s - loss: 0.1048 - dice_coefficient: 0.9543\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "389/400 [============================>.] - ETA: 19s - loss: 0.1047 - dice_coefficient: 0.9544\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "390/400 [============================>.] - ETA: 17s - loss: 0.1049 - dice_coefficient: 0.9541\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "391/400 [============================>.] - ETA: 15s - loss: 0.1049 - dice_coefficient: 0.9540\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "392/400 [============================>.] - ETA: 13s - loss: 0.1048 - dice_coefficient: 0.9540\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "393/400 [============================>.] - ETA: 12s - loss: 0.1048 - dice_coefficient: 0.9539\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "394/400 [============================>.] - ETA: 10s - loss: 0.1047 - dice_coefficient: 0.9539\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "395/400 [============================>.] - ETA: 8s - loss: 0.1046 - dice_coefficient: 0.9539 \n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "396/400 [============================>.] - ETA: 6s - loss: 0.1044 - dice_coefficient: 0.9540\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "397/400 [============================>.] - ETA: 5s - loss: 0.1042 - dice_coefficient: 0.9541\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "398/400 [============================>.] - ETA: 3s - loss: 0.1041 - dice_coefficient: 0.9541\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "399/400 [============================>.] - ETA: 1s - loss: 0.1039 - dice_coefficient: 0.9541\n",
      "Epoch 00026: loss did not improve from 0.04222\n",
      "400/400 [==============================] - 703s 2s/step - loss: 0.1038 - dice_coefficient: 0.9542 - val_loss: 1.0594 - val_dice_coefficient: 0.5455\n",
      "Epoch 27/30\n",
      "\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "  1/400 [..............................] - ETA: 0s - loss: 0.0757 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "  2/400 [..............................] - ETA: 5:49 - loss: 0.0784 - dice_coefficient: 0.9469\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "  3/400 [..............................] - ETA: 7:40 - loss: 0.0717 - dice_coefficient: 0.9483\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "  4/400 [..............................] - ETA: 8:38 - loss: 0.0681 - dice_coefficient: 0.9502\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "  5/400 [..............................] - ETA: 9:13 - loss: 0.0654 - dice_coefficient: 0.9550\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "  6/400 [..............................] - ETA: 9:33 - loss: 0.0618 - dice_coefficient: 0.9567\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "  7/400 [..............................] - ETA: 9:47 - loss: 0.0669 - dice_coefficient: 0.9526\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "  8/400 [..............................] - ETA: 9:58 - loss: 0.0612 - dice_coefficient: 0.9568\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "  9/400 [..............................] - ETA: 10:06 - loss: 0.0664 - dice_coefficient: 0.9536\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 10/400 [..............................] - ETA: 10:11 - loss: 0.0630 - dice_coefficient: 0.9552\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 11/400 [..............................] - ETA: 10:15 - loss: 0.0700 - dice_coefficient: 0.9518\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 12/400 [..............................] - ETA: 10:19 - loss: 0.0724 - dice_coefficient: 0.9498\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 13/400 [..............................] - ETA: 10:22 - loss: 0.0706 - dice_coefficient: 0.9514\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 14/400 [>.............................] - ETA: 10:24 - loss: 0.0678 - dice_coefficient: 0.9529\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 15/400 [>.............................] - ETA: 10:26 - loss: 0.0684 - dice_coefficient: 0.9526\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 16/400 [>.............................] - ETA: 10:28 - loss: 0.0682 - dice_coefficient: 0.9527\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 17/400 [>.............................] - ETA: 10:29 - loss: 0.0685 - dice_coefficient: 0.9535\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 18/400 [>.............................] - ETA: 10:30 - loss: 0.0658 - dice_coefficient: 0.9553\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 19/400 [>.............................] - ETA: 10:30 - loss: 0.0681 - dice_coefficient: 0.9553\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 20/400 [>.............................] - ETA: 10:31 - loss: 0.0725 - dice_coefficient: 0.9510\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 21/400 [>.............................] - ETA: 10:31 - loss: 0.0696 - dice_coefficient: 0.9528\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 22/400 [>.............................] - ETA: 10:31 - loss: 0.0695 - dice_coefficient: 0.9529\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 23/400 [>.............................] - ETA: 10:30 - loss: 0.0760 - dice_coefficient: 0.9500\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 24/400 [>.............................] - ETA: 10:29 - loss: 0.0756 - dice_coefficient: 0.9505\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 25/400 [>.............................] - ETA: 10:29 - loss: 0.0739 - dice_coefficient: 0.9518\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 26/400 [>.............................] - ETA: 10:28 - loss: 0.0745 - dice_coefficient: 0.9514\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 27/400 [=>............................] - ETA: 10:27 - loss: 0.0733 - dice_coefficient: 0.9521\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 28/400 [=>............................] - ETA: 10:26 - loss: 0.0720 - dice_coefficient: 0.9530\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 29/400 [=>............................] - ETA: 10:25 - loss: 0.0709 - dice_coefficient: 0.9538\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 30/400 [=>............................] - ETA: 10:24 - loss: 0.0728 - dice_coefficient: 0.9518\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 31/400 [=>............................] - ETA: 10:23 - loss: 0.0716 - dice_coefficient: 0.9528\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 32/400 [=>............................] - ETA: 10:23 - loss: 0.0719 - dice_coefficient: 0.9522\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 33/400 [=>............................] - ETA: 10:22 - loss: 0.0713 - dice_coefficient: 0.9527\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 34/400 [=>............................] - ETA: 10:21 - loss: 0.0705 - dice_coefficient: 0.9530\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 35/400 [=>............................] - ETA: 10:20 - loss: 0.0699 - dice_coefficient: 0.9533\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 36/400 [=>............................] - ETA: 10:18 - loss: 0.0695 - dice_coefficient: 0.9536\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 37/400 [=>............................] - ETA: 10:17 - loss: 0.0689 - dice_coefficient: 0.9541\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 38/400 [=>............................] - ETA: 10:16 - loss: 0.0686 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 39/400 [=>............................] - ETA: 10:15 - loss: 0.0683 - dice_coefficient: 0.9545\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 40/400 [==>...........................] - ETA: 10:14 - loss: 0.0677 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 41/400 [==>...........................] - ETA: 10:13 - loss: 0.0677 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 42/400 [==>...........................] - ETA: 10:12 - loss: 0.0671 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 43/400 [==>...........................] - ETA: 10:11 - loss: 0.0664 - dice_coefficient: 0.9554\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 44/400 [==>...........................] - ETA: 10:10 - loss: 0.0671 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 45/400 [==>...........................] - ETA: 10:08 - loss: 0.0670 - dice_coefficient: 0.9552\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 46/400 [==>...........................] - ETA: 10:07 - loss: 0.0676 - dice_coefficient: 0.9551\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 47/400 [==>...........................] - ETA: 10:06 - loss: 0.0672 - dice_coefficient: 0.9556\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 48/400 [==>...........................] - ETA: 10:05 - loss: 0.0670 - dice_coefficient: 0.9555\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 49/400 [==>...........................] - ETA: 10:03 - loss: 0.0677 - dice_coefficient: 0.9556\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 50/400 [==>...........................] - ETA: 10:02 - loss: 0.0678 - dice_coefficient: 0.9555\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 51/400 [==>...........................] - ETA: 10:01 - loss: 0.0679 - dice_coefficient: 0.9553\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 52/400 [==>...........................] - ETA: 10:00 - loss: 0.0678 - dice_coefficient: 0.9556\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 53/400 [==>...........................] - ETA: 9:58 - loss: 0.0679 - dice_coefficient: 0.9554 \n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 54/400 [===>..........................] - ETA: 9:57 - loss: 0.0677 - dice_coefficient: 0.9556\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 55/400 [===>..........................] - ETA: 9:55 - loss: 0.0697 - dice_coefficient: 0.9536\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 56/400 [===>..........................] - ETA: 9:54 - loss: 0.0694 - dice_coefficient: 0.9539\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 57/400 [===>..........................] - ETA: 9:52 - loss: 0.0688 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 58/400 [===>..........................] - ETA: 9:51 - loss: 0.0683 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 59/400 [===>..........................] - ETA: 9:49 - loss: 0.0683 - dice_coefficient: 0.9545\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 60/400 [===>..........................] - ETA: 9:48 - loss: 0.0690 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 61/400 [===>..........................] - ETA: 9:46 - loss: 0.0684 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 62/400 [===>..........................] - ETA: 9:45 - loss: 0.0683 - dice_coefficient: 0.9551\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 63/400 [===>..........................] - ETA: 9:44 - loss: 0.0686 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 64/400 [===>..........................] - ETA: 9:42 - loss: 0.0683 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 65/400 [===>..........................] - ETA: 9:40 - loss: 0.0683 - dice_coefficient: 0.9551\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 66/400 [===>..........................] - ETA: 9:39 - loss: 0.0680 - dice_coefficient: 0.9554\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 67/400 [====>.........................] - ETA: 9:38 - loss: 0.0683 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 68/400 [====>.........................] - ETA: 9:36 - loss: 0.0678 - dice_coefficient: 0.9552\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 69/400 [====>.........................] - ETA: 9:35 - loss: 0.0676 - dice_coefficient: 0.9554\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 70/400 [====>.........................] - ETA: 9:33 - loss: 0.0672 - dice_coefficient: 0.9556\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 71/400 [====>.........................] - ETA: 9:32 - loss: 0.0672 - dice_coefficient: 0.9559\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 72/400 [====>.........................] - ETA: 9:30 - loss: 0.0667 - dice_coefficient: 0.9562\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 73/400 [====>.........................] - ETA: 9:28 - loss: 0.0669 - dice_coefficient: 0.9559\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 74/400 [====>.........................] - ETA: 9:27 - loss: 0.0670 - dice_coefficient: 0.9557\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 75/400 [====>.........................] - ETA: 9:25 - loss: 0.0668 - dice_coefficient: 0.9558\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 76/400 [====>.........................] - ETA: 9:23 - loss: 0.0668 - dice_coefficient: 0.9558\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 77/400 [====>.........................] - ETA: 9:22 - loss: 0.0667 - dice_coefficient: 0.9560\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 78/400 [====>.........................] - ETA: 9:20 - loss: 0.0667 - dice_coefficient: 0.9559\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 79/400 [====>.........................] - ETA: 9:18 - loss: 0.0664 - dice_coefficient: 0.9561\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 80/400 [=====>........................] - ETA: 9:17 - loss: 0.0662 - dice_coefficient: 0.9561\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 81/400 [=====>........................] - ETA: 9:15 - loss: 0.0662 - dice_coefficient: 0.9560\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 82/400 [=====>........................] - ETA: 9:14 - loss: 0.0664 - dice_coefficient: 0.9559\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 83/400 [=====>........................] - ETA: 9:12 - loss: 0.0664 - dice_coefficient: 0.9560\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 84/400 [=====>........................] - ETA: 9:11 - loss: 0.0664 - dice_coefficient: 0.9559\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 85/400 [=====>........................] - ETA: 9:09 - loss: 0.0663 - dice_coefficient: 0.9560\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 86/400 [=====>........................] - ETA: 9:07 - loss: 0.0663 - dice_coefficient: 0.9559\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 87/400 [=====>........................] - ETA: 9:06 - loss: 0.0664 - dice_coefficient: 0.9559\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 88/400 [=====>........................] - ETA: 9:04 - loss: 0.0666 - dice_coefficient: 0.9557\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 89/400 [=====>........................] - ETA: 9:02 - loss: 0.0666 - dice_coefficient: 0.9558\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 90/400 [=====>........................] - ETA: 9:01 - loss: 0.0665 - dice_coefficient: 0.9560\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 91/400 [=====>........................] - ETA: 8:59 - loss: 0.0665 - dice_coefficient: 0.9560\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 92/400 [=====>........................] - ETA: 8:57 - loss: 0.0662 - dice_coefficient: 0.9562\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 93/400 [=====>........................] - ETA: 8:56 - loss: 0.0659 - dice_coefficient: 0.9564\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 94/400 [======>.......................] - ETA: 8:54 - loss: 0.0660 - dice_coefficient: 0.9563\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 95/400 [======>.......................] - ETA: 8:52 - loss: 0.0664 - dice_coefficient: 0.9561\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 96/400 [======>.......................] - ETA: 8:50 - loss: 0.0663 - dice_coefficient: 0.9561\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 97/400 [======>.......................] - ETA: 8:49 - loss: 0.0662 - dice_coefficient: 0.9560\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 98/400 [======>.......................] - ETA: 8:47 - loss: 0.0662 - dice_coefficient: 0.9559\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      " 99/400 [======>.......................] - ETA: 8:45 - loss: 0.0660 - dice_coefficient: 0.9561\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "100/400 [======>.......................] - ETA: 8:44 - loss: 0.0659 - dice_coefficient: 0.9562\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "101/400 [======>.......................] - ETA: 8:42 - loss: 0.0659 - dice_coefficient: 0.9563\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "102/400 [======>.......................] - ETA: 8:40 - loss: 0.0659 - dice_coefficient: 0.9563\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "103/400 [======>.......................] - ETA: 8:38 - loss: 0.0657 - dice_coefficient: 0.9564\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "104/400 [======>.......................] - ETA: 8:37 - loss: 0.0657 - dice_coefficient: 0.9565\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "105/400 [======>.......................] - ETA: 8:35 - loss: 0.0658 - dice_coefficient: 0.9564\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "106/400 [======>.......................] - ETA: 8:33 - loss: 0.0655 - dice_coefficient: 0.9567\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "107/400 [=======>......................] - ETA: 8:32 - loss: 0.0653 - dice_coefficient: 0.9568\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "108/400 [=======>......................] - ETA: 8:30 - loss: 0.0650 - dice_coefficient: 0.9570\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "109/400 [=======>......................] - ETA: 8:28 - loss: 0.0651 - dice_coefficient: 0.9570\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "110/400 [=======>......................] - ETA: 8:27 - loss: 0.0653 - dice_coefficient: 0.9569\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "111/400 [=======>......................] - ETA: 8:25 - loss: 0.0655 - dice_coefficient: 0.9567\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "112/400 [=======>......................] - ETA: 8:23 - loss: 0.0662 - dice_coefficient: 0.9565\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "113/400 [=======>......................] - ETA: 8:21 - loss: 0.0660 - dice_coefficient: 0.9567\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "114/400 [=======>......................] - ETA: 8:20 - loss: 0.0663 - dice_coefficient: 0.9562\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "115/400 [=======>......................] - ETA: 8:18 - loss: 0.0661 - dice_coefficient: 0.9564\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "116/400 [=======>......................] - ETA: 8:16 - loss: 0.0662 - dice_coefficient: 0.9563\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "117/400 [=======>......................] - ETA: 8:15 - loss: 0.0662 - dice_coefficient: 0.9563\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "118/400 [=======>......................] - ETA: 8:13 - loss: 0.0660 - dice_coefficient: 0.9564\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "119/400 [=======>......................] - ETA: 8:11 - loss: 0.0658 - dice_coefficient: 0.9566\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "120/400 [========>.....................] - ETA: 8:10 - loss: 0.0658 - dice_coefficient: 0.9565\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "121/400 [========>.....................] - ETA: 8:08 - loss: 0.0656 - dice_coefficient: 0.9566\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "122/400 [========>.....................] - ETA: 8:06 - loss: 0.0655 - dice_coefficient: 0.9568\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "123/400 [========>.....................] - ETA: 8:04 - loss: 0.0657 - dice_coefficient: 0.9566\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "124/400 [========>.....................] - ETA: 8:03 - loss: 0.0654 - dice_coefficient: 0.9568\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "125/400 [========>.....................] - ETA: 8:01 - loss: 0.0658 - dice_coefficient: 0.9566\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "126/400 [========>.....................] - ETA: 7:59 - loss: 0.0659 - dice_coefficient: 0.9565\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "127/400 [========>.....................] - ETA: 7:58 - loss: 0.0665 - dice_coefficient: 0.9563\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "128/400 [========>.....................] - ETA: 7:56 - loss: 0.0670 - dice_coefficient: 0.9560\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "129/400 [========>.....................] - ETA: 7:54 - loss: 0.0669 - dice_coefficient: 0.9560\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "130/400 [========>.....................] - ETA: 7:52 - loss: 0.0670 - dice_coefficient: 0.9560\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "131/400 [========>.....................] - ETA: 7:51 - loss: 0.0670 - dice_coefficient: 0.9561\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "132/400 [========>.....................] - ETA: 7:49 - loss: 0.0667 - dice_coefficient: 0.9562\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "133/400 [========>.....................] - ETA: 7:47 - loss: 0.0667 - dice_coefficient: 0.9562\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "134/400 [=========>....................] - ETA: 7:45 - loss: 0.0667 - dice_coefficient: 0.9561\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "135/400 [=========>....................] - ETA: 7:44 - loss: 0.0670 - dice_coefficient: 0.9562\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "136/400 [=========>....................] - ETA: 7:42 - loss: 0.0667 - dice_coefficient: 0.9564\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "137/400 [=========>....................] - ETA: 7:40 - loss: 0.0669 - dice_coefficient: 0.9563\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "138/400 [=========>....................] - ETA: 7:38 - loss: 0.0666 - dice_coefficient: 0.9564\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "139/400 [=========>....................] - ETA: 7:37 - loss: 0.0666 - dice_coefficient: 0.9564\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "140/400 [=========>....................] - ETA: 7:35 - loss: 0.0664 - dice_coefficient: 0.9566\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "141/400 [=========>....................] - ETA: 7:33 - loss: 0.0665 - dice_coefficient: 0.9567\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "142/400 [=========>....................] - ETA: 7:31 - loss: 0.0663 - dice_coefficient: 0.9568\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "143/400 [=========>....................] - ETA: 7:30 - loss: 0.0664 - dice_coefficient: 0.9568\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "144/400 [=========>....................] - ETA: 7:28 - loss: 0.0663 - dice_coefficient: 0.9569\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "145/400 [=========>....................] - ETA: 7:26 - loss: 0.0661 - dice_coefficient: 0.9571\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "146/400 [=========>....................] - ETA: 7:25 - loss: 0.0660 - dice_coefficient: 0.9572\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "147/400 [==========>...................] - ETA: 7:23 - loss: 0.0658 - dice_coefficient: 0.9574\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "148/400 [==========>...................] - ETA: 7:21 - loss: 0.0657 - dice_coefficient: 0.9574\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "149/400 [==========>...................] - ETA: 7:19 - loss: 0.0656 - dice_coefficient: 0.9574\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "150/400 [==========>...................] - ETA: 7:18 - loss: 0.0654 - dice_coefficient: 0.9575\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "151/400 [==========>...................] - ETA: 7:16 - loss: 0.0654 - dice_coefficient: 0.9574\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "152/400 [==========>...................] - ETA: 7:14 - loss: 0.0655 - dice_coefficient: 0.9572\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "153/400 [==========>...................] - ETA: 7:12 - loss: 0.0657 - dice_coefficient: 0.9571\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "154/400 [==========>...................] - ETA: 7:11 - loss: 0.0657 - dice_coefficient: 0.9572\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "155/400 [==========>...................] - ETA: 7:09 - loss: 0.0657 - dice_coefficient: 0.9572\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "156/400 [==========>...................] - ETA: 7:07 - loss: 0.0655 - dice_coefficient: 0.9573\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "157/400 [==========>...................] - ETA: 7:05 - loss: 0.0653 - dice_coefficient: 0.9574\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "158/400 [==========>...................] - ETA: 7:03 - loss: 0.0653 - dice_coefficient: 0.9574\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "159/400 [==========>...................] - ETA: 7:02 - loss: 0.0655 - dice_coefficient: 0.9572\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "160/400 [===========>..................] - ETA: 7:00 - loss: 0.0657 - dice_coefficient: 0.9572\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "161/400 [===========>..................] - ETA: 6:58 - loss: 0.1658 - dice_coefficient: 0.9512\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "162/400 [===========>..................] - ETA: 6:56 - loss: 0.1650 - dice_coefficient: 0.9514\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "163/400 [===========>..................] - ETA: 6:55 - loss: 0.1644 - dice_coefficient: 0.9515\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "164/400 [===========>..................] - ETA: 6:53 - loss: 0.1637 - dice_coefficient: 0.9516\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "165/400 [===========>..................] - ETA: 6:51 - loss: 0.1629 - dice_coefficient: 0.9517\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "166/400 [===========>..................] - ETA: 6:49 - loss: 0.1623 - dice_coefficient: 0.9517\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "167/400 [===========>..................] - ETA: 6:48 - loss: 0.1616 - dice_coefficient: 0.9518\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "168/400 [===========>..................] - ETA: 6:46 - loss: 0.1611 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "169/400 [===========>..................] - ETA: 6:44 - loss: 0.1604 - dice_coefficient: 0.9521\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "170/400 [===========>..................] - ETA: 6:42 - loss: 0.1597 - dice_coefficient: 0.9521\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "171/400 [===========>..................] - ETA: 6:40 - loss: 0.1592 - dice_coefficient: 0.9521\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "172/400 [===========>..................] - ETA: 6:39 - loss: 0.1587 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "173/400 [===========>..................] - ETA: 6:37 - loss: 0.1586 - dice_coefficient: 0.9519\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "174/400 [============>.................] - ETA: 6:35 - loss: 0.1580 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "175/400 [============>.................] - ETA: 6:33 - loss: 0.1577 - dice_coefficient: 0.9518\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "176/400 [============>.................] - ETA: 6:32 - loss: 0.1570 - dice_coefficient: 0.9519\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "177/400 [============>.................] - ETA: 6:30 - loss: 0.1565 - dice_coefficient: 0.9518\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "178/400 [============>.................] - ETA: 6:28 - loss: 0.1560 - dice_coefficient: 0.9518\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "179/400 [============>.................] - ETA: 6:26 - loss: 0.1558 - dice_coefficient: 0.9517\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "180/400 [============>.................] - ETA: 6:25 - loss: 0.1551 - dice_coefficient: 0.9518\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "181/400 [============>.................] - ETA: 6:23 - loss: 0.1549 - dice_coefficient: 0.9516\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "182/400 [============>.................] - ETA: 6:21 - loss: 0.1543 - dice_coefficient: 0.9517\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "183/400 [============>.................] - ETA: 6:19 - loss: 0.1536 - dice_coefficient: 0.9518\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "184/400 [============>.................] - ETA: 6:17 - loss: 0.1532 - dice_coefficient: 0.9519\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "185/400 [============>.................] - ETA: 6:16 - loss: 0.1526 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "186/400 [============>.................] - ETA: 6:14 - loss: 0.1521 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "187/400 [=============>................] - ETA: 6:12 - loss: 0.1516 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "188/400 [=============>................] - ETA: 6:10 - loss: 0.1511 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "189/400 [=============>................] - ETA: 6:09 - loss: 0.1505 - dice_coefficient: 0.9522\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "190/400 [=============>................] - ETA: 6:07 - loss: 0.1502 - dice_coefficient: 0.9522\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "191/400 [=============>................] - ETA: 6:05 - loss: 0.1498 - dice_coefficient: 0.9523\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "192/400 [=============>................] - ETA: 6:03 - loss: 0.1492 - dice_coefficient: 0.9524\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "193/400 [=============>................] - ETA: 6:02 - loss: 0.1486 - dice_coefficient: 0.9525\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "194/400 [=============>................] - ETA: 6:00 - loss: 0.1483 - dice_coefficient: 0.9523\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "195/400 [=============>................] - ETA: 5:58 - loss: 0.1480 - dice_coefficient: 0.9522\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "196/400 [=============>................] - ETA: 5:56 - loss: 0.1480 - dice_coefficient: 0.9518\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "197/400 [=============>................] - ETA: 5:55 - loss: 0.1475 - dice_coefficient: 0.9519\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "198/400 [=============>................] - ETA: 5:53 - loss: 0.1471 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "199/400 [=============>................] - ETA: 5:51 - loss: 0.1465 - dice_coefficient: 0.9521\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "200/400 [==============>...............] - ETA: 5:49 - loss: 0.1464 - dice_coefficient: 0.9519\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "201/400 [==============>...............] - ETA: 5:48 - loss: 0.1460 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "202/400 [==============>...............] - ETA: 5:46 - loss: 0.1455 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "203/400 [==============>...............] - ETA: 5:44 - loss: 0.1453 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "204/400 [==============>...............] - ETA: 5:43 - loss: 0.1448 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "205/400 [==============>...............] - ETA: 5:41 - loss: 0.1445 - dice_coefficient: 0.9520\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "206/400 [==============>...............] - ETA: 5:39 - loss: 0.1441 - dice_coefficient: 0.9521\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "207/400 [==============>...............] - ETA: 5:37 - loss: 0.1435 - dice_coefficient: 0.9522\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "208/400 [==============>...............] - ETA: 5:36 - loss: 0.1433 - dice_coefficient: 0.9521\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "209/400 [==============>...............] - ETA: 5:34 - loss: 0.1428 - dice_coefficient: 0.9522\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "210/400 [==============>...............] - ETA: 5:32 - loss: 0.1425 - dice_coefficient: 0.9522\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "211/400 [==============>...............] - ETA: 5:30 - loss: 0.1419 - dice_coefficient: 0.9523\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "212/400 [==============>...............] - ETA: 5:29 - loss: 0.1416 - dice_coefficient: 0.9524\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "213/400 [==============>...............] - ETA: 5:27 - loss: 0.1413 - dice_coefficient: 0.9524\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "214/400 [===============>..............] - ETA: 5:25 - loss: 0.1408 - dice_coefficient: 0.9525\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "215/400 [===============>..............] - ETA: 5:23 - loss: 0.1405 - dice_coefficient: 0.9526\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "216/400 [===============>..............] - ETA: 5:22 - loss: 0.1399 - dice_coefficient: 0.9527\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "217/400 [===============>..............] - ETA: 5:20 - loss: 0.1395 - dice_coefficient: 0.9528\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "218/400 [===============>..............] - ETA: 5:18 - loss: 0.1390 - dice_coefficient: 0.9528\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "219/400 [===============>..............] - ETA: 5:17 - loss: 0.1386 - dice_coefficient: 0.9530\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "220/400 [===============>..............] - ETA: 5:15 - loss: 0.1382 - dice_coefficient: 0.9530\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "221/400 [===============>..............] - ETA: 5:13 - loss: 0.1379 - dice_coefficient: 0.9529\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "222/400 [===============>..............] - ETA: 5:11 - loss: 0.1375 - dice_coefficient: 0.9530\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "223/400 [===============>..............] - ETA: 5:10 - loss: 0.1372 - dice_coefficient: 0.9529\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "224/400 [===============>..............] - ETA: 5:08 - loss: 0.1368 - dice_coefficient: 0.9529\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "225/400 [===============>..............] - ETA: 5:06 - loss: 0.1365 - dice_coefficient: 0.9529\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "226/400 [===============>..............] - ETA: 5:04 - loss: 0.1360 - dice_coefficient: 0.9530\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "227/400 [================>.............] - ETA: 5:03 - loss: 0.1358 - dice_coefficient: 0.9530\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "228/400 [================>.............] - ETA: 5:01 - loss: 0.1354 - dice_coefficient: 0.9530\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "229/400 [================>.............] - ETA: 4:59 - loss: 0.1351 - dice_coefficient: 0.9530\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "230/400 [================>.............] - ETA: 4:57 - loss: 0.1348 - dice_coefficient: 0.9531\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "231/400 [================>.............] - ETA: 4:56 - loss: 0.1346 - dice_coefficient: 0.9530\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "232/400 [================>.............] - ETA: 4:54 - loss: 0.1342 - dice_coefficient: 0.9531\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "233/400 [================>.............] - ETA: 4:52 - loss: 0.1339 - dice_coefficient: 0.9531\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "234/400 [================>.............] - ETA: 4:50 - loss: 0.1335 - dice_coefficient: 0.9532\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "235/400 [================>.............] - ETA: 4:49 - loss: 0.1330 - dice_coefficient: 0.9534\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "236/400 [================>.............] - ETA: 4:47 - loss: 0.1326 - dice_coefficient: 0.9535\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "237/400 [================>.............] - ETA: 4:45 - loss: 0.1322 - dice_coefficient: 0.9536\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "238/400 [================>.............] - ETA: 4:43 - loss: 0.1319 - dice_coefficient: 0.9536\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "239/400 [================>.............] - ETA: 4:42 - loss: 0.1316 - dice_coefficient: 0.9535\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "240/400 [=================>............] - ETA: 4:40 - loss: 0.1312 - dice_coefficient: 0.9536\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "241/400 [=================>............] - ETA: 4:38 - loss: 0.1309 - dice_coefficient: 0.9536\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "242/400 [=================>............] - ETA: 4:36 - loss: 0.1306 - dice_coefficient: 0.9537\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "243/400 [=================>............] - ETA: 4:35 - loss: 0.1302 - dice_coefficient: 0.9538\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "244/400 [=================>............] - ETA: 4:33 - loss: 0.1299 - dice_coefficient: 0.9538\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "245/400 [=================>............] - ETA: 4:31 - loss: 0.1295 - dice_coefficient: 0.9540\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "246/400 [=================>............] - ETA: 4:29 - loss: 0.1291 - dice_coefficient: 0.9540\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "247/400 [=================>............] - ETA: 4:28 - loss: 0.1288 - dice_coefficient: 0.9540\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "248/400 [=================>............] - ETA: 4:26 - loss: 0.1286 - dice_coefficient: 0.9541\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "249/400 [=================>............] - ETA: 4:24 - loss: 0.1283 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "250/400 [=================>............] - ETA: 4:22 - loss: 0.1279 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "251/400 [=================>............] - ETA: 4:21 - loss: 0.1276 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "252/400 [=================>............] - ETA: 4:19 - loss: 0.1272 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "253/400 [=================>............] - ETA: 4:17 - loss: 0.1270 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "254/400 [==================>...........] - ETA: 4:15 - loss: 0.1267 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "255/400 [==================>...........] - ETA: 4:14 - loss: 0.1264 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "256/400 [==================>...........] - ETA: 4:12 - loss: 0.1260 - dice_coefficient: 0.9545\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "257/400 [==================>...........] - ETA: 4:10 - loss: 0.1258 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "258/400 [==================>...........] - ETA: 4:08 - loss: 0.1254 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "259/400 [==================>...........] - ETA: 4:07 - loss: 0.1251 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "260/400 [==================>...........] - ETA: 4:05 - loss: 0.1249 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "261/400 [==================>...........] - ETA: 4:03 - loss: 0.1245 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "262/400 [==================>...........] - ETA: 4:01 - loss: 0.1242 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "263/400 [==================>...........] - ETA: 4:00 - loss: 0.1240 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "264/400 [==================>...........] - ETA: 3:58 - loss: 0.1237 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "265/400 [==================>...........] - ETA: 3:56 - loss: 0.1233 - dice_coefficient: 0.9551\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "266/400 [==================>...........] - ETA: 3:54 - loss: 0.1231 - dice_coefficient: 0.9551\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "267/400 [===================>..........] - ETA: 3:53 - loss: 0.1228 - dice_coefficient: 0.9552\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "268/400 [===================>..........] - ETA: 3:51 - loss: 0.1231 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "269/400 [===================>..........] - ETA: 3:49 - loss: 0.1228 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "270/400 [===================>..........] - ETA: 3:47 - loss: 0.1225 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "271/400 [===================>..........] - ETA: 3:46 - loss: 0.1223 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "272/400 [===================>..........] - ETA: 3:44 - loss: 0.1222 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "273/400 [===================>..........] - ETA: 3:42 - loss: 0.1219 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "274/400 [===================>..........] - ETA: 3:40 - loss: 0.1218 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "275/400 [===================>..........] - ETA: 3:39 - loss: 0.1216 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "276/400 [===================>..........] - ETA: 3:37 - loss: 0.1213 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "277/400 [===================>..........] - ETA: 3:35 - loss: 0.1211 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "278/400 [===================>..........] - ETA: 3:33 - loss: 0.1210 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "279/400 [===================>..........] - ETA: 3:32 - loss: 0.1208 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "280/400 [====================>.........] - ETA: 3:30 - loss: 0.1205 - dice_coefficient: 0.9550\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "281/400 [====================>.........] - ETA: 3:28 - loss: 0.1204 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "282/400 [====================>.........] - ETA: 3:26 - loss: 0.1201 - dice_coefficient: 0.9550\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "283/400 [====================>.........] - ETA: 3:25 - loss: 0.1201 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "284/400 [====================>.........] - ETA: 3:23 - loss: 0.1199 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "285/400 [====================>.........] - ETA: 3:21 - loss: 0.1197 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "286/400 [====================>.........] - ETA: 3:19 - loss: 0.1194 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "287/400 [====================>.........] - ETA: 3:17 - loss: 0.1193 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "288/400 [====================>.........] - ETA: 3:16 - loss: 0.1191 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "289/400 [====================>.........] - ETA: 3:14 - loss: 0.1188 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "290/400 [====================>.........] - ETA: 3:12 - loss: 0.1189 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "291/400 [====================>.........] - ETA: 3:10 - loss: 0.1187 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "292/400 [====================>.........] - ETA: 3:09 - loss: 0.1184 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "293/400 [====================>.........] - ETA: 3:07 - loss: 0.1185 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "294/400 [=====================>........] - ETA: 3:05 - loss: 0.1184 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "295/400 [=====================>........] - ETA: 3:03 - loss: 0.1181 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "296/400 [=====================>........] - ETA: 3:02 - loss: 0.1180 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "297/400 [=====================>........] - ETA: 3:00 - loss: 0.1179 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "298/400 [=====================>........] - ETA: 2:58 - loss: 0.1176 - dice_coefficient: 0.9545\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "299/400 [=====================>........] - ETA: 2:56 - loss: 0.1174 - dice_coefficient: 0.9545\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "300/400 [=====================>........] - ETA: 2:55 - loss: 0.1171 - dice_coefficient: 0.9545\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "301/400 [=====================>........] - ETA: 2:53 - loss: 0.1170 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "302/400 [=====================>........] - ETA: 2:51 - loss: 0.1167 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "303/400 [=====================>........] - ETA: 2:49 - loss: 0.1165 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "304/400 [=====================>........] - ETA: 2:48 - loss: 0.1163 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "305/400 [=====================>........] - ETA: 2:46 - loss: 0.1162 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "306/400 [=====================>........] - ETA: 2:44 - loss: 0.1160 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "307/400 [======================>.......] - ETA: 2:42 - loss: 0.1159 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "308/400 [======================>.......] - ETA: 2:41 - loss: 0.1157 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "309/400 [======================>.......] - ETA: 2:39 - loss: 0.1155 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "310/400 [======================>.......] - ETA: 2:37 - loss: 0.1153 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "311/400 [======================>.......] - ETA: 2:35 - loss: 0.1152 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "312/400 [======================>.......] - ETA: 2:34 - loss: 0.1149 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "313/400 [======================>.......] - ETA: 2:32 - loss: 0.1147 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "314/400 [======================>.......] - ETA: 2:30 - loss: 0.1148 - dice_coefficient: 0.9545\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "315/400 [======================>.......] - ETA: 2:28 - loss: 0.1147 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "316/400 [======================>.......] - ETA: 2:27 - loss: 0.1145 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "317/400 [======================>.......] - ETA: 2:25 - loss: 0.1144 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "318/400 [======================>.......] - ETA: 2:23 - loss: 0.1144 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "319/400 [======================>.......] - ETA: 2:21 - loss: 0.1143 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "320/400 [=======================>......] - ETA: 2:20 - loss: 0.1141 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "321/400 [=======================>......] - ETA: 2:18 - loss: 0.1140 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "322/400 [=======================>......] - ETA: 2:16 - loss: 0.1139 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "323/400 [=======================>......] - ETA: 2:14 - loss: 0.1139 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "324/400 [=======================>......] - ETA: 2:13 - loss: 0.1137 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "325/400 [=======================>......] - ETA: 2:11 - loss: 0.1136 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "326/400 [=======================>......] - ETA: 2:09 - loss: 0.1134 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "327/400 [=======================>......] - ETA: 2:07 - loss: 0.1131 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "328/400 [=======================>......] - ETA: 2:06 - loss: 0.1133 - dice_coefficient: 0.9541\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "329/400 [=======================>......] - ETA: 2:04 - loss: 0.1132 - dice_coefficient: 0.9541\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "330/400 [=======================>......] - ETA: 2:02 - loss: 0.1129 - dice_coefficient: 0.9541\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "331/400 [=======================>......] - ETA: 2:00 - loss: 0.1128 - dice_coefficient: 0.9541\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "332/400 [=======================>......] - ETA: 1:59 - loss: 0.1126 - dice_coefficient: 0.9541\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "333/400 [=======================>......] - ETA: 1:57 - loss: 0.1124 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "334/400 [========================>.....] - ETA: 1:55 - loss: 0.1121 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "335/400 [========================>.....] - ETA: 1:53 - loss: 0.1118 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "336/400 [========================>.....] - ETA: 1:52 - loss: 0.1117 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "337/400 [========================>.....] - ETA: 1:50 - loss: 0.1118 - dice_coefficient: 0.9541\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "338/400 [========================>.....] - ETA: 1:48 - loss: 0.1116 - dice_coefficient: 0.9542\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "339/400 [========================>.....] - ETA: 1:46 - loss: 0.1113 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "340/400 [========================>.....] - ETA: 1:45 - loss: 0.1112 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "341/400 [========================>.....] - ETA: 1:43 - loss: 0.1111 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "342/400 [========================>.....] - ETA: 1:41 - loss: 0.1108 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "343/400 [========================>.....] - ETA: 1:39 - loss: 0.1107 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "344/400 [========================>.....] - ETA: 1:37 - loss: 0.1106 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "345/400 [========================>.....] - ETA: 1:36 - loss: 0.1104 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "346/400 [========================>.....] - ETA: 1:34 - loss: 0.1102 - dice_coefficient: 0.9545\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "347/400 [=========================>....] - ETA: 1:32 - loss: 0.1100 - dice_coefficient: 0.9545\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "348/400 [=========================>....] - ETA: 1:30 - loss: 0.1097 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "349/400 [=========================>....] - ETA: 1:29 - loss: 0.1096 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "350/400 [=========================>....] - ETA: 1:27 - loss: 0.1097 - dice_coefficient: 0.9545\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "351/400 [=========================>....] - ETA: 1:25 - loss: 0.1094 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "352/400 [=========================>....] - ETA: 1:23 - loss: 0.1093 - dice_coefficient: 0.9546\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "353/400 [=========================>....] - ETA: 1:22 - loss: 0.1090 - dice_coefficient: 0.9547\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "354/400 [=========================>....] - ETA: 1:20 - loss: 0.1088 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "355/400 [=========================>....] - ETA: 1:18 - loss: 0.1086 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "356/400 [=========================>....] - ETA: 1:16 - loss: 0.1084 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "357/400 [=========================>....] - ETA: 1:15 - loss: 0.1082 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "358/400 [=========================>....] - ETA: 1:13 - loss: 0.1080 - dice_coefficient: 0.9550\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "359/400 [=========================>....] - ETA: 1:11 - loss: 0.1079 - dice_coefficient: 0.9550\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "360/400 [==========================>...] - ETA: 1:09 - loss: 0.1076 - dice_coefficient: 0.9551\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "361/400 [==========================>...] - ETA: 1:08 - loss: 0.1075 - dice_coefficient: 0.9551\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "362/400 [==========================>...] - ETA: 1:06 - loss: 0.1073 - dice_coefficient: 0.9551\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "363/400 [==========================>...] - ETA: 1:04 - loss: 0.1071 - dice_coefficient: 0.9552\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "364/400 [==========================>...] - ETA: 1:02 - loss: 0.1071 - dice_coefficient: 0.9551\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "365/400 [==========================>...] - ETA: 1:01 - loss: 0.1069 - dice_coefficient: 0.9552\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "366/400 [==========================>...] - ETA: 59s - loss: 0.1068 - dice_coefficient: 0.9551 \n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "367/400 [==========================>...] - ETA: 57s - loss: 0.1066 - dice_coefficient: 0.9552\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "368/400 [==========================>...] - ETA: 55s - loss: 0.1064 - dice_coefficient: 0.9552\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "369/400 [==========================>...] - ETA: 54s - loss: 0.1063 - dice_coefficient: 0.9552\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "370/400 [==========================>...] - ETA: 52s - loss: 0.1062 - dice_coefficient: 0.9552\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "371/400 [==========================>...] - ETA: 50s - loss: 0.1059 - dice_coefficient: 0.9553\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "372/400 [==========================>...] - ETA: 48s - loss: 0.1062 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "373/400 [==========================>...] - ETA: 47s - loss: 0.1061 - dice_coefficient: 0.9550\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "374/400 [===========================>..] - ETA: 45s - loss: 0.1059 - dice_coefficient: 0.9550\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "375/400 [===========================>..] - ETA: 43s - loss: 0.1058 - dice_coefficient: 0.9550\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "376/400 [===========================>..] - ETA: 41s - loss: 0.1058 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "377/400 [===========================>..] - ETA: 40s - loss: 0.1057 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "378/400 [===========================>..] - ETA: 38s - loss: 0.1056 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "379/400 [===========================>..] - ETA: 36s - loss: 0.1055 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "380/400 [===========================>..] - ETA: 34s - loss: 0.1054 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "381/400 [===========================>..] - ETA: 33s - loss: 0.1052 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "382/400 [===========================>..] - ETA: 31s - loss: 0.1051 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "383/400 [===========================>..] - ETA: 29s - loss: 0.1049 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "384/400 [===========================>..] - ETA: 27s - loss: 0.1049 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "385/400 [===========================>..] - ETA: 26s - loss: 0.1048 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "386/400 [===========================>..] - ETA: 24s - loss: 0.1046 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "387/400 [============================>.] - ETA: 22s - loss: 0.1045 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "388/400 [============================>.] - ETA: 20s - loss: 0.1045 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "389/400 [============================>.] - ETA: 19s - loss: 0.1043 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "390/400 [============================>.] - ETA: 17s - loss: 0.1041 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "391/400 [============================>.] - ETA: 15s - loss: 0.1040 - dice_coefficient: 0.9549\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "392/400 [============================>.] - ETA: 13s - loss: 0.1040 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "393/400 [============================>.] - ETA: 12s - loss: 0.1040 - dice_coefficient: 0.9548\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "394/400 [============================>.] - ETA: 10s - loss: 0.1044 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "395/400 [============================>.] - ETA: 8s - loss: 0.1043 - dice_coefficient: 0.9544 \n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "396/400 [============================>.] - ETA: 6s - loss: 0.1041 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "397/400 [============================>.] - ETA: 5s - loss: 0.1040 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "398/400 [============================>.] - ETA: 3s - loss: 0.1039 - dice_coefficient: 0.9544\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "399/400 [============================>.] - ETA: 1s - loss: 0.1038 - dice_coefficient: 0.9543\n",
      "Epoch 00027: loss did not improve from 0.04222\n",
      "400/400 [==============================] - 703s 2s/step - loss: 0.1037 - dice_coefficient: 0.9544 - val_loss: 1.3674 - val_dice_coefficient: 0.5042\n",
      "Epoch 28/30\n",
      "\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "  1/400 [..............................] - ETA: 0s - loss: 0.0546 - dice_coefficient: 0.9675\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "  2/400 [..............................] - ETA: 5:38 - loss: 0.0611 - dice_coefficient: 0.9633\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "  3/400 [..............................] - ETA: 7:33 - loss: 0.0488 - dice_coefficient: 0.9684\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "  4/400 [..............................] - ETA: 8:28 - loss: 0.0425 - dice_coefficient: 0.9711\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "  5/400 [..............................] - ETA: 9:01 - loss: 0.0566 - dice_coefficient: 0.9659\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "  6/400 [..............................] - ETA: 9:24 - loss: 0.0601 - dice_coefficient: 0.9661\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "  7/400 [..............................] - ETA: 9:39 - loss: 0.0643 - dice_coefficient: 0.9634\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "  8/400 [..............................] - ETA: 9:49 - loss: 0.0646 - dice_coefficient: 0.9640\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "  9/400 [..............................] - ETA: 9:58 - loss: 0.0637 - dice_coefficient: 0.9645\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 10/400 [..............................] - ETA: 10:04 - loss: 0.0616 - dice_coefficient: 0.9647\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 11/400 [..............................] - ETA: 10:09 - loss: 0.0609 - dice_coefficient: 0.9640\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 12/400 [..............................] - ETA: 10:13 - loss: 0.0609 - dice_coefficient: 0.9628\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 13/400 [..............................] - ETA: 10:15 - loss: 0.0597 - dice_coefficient: 0.9625\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 14/400 [>.............................] - ETA: 10:18 - loss: 0.0604 - dice_coefficient: 0.9623\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 15/400 [>.............................] - ETA: 10:19 - loss: 0.0582 - dice_coefficient: 0.9634\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 16/400 [>.............................] - ETA: 10:20 - loss: 0.0610 - dice_coefficient: 0.9633\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 17/400 [>.............................] - ETA: 10:21 - loss: 0.0596 - dice_coefficient: 0.9633\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 18/400 [>.............................] - ETA: 10:22 - loss: 0.0577 - dice_coefficient: 0.9644\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 19/400 [>.............................] - ETA: 10:22 - loss: 0.0609 - dice_coefficient: 0.9634\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 20/400 [>.............................] - ETA: 10:23 - loss: 0.0613 - dice_coefficient: 0.9622\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 21/400 [>.............................] - ETA: 10:22 - loss: 0.0622 - dice_coefficient: 0.9616\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 22/400 [>.............................] - ETA: 10:22 - loss: 0.0616 - dice_coefficient: 0.9619\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 23/400 [>.............................] - ETA: 10:22 - loss: 0.0617 - dice_coefficient: 0.9624\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 24/400 [>.............................] - ETA: 10:21 - loss: 0.0610 - dice_coefficient: 0.9630\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 25/400 [>.............................] - ETA: 10:21 - loss: 0.0606 - dice_coefficient: 0.9634\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 26/400 [>.............................] - ETA: 10:20 - loss: 0.0613 - dice_coefficient: 0.9626\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 27/400 [=>............................] - ETA: 10:19 - loss: 0.0624 - dice_coefficient: 0.9626\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 28/400 [=>............................] - ETA: 10:18 - loss: 0.0648 - dice_coefficient: 0.9607\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 29/400 [=>............................] - ETA: 10:17 - loss: 0.0672 - dice_coefficient: 0.9582\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 30/400 [=>............................] - ETA: 10:17 - loss: 0.0686 - dice_coefficient: 0.9566\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 31/400 [=>............................] - ETA: 10:16 - loss: 0.0697 - dice_coefficient: 0.9552\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 32/400 [=>............................] - ETA: 10:15 - loss: 0.0696 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 33/400 [=>............................] - ETA: 10:14 - loss: 0.0690 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 34/400 [=>............................] - ETA: 10:12 - loss: 0.0694 - dice_coefficient: 0.9554\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 35/400 [=>............................] - ETA: 10:11 - loss: 0.0693 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 36/400 [=>............................] - ETA: 10:10 - loss: 0.0699 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 37/400 [=>............................] - ETA: 10:09 - loss: 0.0696 - dice_coefficient: 0.9561\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 38/400 [=>............................] - ETA: 10:08 - loss: 0.0689 - dice_coefficient: 0.9564\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 39/400 [=>............................] - ETA: 10:07 - loss: 0.0688 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 40/400 [==>...........................] - ETA: 10:06 - loss: 0.0676 - dice_coefficient: 0.9568\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 41/400 [==>...........................] - ETA: 10:05 - loss: 0.0685 - dice_coefficient: 0.9566\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 42/400 [==>...........................] - ETA: 10:03 - loss: 0.0678 - dice_coefficient: 0.9570\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 43/400 [==>...........................] - ETA: 10:02 - loss: 0.0698 - dice_coefficient: 0.9562\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 44/400 [==>...........................] - ETA: 10:01 - loss: 0.0693 - dice_coefficient: 0.9563\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 45/400 [==>...........................] - ETA: 9:59 - loss: 0.0690 - dice_coefficient: 0.9564 \n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 46/400 [==>...........................] - ETA: 9:58 - loss: 0.0680 - dice_coefficient: 0.9568\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 47/400 [==>...........................] - ETA: 9:57 - loss: 0.0683 - dice_coefficient: 0.9570\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 48/400 [==>...........................] - ETA: 9:55 - loss: 0.0683 - dice_coefficient: 0.9571\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 49/400 [==>...........................] - ETA: 9:54 - loss: 0.0695 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 50/400 [==>...........................] - ETA: 9:53 - loss: 0.0703 - dice_coefficient: 0.9550\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 51/400 [==>...........................] - ETA: 9:51 - loss: 0.0704 - dice_coefficient: 0.9550\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 52/400 [==>...........................] - ETA: 9:50 - loss: 0.0710 - dice_coefficient: 0.9549\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 53/400 [==>...........................] - ETA: 9:48 - loss: 0.0707 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 54/400 [===>..........................] - ETA: 9:47 - loss: 0.0712 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 55/400 [===>..........................] - ETA: 9:45 - loss: 0.0760 - dice_coefficient: 0.9502\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 56/400 [===>..........................] - ETA: 9:44 - loss: 0.0761 - dice_coefficient: 0.9499\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 57/400 [===>..........................] - ETA: 9:42 - loss: 0.0781 - dice_coefficient: 0.9478\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 58/400 [===>..........................] - ETA: 9:41 - loss: 0.0771 - dice_coefficient: 0.9485\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 59/400 [===>..........................] - ETA: 9:39 - loss: 0.0764 - dice_coefficient: 0.9489\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 60/400 [===>..........................] - ETA: 9:38 - loss: 0.0761 - dice_coefficient: 0.9491\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 61/400 [===>..........................] - ETA: 9:37 - loss: 0.0754 - dice_coefficient: 0.9494\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 62/400 [===>..........................] - ETA: 9:35 - loss: 0.0747 - dice_coefficient: 0.9500\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 63/400 [===>..........................] - ETA: 9:33 - loss: 0.0744 - dice_coefficient: 0.9502\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 64/400 [===>..........................] - ETA: 9:32 - loss: 0.0743 - dice_coefficient: 0.9504\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 65/400 [===>..........................] - ETA: 9:30 - loss: 0.0743 - dice_coefficient: 0.9505\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 66/400 [===>..........................] - ETA: 9:29 - loss: 0.0745 - dice_coefficient: 0.9502\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 67/400 [====>.........................] - ETA: 9:27 - loss: 0.0743 - dice_coefficient: 0.9502\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 68/400 [====>.........................] - ETA: 9:26 - loss: 0.0736 - dice_coefficient: 0.9507\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 69/400 [====>.........................] - ETA: 9:24 - loss: 0.0733 - dice_coefficient: 0.9511\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 70/400 [====>.........................] - ETA: 9:23 - loss: 0.0732 - dice_coefficient: 0.9511\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 71/400 [====>.........................] - ETA: 9:21 - loss: 0.0725 - dice_coefficient: 0.9515\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 72/400 [====>.........................] - ETA: 9:19 - loss: 0.0725 - dice_coefficient: 0.9514\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 73/400 [====>.........................] - ETA: 9:18 - loss: 0.0726 - dice_coefficient: 0.9512\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 74/400 [====>.........................] - ETA: 9:16 - loss: 0.0720 - dice_coefficient: 0.9516\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 75/400 [====>.........................] - ETA: 9:15 - loss: 0.0714 - dice_coefficient: 0.9521\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 76/400 [====>.........................] - ETA: 9:13 - loss: 0.0716 - dice_coefficient: 0.9518\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 77/400 [====>.........................] - ETA: 9:12 - loss: 0.0713 - dice_coefficient: 0.9519\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 78/400 [====>.........................] - ETA: 9:10 - loss: 0.0711 - dice_coefficient: 0.9522\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 79/400 [====>.........................] - ETA: 9:08 - loss: 0.0717 - dice_coefficient: 0.9520\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 80/400 [=====>........................] - ETA: 9:07 - loss: 0.0715 - dice_coefficient: 0.9522\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 81/400 [=====>........................] - ETA: 9:05 - loss: 0.0714 - dice_coefficient: 0.9523\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 82/400 [=====>........................] - ETA: 9:04 - loss: 0.0713 - dice_coefficient: 0.9523\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 83/400 [=====>........................] - ETA: 9:02 - loss: 0.0710 - dice_coefficient: 0.9526\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 84/400 [=====>........................] - ETA: 9:00 - loss: 0.0707 - dice_coefficient: 0.9529\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 85/400 [=====>........................] - ETA: 8:59 - loss: 0.0707 - dice_coefficient: 0.9527\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 86/400 [=====>........................] - ETA: 8:57 - loss: 0.0711 - dice_coefficient: 0.9524\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 87/400 [=====>........................] - ETA: 8:56 - loss: 0.0709 - dice_coefficient: 0.9526\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 88/400 [=====>........................] - ETA: 8:54 - loss: 0.0712 - dice_coefficient: 0.9525\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 89/400 [=====>........................] - ETA: 8:52 - loss: 0.0709 - dice_coefficient: 0.9527\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 90/400 [=====>........................] - ETA: 8:51 - loss: 0.0711 - dice_coefficient: 0.9523\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 91/400 [=====>........................] - ETA: 8:49 - loss: 0.0708 - dice_coefficient: 0.9526\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 92/400 [=====>........................] - ETA: 8:47 - loss: 0.0707 - dice_coefficient: 0.9526\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 93/400 [=====>........................] - ETA: 8:46 - loss: 0.0714 - dice_coefficient: 0.9518\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 94/400 [======>.......................] - ETA: 8:44 - loss: 0.0713 - dice_coefficient: 0.9517\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 95/400 [======>.......................] - ETA: 8:43 - loss: 0.0710 - dice_coefficient: 0.9520\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 96/400 [======>.......................] - ETA: 8:41 - loss: 0.0707 - dice_coefficient: 0.9523\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 97/400 [======>.......................] - ETA: 8:39 - loss: 0.0717 - dice_coefficient: 0.9513\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 98/400 [======>.......................] - ETA: 8:38 - loss: 0.0714 - dice_coefficient: 0.9515\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      " 99/400 [======>.......................] - ETA: 8:36 - loss: 0.0715 - dice_coefficient: 0.9517\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "100/400 [======>.......................] - ETA: 8:34 - loss: 0.0718 - dice_coefficient: 0.9515\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "101/400 [======>.......................] - ETA: 8:33 - loss: 0.0716 - dice_coefficient: 0.9516\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "102/400 [======>.......................] - ETA: 8:31 - loss: 0.0715 - dice_coefficient: 0.9518\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "103/400 [======>.......................] - ETA: 8:29 - loss: 0.0717 - dice_coefficient: 0.9518\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "104/400 [======>.......................] - ETA: 8:28 - loss: 0.0712 - dice_coefficient: 0.9521\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "105/400 [======>.......................] - ETA: 8:26 - loss: 0.0708 - dice_coefficient: 0.9524\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "106/400 [======>.......................] - ETA: 8:25 - loss: 0.0706 - dice_coefficient: 0.9525\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "107/400 [=======>......................] - ETA: 8:23 - loss: 0.0710 - dice_coefficient: 0.9522\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "108/400 [=======>......................] - ETA: 8:21 - loss: 0.0710 - dice_coefficient: 0.9523\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "109/400 [=======>......................] - ETA: 8:20 - loss: 0.0708 - dice_coefficient: 0.9523\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "110/400 [=======>......................] - ETA: 8:18 - loss: 0.0705 - dice_coefficient: 0.9525\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "111/400 [=======>......................] - ETA: 8:16 - loss: 0.0703 - dice_coefficient: 0.9527\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "112/400 [=======>......................] - ETA: 8:15 - loss: 0.0699 - dice_coefficient: 0.9530\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "113/400 [=======>......................] - ETA: 8:13 - loss: 0.0698 - dice_coefficient: 0.9532\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "114/400 [=======>......................] - ETA: 8:11 - loss: 0.0694 - dice_coefficient: 0.9534\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "115/400 [=======>......................] - ETA: 8:10 - loss: 0.0691 - dice_coefficient: 0.9536\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "116/400 [=======>......................] - ETA: 8:08 - loss: 0.0690 - dice_coefficient: 0.9537\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "117/400 [=======>......................] - ETA: 8:06 - loss: 0.0692 - dice_coefficient: 0.9535\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "118/400 [=======>......................] - ETA: 8:05 - loss: 0.0689 - dice_coefficient: 0.9536\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "119/400 [=======>......................] - ETA: 8:03 - loss: 0.0688 - dice_coefficient: 0.9536\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "120/400 [========>.....................] - ETA: 8:01 - loss: 0.0691 - dice_coefficient: 0.9534\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "121/400 [========>.....................] - ETA: 8:00 - loss: 0.0688 - dice_coefficient: 0.9536\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "122/400 [========>.....................] - ETA: 7:58 - loss: 0.0685 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "123/400 [========>.....................] - ETA: 7:56 - loss: 0.0684 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "124/400 [========>.....................] - ETA: 7:55 - loss: 0.0682 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "125/400 [========>.....................] - ETA: 7:53 - loss: 0.0683 - dice_coefficient: 0.9539\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "126/400 [========>.....................] - ETA: 7:51 - loss: 0.0683 - dice_coefficient: 0.9539\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "127/400 [========>.....................] - ETA: 7:50 - loss: 0.0681 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "128/400 [========>.....................] - ETA: 7:48 - loss: 0.0682 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "129/400 [========>.....................] - ETA: 7:46 - loss: 0.0683 - dice_coefficient: 0.9536\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "130/400 [========>.....................] - ETA: 7:45 - loss: 0.0685 - dice_coefficient: 0.9536\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "131/400 [========>.....................] - ETA: 7:43 - loss: 0.0682 - dice_coefficient: 0.9537\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "132/400 [========>.....................] - ETA: 7:41 - loss: 0.0678 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "133/400 [========>.....................] - ETA: 7:40 - loss: 0.0676 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "134/400 [=========>....................] - ETA: 7:38 - loss: 0.0679 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "135/400 [=========>....................] - ETA: 7:36 - loss: 0.0678 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "136/400 [=========>....................] - ETA: 7:35 - loss: 0.0678 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "137/400 [=========>....................] - ETA: 7:33 - loss: 0.0678 - dice_coefficient: 0.9543\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "138/400 [=========>....................] - ETA: 7:31 - loss: 0.0676 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "139/400 [=========>....................] - ETA: 7:30 - loss: 0.0678 - dice_coefficient: 0.9543\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "140/400 [=========>....................] - ETA: 7:28 - loss: 0.0677 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "141/400 [=========>....................] - ETA: 7:26 - loss: 0.0679 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "142/400 [=========>....................] - ETA: 7:25 - loss: 0.0681 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "143/400 [=========>....................] - ETA: 7:23 - loss: 0.0678 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "144/400 [=========>....................] - ETA: 7:21 - loss: 0.0679 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "145/400 [=========>....................] - ETA: 7:20 - loss: 0.0676 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "146/400 [=========>....................] - ETA: 7:18 - loss: 0.0673 - dice_coefficient: 0.9546\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "147/400 [==========>...................] - ETA: 7:16 - loss: 0.0671 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "148/400 [==========>...................] - ETA: 7:15 - loss: 0.0670 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "149/400 [==========>...................] - ETA: 7:13 - loss: 0.0670 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "150/400 [==========>...................] - ETA: 7:12 - loss: 0.0669 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "151/400 [==========>...................] - ETA: 7:10 - loss: 0.0670 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "152/400 [==========>...................] - ETA: 7:08 - loss: 0.0670 - dice_coefficient: 0.9546\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "153/400 [==========>...................] - ETA: 7:06 - loss: 0.0675 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "154/400 [==========>...................] - ETA: 7:05 - loss: 0.0672 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "155/400 [==========>...................] - ETA: 7:03 - loss: 0.0670 - dice_coefficient: 0.9545\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "156/400 [==========>...................] - ETA: 7:01 - loss: 0.0668 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "157/400 [==========>...................] - ETA: 7:00 - loss: 0.0666 - dice_coefficient: 0.9549\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "158/400 [==========>...................] - ETA: 6:58 - loss: 0.0667 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "159/400 [==========>...................] - ETA: 6:56 - loss: 0.0667 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "160/400 [===========>..................] - ETA: 6:55 - loss: 0.0668 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "161/400 [===========>..................] - ETA: 6:53 - loss: 0.0667 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "162/400 [===========>..................] - ETA: 6:51 - loss: 0.0667 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "163/400 [===========>..................] - ETA: 6:49 - loss: 0.0667 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "164/400 [===========>..................] - ETA: 6:48 - loss: 0.0664 - dice_coefficient: 0.9550\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "165/400 [===========>..................] - ETA: 6:46 - loss: 0.0664 - dice_coefficient: 0.9549\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "166/400 [===========>..................] - ETA: 6:44 - loss: 0.0664 - dice_coefficient: 0.9549\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "167/400 [===========>..................] - ETA: 6:43 - loss: 0.0663 - dice_coefficient: 0.9550\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "168/400 [===========>..................] - ETA: 6:41 - loss: 0.0660 - dice_coefficient: 0.9552\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "169/400 [===========>..................] - ETA: 6:39 - loss: 0.0658 - dice_coefficient: 0.9553\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "170/400 [===========>..................] - ETA: 6:37 - loss: 0.0658 - dice_coefficient: 0.9554\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "171/400 [===========>..................] - ETA: 6:36 - loss: 0.0656 - dice_coefficient: 0.9555\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "172/400 [===========>..................] - ETA: 6:34 - loss: 0.0656 - dice_coefficient: 0.9555\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "173/400 [===========>..................] - ETA: 6:32 - loss: 0.0657 - dice_coefficient: 0.9553\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "174/400 [============>.................] - ETA: 6:30 - loss: 0.0658 - dice_coefficient: 0.9552\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "175/400 [============>.................] - ETA: 6:29 - loss: 0.0656 - dice_coefficient: 0.9553\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "176/400 [============>.................] - ETA: 6:27 - loss: 0.0654 - dice_coefficient: 0.9554\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "177/400 [============>.................] - ETA: 6:25 - loss: 0.0654 - dice_coefficient: 0.9554\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "178/400 [============>.................] - ETA: 6:24 - loss: 0.0654 - dice_coefficient: 0.9555\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "179/400 [============>.................] - ETA: 6:22 - loss: 0.0653 - dice_coefficient: 0.9555\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "180/400 [============>.................] - ETA: 6:20 - loss: 0.0652 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "181/400 [============>.................] - ETA: 6:19 - loss: 0.0650 - dice_coefficient: 0.9557\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "182/400 [============>.................] - ETA: 6:17 - loss: 0.0650 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "183/400 [============>.................] - ETA: 6:15 - loss: 0.0648 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "184/400 [============>.................] - ETA: 6:13 - loss: 0.0649 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "185/400 [============>.................] - ETA: 6:12 - loss: 0.0650 - dice_coefficient: 0.9557\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "186/400 [============>.................] - ETA: 6:10 - loss: 0.0652 - dice_coefficient: 0.9554\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "187/400 [=============>................] - ETA: 6:08 - loss: 0.0652 - dice_coefficient: 0.9554\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "188/400 [=============>................] - ETA: 6:06 - loss: 0.0651 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "189/400 [=============>................] - ETA: 6:05 - loss: 0.0649 - dice_coefficient: 0.9557\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "190/400 [=============>................] - ETA: 6:03 - loss: 0.0648 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "191/400 [=============>................] - ETA: 6:01 - loss: 0.0649 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "192/400 [=============>................] - ETA: 6:00 - loss: 0.0647 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "193/400 [=============>................] - ETA: 5:58 - loss: 0.0646 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "194/400 [=============>................] - ETA: 5:56 - loss: 0.0645 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "195/400 [=============>................] - ETA: 5:54 - loss: 0.0645 - dice_coefficient: 0.9559\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "196/400 [=============>................] - ETA: 5:53 - loss: 0.0645 - dice_coefficient: 0.9559\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "197/400 [=============>................] - ETA: 5:51 - loss: 0.0644 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "198/400 [=============>................] - ETA: 5:49 - loss: 0.0643 - dice_coefficient: 0.9561\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "199/400 [=============>................] - ETA: 5:48 - loss: 0.0644 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "200/400 [==============>...............] - ETA: 5:46 - loss: 0.0645 - dice_coefficient: 0.9559\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "201/400 [==============>...............] - ETA: 5:44 - loss: 0.0645 - dice_coefficient: 0.9559\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "202/400 [==============>...............] - ETA: 5:42 - loss: 0.0643 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "203/400 [==============>...............] - ETA: 5:41 - loss: 0.0644 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "204/400 [==============>...............] - ETA: 5:39 - loss: 0.0645 - dice_coefficient: 0.9559\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "205/400 [==============>...............] - ETA: 5:37 - loss: 0.0648 - dice_coefficient: 0.9559\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "206/400 [==============>...............] - ETA: 5:36 - loss: 0.0647 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "207/400 [==============>...............] - ETA: 5:34 - loss: 0.0649 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "208/400 [==============>...............] - ETA: 5:32 - loss: 0.0647 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "209/400 [==============>...............] - ETA: 5:30 - loss: 0.1418 - dice_coefficient: 0.9514\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "210/400 [==============>...............] - ETA: 5:29 - loss: 0.1413 - dice_coefficient: 0.9515\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "211/400 [==============>...............] - ETA: 5:27 - loss: 0.1409 - dice_coefficient: 0.9516\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "212/400 [==============>...............] - ETA: 5:25 - loss: 0.1404 - dice_coefficient: 0.9516\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "213/400 [==============>...............] - ETA: 5:24 - loss: 0.1401 - dice_coefficient: 0.9517\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "214/400 [===============>..............] - ETA: 5:22 - loss: 0.1399 - dice_coefficient: 0.9517\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "215/400 [===============>..............] - ETA: 5:20 - loss: 0.1395 - dice_coefficient: 0.9518\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "216/400 [===============>..............] - ETA: 5:18 - loss: 0.1390 - dice_coefficient: 0.9519\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "217/400 [===============>..............] - ETA: 5:17 - loss: 0.1386 - dice_coefficient: 0.9519\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "218/400 [===============>..............] - ETA: 5:15 - loss: 0.1381 - dice_coefficient: 0.9520\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "219/400 [===============>..............] - ETA: 5:13 - loss: 0.1378 - dice_coefficient: 0.9521\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "220/400 [===============>..............] - ETA: 5:12 - loss: 0.1373 - dice_coefficient: 0.9522\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "221/400 [===============>..............] - ETA: 5:10 - loss: 0.1372 - dice_coefficient: 0.9521\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "222/400 [===============>..............] - ETA: 5:08 - loss: 0.1368 - dice_coefficient: 0.9522\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "223/400 [===============>..............] - ETA: 5:06 - loss: 0.1363 - dice_coefficient: 0.9523\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "224/400 [===============>..............] - ETA: 5:05 - loss: 0.1359 - dice_coefficient: 0.9524\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "225/400 [===============>..............] - ETA: 5:03 - loss: 0.1357 - dice_coefficient: 0.9524\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "226/400 [===============>..............] - ETA: 5:01 - loss: 0.1353 - dice_coefficient: 0.9525\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "227/400 [================>.............] - ETA: 5:00 - loss: 0.1348 - dice_coefficient: 0.9526\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "228/400 [================>.............] - ETA: 4:58 - loss: 0.1344 - dice_coefficient: 0.9527\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "229/400 [================>.............] - ETA: 4:56 - loss: 0.1340 - dice_coefficient: 0.9528\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "230/400 [================>.............] - ETA: 4:54 - loss: 0.1336 - dice_coefficient: 0.9528\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "231/400 [================>.............] - ETA: 4:53 - loss: 0.1333 - dice_coefficient: 0.9528\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "232/400 [================>.............] - ETA: 4:51 - loss: 0.1328 - dice_coefficient: 0.9529\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "233/400 [================>.............] - ETA: 4:49 - loss: 0.1324 - dice_coefficient: 0.9530\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "234/400 [================>.............] - ETA: 4:48 - loss: 0.1320 - dice_coefficient: 0.9531\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "235/400 [================>.............] - ETA: 4:46 - loss: 0.1317 - dice_coefficient: 0.9531\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "236/400 [================>.............] - ETA: 4:44 - loss: 0.1315 - dice_coefficient: 0.9531\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "237/400 [================>.............] - ETA: 4:42 - loss: 0.1312 - dice_coefficient: 0.9531\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "238/400 [================>.............] - ETA: 4:41 - loss: 0.1308 - dice_coefficient: 0.9532\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "239/400 [================>.............] - ETA: 4:39 - loss: 0.1304 - dice_coefficient: 0.9533\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "240/400 [=================>............] - ETA: 4:37 - loss: 0.1302 - dice_coefficient: 0.9533\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "241/400 [=================>............] - ETA: 4:35 - loss: 0.1297 - dice_coefficient: 0.9535\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "242/400 [=================>............] - ETA: 4:34 - loss: 0.1295 - dice_coefficient: 0.9535\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "243/400 [=================>............] - ETA: 4:32 - loss: 0.1293 - dice_coefficient: 0.9534\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "244/400 [=================>............] - ETA: 4:30 - loss: 0.1292 - dice_coefficient: 0.9534\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "245/400 [=================>............] - ETA: 4:29 - loss: 0.1289 - dice_coefficient: 0.9534\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "246/400 [=================>............] - ETA: 4:27 - loss: 0.1285 - dice_coefficient: 0.9535\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "247/400 [=================>............] - ETA: 4:25 - loss: 0.1281 - dice_coefficient: 0.9536\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "248/400 [=================>............] - ETA: 4:23 - loss: 0.1277 - dice_coefficient: 0.9537\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "249/400 [=================>............] - ETA: 4:22 - loss: 0.1273 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "250/400 [=================>............] - ETA: 4:20 - loss: 0.1269 - dice_coefficient: 0.9539\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "251/400 [=================>............] - ETA: 4:18 - loss: 0.1268 - dice_coefficient: 0.9537\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "252/400 [=================>............] - ETA: 4:17 - loss: 0.1266 - dice_coefficient: 0.9537\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "253/400 [=================>............] - ETA: 4:15 - loss: 0.1262 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "254/400 [==================>...........] - ETA: 4:13 - loss: 0.1260 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "255/400 [==================>...........] - ETA: 4:11 - loss: 0.1257 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "256/400 [==================>...........] - ETA: 4:10 - loss: 0.1254 - dice_coefficient: 0.9539\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "257/400 [==================>...........] - ETA: 4:08 - loss: 0.1251 - dice_coefficient: 0.9539\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "258/400 [==================>...........] - ETA: 4:06 - loss: 0.1248 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "259/400 [==================>...........] - ETA: 4:05 - loss: 0.1245 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "260/400 [==================>...........] - ETA: 4:03 - loss: 0.1242 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "261/400 [==================>...........] - ETA: 4:01 - loss: 0.1239 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "262/400 [==================>...........] - ETA: 3:59 - loss: 0.1237 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "263/400 [==================>...........] - ETA: 3:58 - loss: 0.1234 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "264/400 [==================>...........] - ETA: 3:56 - loss: 0.1230 - dice_coefficient: 0.9543\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "265/400 [==================>...........] - ETA: 3:54 - loss: 0.1228 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "266/400 [==================>...........] - ETA: 3:52 - loss: 0.1226 - dice_coefficient: 0.9543\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "267/400 [===================>..........] - ETA: 3:51 - loss: 0.1223 - dice_coefficient: 0.9543\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "268/400 [===================>..........] - ETA: 3:49 - loss: 0.1220 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "269/400 [===================>..........] - ETA: 3:47 - loss: 0.1218 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "270/400 [===================>..........] - ETA: 3:46 - loss: 0.1217 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "271/400 [===================>..........] - ETA: 3:44 - loss: 0.1214 - dice_coefficient: 0.9543\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "272/400 [===================>..........] - ETA: 3:42 - loss: 0.1212 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "273/400 [===================>..........] - ETA: 3:40 - loss: 0.1211 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "274/400 [===================>..........] - ETA: 3:39 - loss: 0.1209 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "275/400 [===================>..........] - ETA: 3:37 - loss: 0.1208 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "276/400 [===================>..........] - ETA: 3:35 - loss: 0.1207 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "277/400 [===================>..........] - ETA: 3:34 - loss: 0.1203 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "278/400 [===================>..........] - ETA: 3:32 - loss: 0.1200 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "279/400 [===================>..........] - ETA: 3:30 - loss: 0.1198 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "280/400 [====================>.........] - ETA: 3:28 - loss: 0.1197 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "281/400 [====================>.........] - ETA: 3:27 - loss: 0.1195 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "282/400 [====================>.........] - ETA: 3:25 - loss: 0.1195 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "283/400 [====================>.........] - ETA: 3:23 - loss: 0.1193 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "284/400 [====================>.........] - ETA: 3:21 - loss: 0.1192 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "285/400 [====================>.........] - ETA: 3:20 - loss: 0.1189 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "286/400 [====================>.........] - ETA: 3:18 - loss: 0.1187 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "287/400 [====================>.........] - ETA: 3:16 - loss: 0.1185 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "288/400 [====================>.........] - ETA: 3:15 - loss: 0.1182 - dice_coefficient: 0.9543\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "289/400 [====================>.........] - ETA: 3:13 - loss: 0.1186 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "290/400 [====================>.........] - ETA: 3:11 - loss: 0.1184 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "291/400 [====================>.........] - ETA: 3:09 - loss: 0.1182 - dice_coefficient: 0.9539\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "292/400 [====================>.........] - ETA: 3:08 - loss: 0.1180 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "293/400 [====================>.........] - ETA: 3:06 - loss: 0.1178 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "294/400 [=====================>........] - ETA: 3:04 - loss: 0.1177 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "295/400 [=====================>........] - ETA: 3:02 - loss: 0.1177 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "296/400 [=====================>........] - ETA: 3:01 - loss: 0.1175 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "297/400 [=====================>........] - ETA: 2:59 - loss: 0.1172 - dice_coefficient: 0.9539\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "298/400 [=====================>........] - ETA: 2:57 - loss: 0.1169 - dice_coefficient: 0.9539\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "299/400 [=====================>........] - ETA: 2:55 - loss: 0.1169 - dice_coefficient: 0.9537\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "300/400 [=====================>........] - ETA: 2:54 - loss: 0.1169 - dice_coefficient: 0.9536\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "301/400 [=====================>........] - ETA: 2:52 - loss: 0.1167 - dice_coefficient: 0.9536\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "302/400 [=====================>........] - ETA: 2:50 - loss: 0.1165 - dice_coefficient: 0.9537\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "303/400 [=====================>........] - ETA: 2:49 - loss: 0.1163 - dice_coefficient: 0.9537\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "304/400 [=====================>........] - ETA: 2:47 - loss: 0.1162 - dice_coefficient: 0.9537\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "305/400 [=====================>........] - ETA: 2:45 - loss: 0.1161 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "306/400 [=====================>........] - ETA: 2:43 - loss: 0.1159 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "307/400 [======================>.......] - ETA: 2:42 - loss: 0.1157 - dice_coefficient: 0.9537\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "308/400 [======================>.......] - ETA: 2:40 - loss: 0.1154 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "309/400 [======================>.......] - ETA: 2:38 - loss: 0.1153 - dice_coefficient: 0.9538\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "310/400 [======================>.......] - ETA: 2:36 - loss: 0.1150 - dice_coefficient: 0.9539\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "311/400 [======================>.......] - ETA: 2:35 - loss: 0.1147 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "312/400 [======================>.......] - ETA: 2:33 - loss: 0.1146 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "313/400 [======================>.......] - ETA: 2:31 - loss: 0.1144 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "314/400 [======================>.......] - ETA: 2:29 - loss: 0.1142 - dice_coefficient: 0.9540\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "315/400 [======================>.......] - ETA: 2:28 - loss: 0.1141 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "316/400 [======================>.......] - ETA: 2:26 - loss: 0.1138 - dice_coefficient: 0.9541\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "317/400 [======================>.......] - ETA: 2:24 - loss: 0.1136 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "318/400 [======================>.......] - ETA: 2:22 - loss: 0.1133 - dice_coefficient: 0.9542\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "319/400 [======================>.......] - ETA: 2:21 - loss: 0.1131 - dice_coefficient: 0.9543\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "320/400 [=======================>......] - ETA: 2:19 - loss: 0.1130 - dice_coefficient: 0.9543\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "321/400 [=======================>......] - ETA: 2:17 - loss: 0.1128 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "322/400 [=======================>......] - ETA: 2:15 - loss: 0.1127 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "323/400 [=======================>......] - ETA: 2:14 - loss: 0.1126 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "324/400 [=======================>......] - ETA: 2:12 - loss: 0.1124 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "325/400 [=======================>......] - ETA: 2:10 - loss: 0.1123 - dice_coefficient: 0.9545\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "326/400 [=======================>......] - ETA: 2:09 - loss: 0.1121 - dice_coefficient: 0.9545\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "327/400 [=======================>......] - ETA: 2:07 - loss: 0.1119 - dice_coefficient: 0.9546\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "328/400 [=======================>......] - ETA: 2:05 - loss: 0.1117 - dice_coefficient: 0.9546\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "329/400 [=======================>......] - ETA: 2:03 - loss: 0.1117 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "330/400 [=======================>......] - ETA: 2:02 - loss: 0.1115 - dice_coefficient: 0.9544\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "331/400 [=======================>......] - ETA: 2:00 - loss: 0.1113 - dice_coefficient: 0.9545\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "332/400 [=======================>......] - ETA: 1:58 - loss: 0.1112 - dice_coefficient: 0.9545\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "333/400 [=======================>......] - ETA: 1:56 - loss: 0.1110 - dice_coefficient: 0.9546\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "334/400 [========================>.....] - ETA: 1:55 - loss: 0.1108 - dice_coefficient: 0.9546\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "335/400 [========================>.....] - ETA: 1:53 - loss: 0.1108 - dice_coefficient: 0.9545\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "336/400 [========================>.....] - ETA: 1:51 - loss: 0.1106 - dice_coefficient: 0.9545\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "337/400 [========================>.....] - ETA: 1:49 - loss: 0.1104 - dice_coefficient: 0.9546\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "338/400 [========================>.....] - ETA: 1:48 - loss: 0.1103 - dice_coefficient: 0.9546\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "339/400 [========================>.....] - ETA: 1:46 - loss: 0.1101 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "340/400 [========================>.....] - ETA: 1:44 - loss: 0.1099 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "341/400 [========================>.....] - ETA: 1:42 - loss: 0.1097 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "342/400 [========================>.....] - ETA: 1:41 - loss: 0.1095 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "343/400 [========================>.....] - ETA: 1:39 - loss: 0.1097 - dice_coefficient: 0.9545\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "344/400 [========================>.....] - ETA: 1:37 - loss: 0.1095 - dice_coefficient: 0.9546\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "345/400 [========================>.....] - ETA: 1:35 - loss: 0.1093 - dice_coefficient: 0.9546\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "346/400 [========================>.....] - ETA: 1:34 - loss: 0.1091 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "347/400 [=========================>....] - ETA: 1:32 - loss: 0.1090 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "348/400 [=========================>....] - ETA: 1:30 - loss: 0.1088 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "349/400 [=========================>....] - ETA: 1:29 - loss: 0.1087 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "350/400 [=========================>....] - ETA: 1:27 - loss: 0.1086 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "351/400 [=========================>....] - ETA: 1:25 - loss: 0.1085 - dice_coefficient: 0.9547\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "352/400 [=========================>....] - ETA: 1:23 - loss: 0.1083 - dice_coefficient: 0.9548\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "353/400 [=========================>....] - ETA: 1:22 - loss: 0.1081 - dice_coefficient: 0.9549\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "354/400 [=========================>....] - ETA: 1:20 - loss: 0.1079 - dice_coefficient: 0.9549\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "355/400 [=========================>....] - ETA: 1:18 - loss: 0.1078 - dice_coefficient: 0.9549\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "356/400 [=========================>....] - ETA: 1:16 - loss: 0.1076 - dice_coefficient: 0.9550\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "357/400 [=========================>....] - ETA: 1:15 - loss: 0.1074 - dice_coefficient: 0.9550\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "358/400 [=========================>....] - ETA: 1:13 - loss: 0.1072 - dice_coefficient: 0.9550\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "359/400 [=========================>....] - ETA: 1:11 - loss: 0.1070 - dice_coefficient: 0.9551\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "360/400 [==========================>...] - ETA: 1:09 - loss: 0.1068 - dice_coefficient: 0.9551\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "361/400 [==========================>...] - ETA: 1:08 - loss: 0.1066 - dice_coefficient: 0.9552\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "362/400 [==========================>...] - ETA: 1:06 - loss: 0.1064 - dice_coefficient: 0.9553\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "363/400 [==========================>...] - ETA: 1:04 - loss: 0.1063 - dice_coefficient: 0.9553\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "364/400 [==========================>...] - ETA: 1:02 - loss: 0.1061 - dice_coefficient: 0.9554\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "365/400 [==========================>...] - ETA: 1:01 - loss: 0.1059 - dice_coefficient: 0.9554\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "366/400 [==========================>...] - ETA: 59s - loss: 0.1058 - dice_coefficient: 0.9555 \n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "367/400 [==========================>...] - ETA: 57s - loss: 0.1057 - dice_coefficient: 0.9555\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "368/400 [==========================>...] - ETA: 55s - loss: 0.1055 - dice_coefficient: 0.9555\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "369/400 [==========================>...] - ETA: 54s - loss: 0.1054 - dice_coefficient: 0.9555\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "370/400 [==========================>...] - ETA: 52s - loss: 0.1051 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "371/400 [==========================>...] - ETA: 50s - loss: 0.1050 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "372/400 [==========================>...] - ETA: 48s - loss: 0.1051 - dice_coefficient: 0.9555\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "373/400 [==========================>...] - ETA: 47s - loss: 0.1050 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "374/400 [===========================>..] - ETA: 45s - loss: 0.1048 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "375/400 [===========================>..] - ETA: 43s - loss: 0.1046 - dice_coefficient: 0.9557\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "376/400 [===========================>..] - ETA: 41s - loss: 0.1045 - dice_coefficient: 0.9557\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "377/400 [===========================>..] - ETA: 40s - loss: 0.1042 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "378/400 [===========================>..] - ETA: 38s - loss: 0.1042 - dice_coefficient: 0.9557\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "379/400 [===========================>..] - ETA: 36s - loss: 0.1040 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "380/400 [===========================>..] - ETA: 34s - loss: 0.1038 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "381/400 [===========================>..] - ETA: 33s - loss: 0.1040 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "382/400 [===========================>..] - ETA: 31s - loss: 0.1039 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "383/400 [===========================>..] - ETA: 29s - loss: 0.1037 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "384/400 [===========================>..] - ETA: 27s - loss: 0.1037 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "385/400 [===========================>..] - ETA: 26s - loss: 0.1036 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "386/400 [===========================>..] - ETA: 24s - loss: 0.1034 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "387/400 [============================>.] - ETA: 22s - loss: 0.1033 - dice_coefficient: 0.9556\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "388/400 [============================>.] - ETA: 20s - loss: 0.1031 - dice_coefficient: 0.9557\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "389/400 [============================>.] - ETA: 19s - loss: 0.1029 - dice_coefficient: 0.9557\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "390/400 [============================>.] - ETA: 17s - loss: 0.1028 - dice_coefficient: 0.9557\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "391/400 [============================>.] - ETA: 15s - loss: 0.1027 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "392/400 [============================>.] - ETA: 13s - loss: 0.1027 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "393/400 [============================>.] - ETA: 12s - loss: 0.1025 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "394/400 [============================>.] - ETA: 10s - loss: 0.1023 - dice_coefficient: 0.9559\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "395/400 [============================>.] - ETA: 8s - loss: 0.1021 - dice_coefficient: 0.9559 \n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "396/400 [============================>.] - ETA: 6s - loss: 0.1020 - dice_coefficient: 0.9559\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "397/400 [============================>.] - ETA: 5s - loss: 0.1019 - dice_coefficient: 0.9560\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "398/400 [============================>.] - ETA: 3s - loss: 0.1020 - dice_coefficient: 0.9558\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "399/400 [============================>.] - ETA: 1s - loss: 0.1019 - dice_coefficient: 0.9559\n",
      "Epoch 00028: loss did not improve from 0.04222\n",
      "400/400 [==============================] - 704s 2s/step - loss: 0.1017 - dice_coefficient: 0.9559 - val_loss: 1.0913 - val_dice_coefficient: 0.5416\n",
      "Epoch 29/30\n",
      "\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "  1/400 [..............................] - ETA: 0s - loss: 0.0961 - dice_coefficient: 0.9469\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "  2/400 [..............................] - ETA: 5:59 - loss: 0.1282 - dice_coefficient: 0.9054\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "  3/400 [..............................] - ETA: 7:52 - loss: 0.1007 - dice_coefficient: 0.9242\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "  4/400 [..............................] - ETA: 8:46 - loss: 0.0967 - dice_coefficient: 0.9262\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "  5/400 [..............................] - ETA: 9:18 - loss: 3.3127 - dice_coefficient: 0.7409\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "  6/400 [..............................] - ETA: 9:40 - loss: 2.7651 - dice_coefficient: 0.7812\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "  7/400 [..............................] - ETA: 9:55 - loss: 2.3835 - dice_coefficient: 0.8023\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "  8/400 [..............................] - ETA: 10:06 - loss: 2.0978 - dice_coefficient: 0.8220\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "  9/400 [..............................] - ETA: 10:14 - loss: 1.8771 - dice_coefficient: 0.8312\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 10/400 [..............................] - ETA: 10:19 - loss: 1.6928 - dice_coefficient: 0.8455\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 11/400 [..............................] - ETA: 10:24 - loss: 1.5479 - dice_coefficient: 0.8525\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 12/400 [..............................] - ETA: 10:28 - loss: 1.4228 - dice_coefficient: 0.8633\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 13/400 [..............................] - ETA: 10:31 - loss: 1.3185 - dice_coefficient: 0.8695\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 14/400 [>.............................] - ETA: 10:32 - loss: 1.2253 - dice_coefficient: 0.8782\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 15/400 [>.............................] - ETA: 10:34 - loss: 1.1496 - dice_coefficient: 0.8824\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 16/400 [>.............................] - ETA: 10:35 - loss: 1.0818 - dice_coefficient: 0.8877\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 17/400 [>.............................] - ETA: 10:36 - loss: 1.0220 - dice_coefficient: 0.8910\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 18/400 [>.............................] - ETA: 10:37 - loss: 0.9691 - dice_coefficient: 0.8952\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 19/400 [>.............................] - ETA: 10:37 - loss: 0.9201 - dice_coefficient: 0.8997\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 20/400 [>.............................] - ETA: 10:37 - loss: 0.8774 - dice_coefficient: 0.9030\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 21/400 [>.............................] - ETA: 10:37 - loss: 0.8384 - dice_coefficient: 0.9053\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 22/400 [>.............................] - ETA: 10:37 - loss: 0.8024 - dice_coefficient: 0.9084\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 23/400 [>.............................] - ETA: 10:36 - loss: 0.7701 - dice_coefficient: 0.9113\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 24/400 [>.............................] - ETA: 10:36 - loss: 0.7387 - dice_coefficient: 0.9145\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 25/400 [>.............................] - ETA: 10:35 - loss: 0.7180 - dice_coefficient: 0.9103\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 26/400 [>.............................] - ETA: 10:35 - loss: 0.6936 - dice_coefficient: 0.9116\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 27/400 [=>............................] - ETA: 10:34 - loss: 0.6694 - dice_coefficient: 0.9139\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 28/400 [=>............................] - ETA: 10:33 - loss: 0.6478 - dice_coefficient: 0.9152\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 29/400 [=>............................] - ETA: 10:32 - loss: 0.6264 - dice_coefficient: 0.9177\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 30/400 [=>............................] - ETA: 10:31 - loss: 0.6080 - dice_coefficient: 0.9188\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 31/400 [=>............................] - ETA: 10:30 - loss: 0.5899 - dice_coefficient: 0.9202\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 32/400 [=>............................] - ETA: 10:29 - loss: 0.5723 - dice_coefficient: 0.9219\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 33/400 [=>............................] - ETA: 10:28 - loss: 0.5564 - dice_coefficient: 0.9233\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 34/400 [=>............................] - ETA: 10:27 - loss: 0.5426 - dice_coefficient: 0.9239\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 35/400 [=>............................] - ETA: 10:25 - loss: 0.5282 - dice_coefficient: 0.9255\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 36/400 [=>............................] - ETA: 10:24 - loss: 0.5152 - dice_coefficient: 0.9268\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 37/400 [=>............................] - ETA: 10:23 - loss: 0.5021 - dice_coefficient: 0.9282\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 38/400 [=>............................] - ETA: 10:22 - loss: 0.4920 - dice_coefficient: 0.9285\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 39/400 [=>............................] - ETA: 10:21 - loss: 0.4808 - dice_coefficient: 0.9293\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 40/400 [==>...........................] - ETA: 10:20 - loss: 0.4698 - dice_coefficient: 0.9302\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 41/400 [==>...........................] - ETA: 10:18 - loss: 0.4600 - dice_coefficient: 0.9306\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 42/400 [==>...........................] - ETA: 10:17 - loss: 0.4504 - dice_coefficient: 0.9312\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 43/400 [==>...........................] - ETA: 10:16 - loss: 0.4407 - dice_coefficient: 0.9321\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 44/400 [==>...........................] - ETA: 10:14 - loss: 0.4312 - dice_coefficient: 0.9332\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 45/400 [==>...........................] - ETA: 10:13 - loss: 0.4243 - dice_coefficient: 0.9332\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 46/400 [==>...........................] - ETA: 10:11 - loss: 0.4156 - dice_coefficient: 0.9343\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 47/400 [==>...........................] - ETA: 10:10 - loss: 0.4088 - dice_coefficient: 0.9342\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 48/400 [==>...........................] - ETA: 10:08 - loss: 0.4016 - dice_coefficient: 0.9347\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 49/400 [==>...........................] - ETA: 10:07 - loss: 0.3946 - dice_coefficient: 0.9350\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 50/400 [==>...........................] - ETA: 10:05 - loss: 0.3871 - dice_coefficient: 0.9360\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 51/400 [==>...........................] - ETA: 10:04 - loss: 0.3802 - dice_coefficient: 0.9367\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 52/400 [==>...........................] - ETA: 10:02 - loss: 0.3742 - dice_coefficient: 0.9373\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 53/400 [==>...........................] - ETA: 10:01 - loss: 0.3687 - dice_coefficient: 0.9374\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 54/400 [===>..........................] - ETA: 9:59 - loss: 0.3630 - dice_coefficient: 0.9377 \n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 55/400 [===>..........................] - ETA: 9:58 - loss: 0.3581 - dice_coefficient: 0.9378\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 56/400 [===>..........................] - ETA: 9:57 - loss: 0.3531 - dice_coefficient: 0.9380\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 57/400 [===>..........................] - ETA: 9:55 - loss: 0.3484 - dice_coefficient: 0.9380\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 58/400 [===>..........................] - ETA: 9:54 - loss: 0.3434 - dice_coefficient: 0.9385\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 59/400 [===>..........................] - ETA: 9:52 - loss: 0.3386 - dice_coefficient: 0.9389\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 60/400 [===>..........................] - ETA: 9:51 - loss: 0.3342 - dice_coefficient: 0.9390\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 61/400 [===>..........................] - ETA: 9:50 - loss: 0.3302 - dice_coefficient: 0.9392\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 62/400 [===>..........................] - ETA: 9:48 - loss: 0.3254 - dice_coefficient: 0.9398\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 63/400 [===>..........................] - ETA: 9:46 - loss: 0.3213 - dice_coefficient: 0.9400\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 64/400 [===>..........................] - ETA: 9:45 - loss: 0.3175 - dice_coefficient: 0.9399\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 65/400 [===>..........................] - ETA: 9:43 - loss: 0.3135 - dice_coefficient: 0.9404\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 66/400 [===>..........................] - ETA: 9:41 - loss: 0.3098 - dice_coefficient: 0.9403\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 67/400 [====>.........................] - ETA: 9:40 - loss: 0.3062 - dice_coefficient: 0.9407\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 68/400 [====>.........................] - ETA: 9:38 - loss: 0.3022 - dice_coefficient: 0.9412\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 69/400 [====>.........................] - ETA: 9:36 - loss: 0.2989 - dice_coefficient: 0.9411\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 70/400 [====>.........................] - ETA: 9:35 - loss: 0.2951 - dice_coefficient: 0.9416\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 71/400 [====>.........................] - ETA: 9:33 - loss: 0.2927 - dice_coefficient: 0.9411\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 72/400 [====>.........................] - ETA: 9:31 - loss: 0.2895 - dice_coefficient: 0.9415\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 73/400 [====>.........................] - ETA: 9:30 - loss: 0.2862 - dice_coefficient: 0.9420\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 74/400 [====>.........................] - ETA: 9:28 - loss: 0.2829 - dice_coefficient: 0.9425\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 75/400 [====>.........................] - ETA: 9:26 - loss: 0.2804 - dice_coefficient: 0.9420\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 76/400 [====>.........................] - ETA: 9:25 - loss: 0.2783 - dice_coefficient: 0.9418\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 77/400 [====>.........................] - ETA: 9:23 - loss: 0.2751 - dice_coefficient: 0.9423\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 78/400 [====>.........................] - ETA: 9:22 - loss: 0.2726 - dice_coefficient: 0.9425\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 79/400 [====>.........................] - ETA: 9:20 - loss: 0.2699 - dice_coefficient: 0.9427\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 80/400 [=====>........................] - ETA: 9:18 - loss: 0.2675 - dice_coefficient: 0.9426\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 81/400 [=====>........................] - ETA: 9:17 - loss: 0.2652 - dice_coefficient: 0.9426\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 82/400 [=====>........................] - ETA: 9:15 - loss: 0.2628 - dice_coefficient: 0.9426\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 83/400 [=====>........................] - ETA: 9:13 - loss: 0.2604 - dice_coefficient: 0.9428\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 84/400 [=====>........................] - ETA: 9:11 - loss: 0.2579 - dice_coefficient: 0.9431\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 85/400 [=====>........................] - ETA: 9:10 - loss: 0.2560 - dice_coefficient: 0.9432\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 86/400 [=====>........................] - ETA: 9:08 - loss: 0.2534 - dice_coefficient: 0.9436\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 87/400 [=====>........................] - ETA: 9:06 - loss: 0.2514 - dice_coefficient: 0.9434\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 88/400 [=====>........................] - ETA: 9:05 - loss: 0.2493 - dice_coefficient: 0.9437\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 89/400 [=====>........................] - ETA: 9:03 - loss: 0.2471 - dice_coefficient: 0.9439\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 90/400 [=====>........................] - ETA: 9:01 - loss: 0.2452 - dice_coefficient: 0.9441\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 91/400 [=====>........................] - ETA: 9:00 - loss: 0.2433 - dice_coefficient: 0.9441\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 92/400 [=====>........................] - ETA: 8:58 - loss: 0.2417 - dice_coefficient: 0.9437\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 93/400 [=====>........................] - ETA: 8:56 - loss: 0.2399 - dice_coefficient: 0.9436\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 94/400 [======>.......................] - ETA: 8:55 - loss: 0.2376 - dice_coefficient: 0.9441\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 95/400 [======>.......................] - ETA: 8:53 - loss: 0.2356 - dice_coefficient: 0.9442\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 96/400 [======>.......................] - ETA: 8:51 - loss: 0.2335 - dice_coefficient: 0.9446\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 97/400 [======>.......................] - ETA: 8:50 - loss: 0.2316 - dice_coefficient: 0.9449\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 98/400 [======>.......................] - ETA: 8:48 - loss: 0.2297 - dice_coefficient: 0.9451\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      " 99/400 [======>.......................] - ETA: 8:46 - loss: 0.2281 - dice_coefficient: 0.9451\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "100/400 [======>.......................] - ETA: 8:44 - loss: 0.2263 - dice_coefficient: 0.9452\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "101/400 [======>.......................] - ETA: 8:42 - loss: 0.2246 - dice_coefficient: 0.9453\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "102/400 [======>.......................] - ETA: 8:41 - loss: 0.2229 - dice_coefficient: 0.9456\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "103/400 [======>.......................] - ETA: 8:39 - loss: 0.2215 - dice_coefficient: 0.9458\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "104/400 [======>.......................] - ETA: 8:37 - loss: 0.2200 - dice_coefficient: 0.9461\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "105/400 [======>.......................] - ETA: 8:36 - loss: 0.2181 - dice_coefficient: 0.9465\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "106/400 [======>.......................] - ETA: 8:34 - loss: 0.2167 - dice_coefficient: 0.9466\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "107/400 [=======>......................] - ETA: 8:32 - loss: 0.2149 - dice_coefficient: 0.9468\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "108/400 [=======>......................] - ETA: 8:31 - loss: 0.2131 - dice_coefficient: 0.9472\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "109/400 [=======>......................] - ETA: 8:29 - loss: 0.2114 - dice_coefficient: 0.9476\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "110/400 [=======>......................] - ETA: 8:27 - loss: 0.2099 - dice_coefficient: 0.9479\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "111/400 [=======>......................] - ETA: 8:25 - loss: 0.2083 - dice_coefficient: 0.9482\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "112/400 [=======>......................] - ETA: 8:24 - loss: 0.2068 - dice_coefficient: 0.9485\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "113/400 [=======>......................] - ETA: 8:22 - loss: 0.2055 - dice_coefficient: 0.9487\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "114/400 [=======>......................] - ETA: 8:20 - loss: 0.2042 - dice_coefficient: 0.9488\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "115/400 [=======>......................] - ETA: 8:18 - loss: 0.2031 - dice_coefficient: 0.9488\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "116/400 [=======>......................] - ETA: 8:17 - loss: 0.2021 - dice_coefficient: 0.9486\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "117/400 [=======>......................] - ETA: 8:15 - loss: 0.2008 - dice_coefficient: 0.9489\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "118/400 [=======>......................] - ETA: 8:13 - loss: 0.1998 - dice_coefficient: 0.9490\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "119/400 [=======>......................] - ETA: 8:11 - loss: 0.1986 - dice_coefficient: 0.9492\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "120/400 [========>.....................] - ETA: 8:10 - loss: 0.1974 - dice_coefficient: 0.9494\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "121/400 [========>.....................] - ETA: 8:08 - loss: 0.1960 - dice_coefficient: 0.9496\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "122/400 [========>.....................] - ETA: 8:07 - loss: 0.1947 - dice_coefficient: 0.9499\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "123/400 [========>.....................] - ETA: 8:05 - loss: 0.1933 - dice_coefficient: 0.9501\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "124/400 [========>.....................] - ETA: 8:03 - loss: 0.1924 - dice_coefficient: 0.9499\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "125/400 [========>.....................] - ETA: 8:01 - loss: 0.1923 - dice_coefficient: 0.9491\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "126/400 [========>.....................] - ETA: 8:00 - loss: 0.1910 - dice_coefficient: 0.9494\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "127/400 [========>.....................] - ETA: 7:58 - loss: 0.1901 - dice_coefficient: 0.9493\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "128/400 [========>.....................] - ETA: 7:56 - loss: 0.1891 - dice_coefficient: 0.9495\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "129/400 [========>.....................] - ETA: 7:55 - loss: 0.1880 - dice_coefficient: 0.9496\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "130/400 [========>.....................] - ETA: 7:53 - loss: 0.1871 - dice_coefficient: 0.9497\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "131/400 [========>.....................] - ETA: 7:51 - loss: 0.1864 - dice_coefficient: 0.9495\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "132/400 [========>.....................] - ETA: 7:49 - loss: 0.1856 - dice_coefficient: 0.9495\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "133/400 [========>.....................] - ETA: 7:48 - loss: 0.1843 - dice_coefficient: 0.9498\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "134/400 [=========>....................] - ETA: 7:46 - loss: 0.1833 - dice_coefficient: 0.9499\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "135/400 [=========>....................] - ETA: 7:44 - loss: 0.1823 - dice_coefficient: 0.9500\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "136/400 [=========>....................] - ETA: 7:42 - loss: 0.1813 - dice_coefficient: 0.9502\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "137/400 [=========>....................] - ETA: 7:41 - loss: 0.1804 - dice_coefficient: 0.9504\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "138/400 [=========>....................] - ETA: 7:39 - loss: 0.1793 - dice_coefficient: 0.9505\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "139/400 [=========>....................] - ETA: 7:37 - loss: 0.1784 - dice_coefficient: 0.9507\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "140/400 [=========>....................] - ETA: 7:36 - loss: 0.1775 - dice_coefficient: 0.9507\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "141/400 [=========>....................] - ETA: 7:34 - loss: 0.1766 - dice_coefficient: 0.9509\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "142/400 [=========>....................] - ETA: 7:32 - loss: 0.1757 - dice_coefficient: 0.9510\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "143/400 [=========>....................] - ETA: 7:30 - loss: 0.1748 - dice_coefficient: 0.9512\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "144/400 [=========>....................] - ETA: 7:29 - loss: 0.1746 - dice_coefficient: 0.9507\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "145/400 [=========>....................] - ETA: 7:27 - loss: 0.1739 - dice_coefficient: 0.9507\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "146/400 [=========>....................] - ETA: 7:25 - loss: 0.1734 - dice_coefficient: 0.9507\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "147/400 [==========>...................] - ETA: 7:23 - loss: 0.1730 - dice_coefficient: 0.9504\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "148/400 [==========>...................] - ETA: 7:22 - loss: 0.1724 - dice_coefficient: 0.9503\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "149/400 [==========>...................] - ETA: 7:20 - loss: 0.1715 - dice_coefficient: 0.9505\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "150/400 [==========>...................] - ETA: 7:18 - loss: 0.1708 - dice_coefficient: 0.9504\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "151/400 [==========>...................] - ETA: 7:16 - loss: 0.1700 - dice_coefficient: 0.9505\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "152/400 [==========>...................] - ETA: 7:14 - loss: 0.1691 - dice_coefficient: 0.9506\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "153/400 [==========>...................] - ETA: 7:13 - loss: 0.1686 - dice_coefficient: 0.9506\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "154/400 [==========>...................] - ETA: 7:11 - loss: 0.1681 - dice_coefficient: 0.9503\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "155/400 [==========>...................] - ETA: 7:09 - loss: 0.1675 - dice_coefficient: 0.9502\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "156/400 [==========>...................] - ETA: 7:08 - loss: 0.1669 - dice_coefficient: 0.9501\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "157/400 [==========>...................] - ETA: 7:06 - loss: 0.1662 - dice_coefficient: 0.9502\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "158/400 [==========>...................] - ETA: 7:04 - loss: 0.1658 - dice_coefficient: 0.9502\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "159/400 [==========>...................] - ETA: 7:02 - loss: 0.1651 - dice_coefficient: 0.9503\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "160/400 [===========>..................] - ETA: 7:00 - loss: 0.1644 - dice_coefficient: 0.9503\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "161/400 [===========>..................] - ETA: 6:59 - loss: 0.1637 - dice_coefficient: 0.9504\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "162/400 [===========>..................] - ETA: 6:57 - loss: 0.1632 - dice_coefficient: 0.9505\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "163/400 [===========>..................] - ETA: 6:55 - loss: 0.1624 - dice_coefficient: 0.9507\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "164/400 [===========>..................] - ETA: 6:53 - loss: 0.1617 - dice_coefficient: 0.9507\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "165/400 [===========>..................] - ETA: 6:52 - loss: 0.1610 - dice_coefficient: 0.9508\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "166/400 [===========>..................] - ETA: 6:50 - loss: 0.1607 - dice_coefficient: 0.9505\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "167/400 [===========>..................] - ETA: 6:48 - loss: 0.1602 - dice_coefficient: 0.9506\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "168/400 [===========>..................] - ETA: 6:46 - loss: 0.1594 - dice_coefficient: 0.9508\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "169/400 [===========>..................] - ETA: 6:45 - loss: 0.1587 - dice_coefficient: 0.9509\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "170/400 [===========>..................] - ETA: 6:43 - loss: 0.1583 - dice_coefficient: 0.9508\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "171/400 [===========>..................] - ETA: 6:41 - loss: 0.1576 - dice_coefficient: 0.9508\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "172/400 [===========>..................] - ETA: 6:39 - loss: 0.1570 - dice_coefficient: 0.9510\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "173/400 [===========>..................] - ETA: 6:37 - loss: 0.1563 - dice_coefficient: 0.9511\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "174/400 [============>.................] - ETA: 6:36 - loss: 0.1557 - dice_coefficient: 0.9512\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "175/400 [============>.................] - ETA: 6:34 - loss: 0.1551 - dice_coefficient: 0.9512\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "176/400 [============>.................] - ETA: 6:32 - loss: 0.1545 - dice_coefficient: 0.9513\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "177/400 [============>.................] - ETA: 6:30 - loss: 0.1539 - dice_coefficient: 0.9514\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "178/400 [============>.................] - ETA: 6:29 - loss: 0.1532 - dice_coefficient: 0.9515\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "179/400 [============>.................] - ETA: 6:27 - loss: 0.1526 - dice_coefficient: 0.9517\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "180/400 [============>.................] - ETA: 6:25 - loss: 0.1520 - dice_coefficient: 0.9518\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "181/400 [============>.................] - ETA: 6:23 - loss: 0.1515 - dice_coefficient: 0.9519\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "182/400 [============>.................] - ETA: 6:22 - loss: 0.1510 - dice_coefficient: 0.9519\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "183/400 [============>.................] - ETA: 6:20 - loss: 0.1504 - dice_coefficient: 0.9521\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "184/400 [============>.................] - ETA: 6:18 - loss: 0.1499 - dice_coefficient: 0.9521\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "185/400 [============>.................] - ETA: 6:16 - loss: 0.1493 - dice_coefficient: 0.9523\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "186/400 [============>.................] - ETA: 6:15 - loss: 0.1486 - dice_coefficient: 0.9524\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "187/400 [=============>................] - ETA: 6:13 - loss: 0.1482 - dice_coefficient: 0.9525\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "188/400 [=============>................] - ETA: 6:11 - loss: 0.1478 - dice_coefficient: 0.9525\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "189/400 [=============>................] - ETA: 6:09 - loss: 0.1473 - dice_coefficient: 0.9526\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "190/400 [=============>................] - ETA: 6:08 - loss: 0.1469 - dice_coefficient: 0.9526\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "191/400 [=============>................] - ETA: 6:06 - loss: 0.1464 - dice_coefficient: 0.9527\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "192/400 [=============>................] - ETA: 6:04 - loss: 0.1461 - dice_coefficient: 0.9527\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "193/400 [=============>................] - ETA: 6:02 - loss: 0.1456 - dice_coefficient: 0.9527\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "194/400 [=============>................] - ETA: 6:01 - loss: 0.1451 - dice_coefficient: 0.9528\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "195/400 [=============>................] - ETA: 5:59 - loss: 0.1447 - dice_coefficient: 0.9527\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "196/400 [=============>................] - ETA: 5:57 - loss: 0.1443 - dice_coefficient: 0.9527\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "197/400 [=============>................] - ETA: 5:55 - loss: 0.1439 - dice_coefficient: 0.9528\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "198/400 [=============>................] - ETA: 5:54 - loss: 0.1432 - dice_coefficient: 0.9530\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "199/400 [=============>................] - ETA: 5:52 - loss: 0.1427 - dice_coefficient: 0.9531\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "200/400 [==============>...............] - ETA: 5:50 - loss: 0.1423 - dice_coefficient: 0.9532\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "201/400 [==============>...............] - ETA: 5:48 - loss: 0.1420 - dice_coefficient: 0.9531\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "202/400 [==============>...............] - ETA: 5:47 - loss: 0.1415 - dice_coefficient: 0.9532\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "203/400 [==============>...............] - ETA: 5:45 - loss: 0.1410 - dice_coefficient: 0.9533\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "204/400 [==============>...............] - ETA: 5:43 - loss: 0.1406 - dice_coefficient: 0.9534\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "205/400 [==============>...............] - ETA: 5:41 - loss: 0.1401 - dice_coefficient: 0.9536\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "206/400 [==============>...............] - ETA: 5:40 - loss: 0.1397 - dice_coefficient: 0.9535\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "207/400 [==============>...............] - ETA: 5:38 - loss: 0.1392 - dice_coefficient: 0.9536\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "208/400 [==============>...............] - ETA: 5:36 - loss: 0.1387 - dice_coefficient: 0.9537\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "209/400 [==============>...............] - ETA: 5:34 - loss: 0.1382 - dice_coefficient: 0.9538\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "210/400 [==============>...............] - ETA: 5:33 - loss: 0.1380 - dice_coefficient: 0.9537\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "211/400 [==============>...............] - ETA: 5:31 - loss: 0.1375 - dice_coefficient: 0.9538\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "212/400 [==============>...............] - ETA: 5:29 - loss: 0.1371 - dice_coefficient: 0.9539\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "213/400 [==============>...............] - ETA: 5:27 - loss: 0.1369 - dice_coefficient: 0.9539\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "214/400 [===============>..............] - ETA: 5:26 - loss: 0.1368 - dice_coefficient: 0.9537\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "215/400 [===============>..............] - ETA: 5:24 - loss: 0.1363 - dice_coefficient: 0.9538\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "216/400 [===============>..............] - ETA: 5:22 - loss: 0.1358 - dice_coefficient: 0.9539\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "217/400 [===============>..............] - ETA: 5:20 - loss: 0.1406 - dice_coefficient: 0.9510\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "218/400 [===============>..............] - ETA: 5:18 - loss: 0.1402 - dice_coefficient: 0.9510\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "219/400 [===============>..............] - ETA: 5:17 - loss: 0.1397 - dice_coefficient: 0.9511\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "220/400 [===============>..............] - ETA: 5:15 - loss: 0.1395 - dice_coefficient: 0.9511\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "221/400 [===============>..............] - ETA: 5:13 - loss: 0.1391 - dice_coefficient: 0.9512\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "222/400 [===============>..............] - ETA: 5:11 - loss: 0.1385 - dice_coefficient: 0.9514\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "223/400 [===============>..............] - ETA: 5:10 - loss: 0.1381 - dice_coefficient: 0.9514\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "224/400 [===============>..............] - ETA: 5:08 - loss: 0.1377 - dice_coefficient: 0.9515\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "225/400 [===============>..............] - ETA: 5:06 - loss: 0.1373 - dice_coefficient: 0.9516\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "226/400 [===============>..............] - ETA: 5:04 - loss: 0.1368 - dice_coefficient: 0.9517\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "227/400 [================>.............] - ETA: 5:03 - loss: 0.1365 - dice_coefficient: 0.9517\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "228/400 [================>.............] - ETA: 5:01 - loss: 0.1362 - dice_coefficient: 0.9517\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "229/400 [================>.............] - ETA: 4:59 - loss: 0.1358 - dice_coefficient: 0.9518\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "230/400 [================>.............] - ETA: 4:57 - loss: 0.1355 - dice_coefficient: 0.9519\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "231/400 [================>.............] - ETA: 4:56 - loss: 0.1353 - dice_coefficient: 0.9517\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "232/400 [================>.............] - ETA: 4:54 - loss: 0.1349 - dice_coefficient: 0.9517\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "233/400 [================>.............] - ETA: 4:52 - loss: 0.1349 - dice_coefficient: 0.9516\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "234/400 [================>.............] - ETA: 4:51 - loss: 0.1347 - dice_coefficient: 0.9515\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "235/400 [================>.............] - ETA: 4:49 - loss: 0.1345 - dice_coefficient: 0.9514\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "236/400 [================>.............] - ETA: 4:47 - loss: 0.1341 - dice_coefficient: 0.9514\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "237/400 [================>.............] - ETA: 4:45 - loss: 0.1337 - dice_coefficient: 0.9516\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "238/400 [================>.............] - ETA: 4:44 - loss: 0.1333 - dice_coefficient: 0.9517\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "239/400 [================>.............] - ETA: 4:42 - loss: 0.1330 - dice_coefficient: 0.9517\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "240/400 [=================>............] - ETA: 4:40 - loss: 0.1326 - dice_coefficient: 0.9518\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "241/400 [=================>............] - ETA: 4:38 - loss: 0.1324 - dice_coefficient: 0.9517\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "242/400 [=================>............] - ETA: 4:37 - loss: 0.1320 - dice_coefficient: 0.9518\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "243/400 [=================>............] - ETA: 4:35 - loss: 0.1316 - dice_coefficient: 0.9519\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "244/400 [=================>............] - ETA: 4:33 - loss: 0.1315 - dice_coefficient: 0.9519\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "245/400 [=================>............] - ETA: 4:31 - loss: 0.1311 - dice_coefficient: 0.9520\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "246/400 [=================>............] - ETA: 4:30 - loss: 0.1310 - dice_coefficient: 0.9519\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "247/400 [=================>............] - ETA: 4:28 - loss: 0.1306 - dice_coefficient: 0.9519\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "248/400 [=================>............] - ETA: 4:26 - loss: 0.1304 - dice_coefficient: 0.9519\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "249/400 [=================>............] - ETA: 4:24 - loss: 0.1302 - dice_coefficient: 0.9519\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "250/400 [=================>............] - ETA: 4:23 - loss: 0.1298 - dice_coefficient: 0.9519\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "251/400 [=================>............] - ETA: 4:21 - loss: 0.1296 - dice_coefficient: 0.9520\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "252/400 [=================>............] - ETA: 4:19 - loss: 0.1291 - dice_coefficient: 0.9521\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "253/400 [=================>............] - ETA: 4:17 - loss: 0.1287 - dice_coefficient: 0.9523\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "254/400 [==================>...........] - ETA: 4:16 - loss: 0.1284 - dice_coefficient: 0.9524\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "255/400 [==================>...........] - ETA: 4:14 - loss: 0.1281 - dice_coefficient: 0.9525\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "256/400 [==================>...........] - ETA: 4:12 - loss: 0.1279 - dice_coefficient: 0.9524\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "257/400 [==================>...........] - ETA: 4:10 - loss: 0.1276 - dice_coefficient: 0.9525\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "258/400 [==================>...........] - ETA: 4:09 - loss: 0.1273 - dice_coefficient: 0.9525\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "259/400 [==================>...........] - ETA: 4:07 - loss: 0.1270 - dice_coefficient: 0.9526\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "260/400 [==================>...........] - ETA: 4:05 - loss: 0.1267 - dice_coefficient: 0.9526\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "261/400 [==================>...........] - ETA: 4:03 - loss: 0.1265 - dice_coefficient: 0.9527\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "262/400 [==================>...........] - ETA: 4:02 - loss: 0.1263 - dice_coefficient: 0.9527\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "263/400 [==================>...........] - ETA: 4:00 - loss: 0.1260 - dice_coefficient: 0.9526\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "264/400 [==================>...........] - ETA: 3:58 - loss: 0.1256 - dice_coefficient: 0.9528\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "265/400 [==================>...........] - ETA: 3:56 - loss: 0.1252 - dice_coefficient: 0.9529\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "266/400 [==================>...........] - ETA: 3:55 - loss: 0.1250 - dice_coefficient: 0.9529\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "267/400 [===================>..........] - ETA: 3:53 - loss: 0.1245 - dice_coefficient: 0.9530\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "268/400 [===================>..........] - ETA: 3:51 - loss: 0.1242 - dice_coefficient: 0.9531\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "269/400 [===================>..........] - ETA: 3:49 - loss: 0.1240 - dice_coefficient: 0.9532\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "270/400 [===================>..........] - ETA: 3:48 - loss: 0.1238 - dice_coefficient: 0.9531\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "271/400 [===================>..........] - ETA: 3:46 - loss: 0.1235 - dice_coefficient: 0.9532\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "272/400 [===================>..........] - ETA: 3:44 - loss: 0.1231 - dice_coefficient: 0.9533\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "273/400 [===================>..........] - ETA: 3:42 - loss: 0.1227 - dice_coefficient: 0.9534\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "274/400 [===================>..........] - ETA: 3:41 - loss: 0.1224 - dice_coefficient: 0.9535\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "275/400 [===================>..........] - ETA: 3:39 - loss: 0.1220 - dice_coefficient: 0.9536\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "276/400 [===================>..........] - ETA: 3:37 - loss: 0.1217 - dice_coefficient: 0.9537\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "277/400 [===================>..........] - ETA: 3:35 - loss: 0.1214 - dice_coefficient: 0.9538\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "278/400 [===================>..........] - ETA: 3:34 - loss: 0.1213 - dice_coefficient: 0.9538\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "279/400 [===================>..........] - ETA: 3:32 - loss: 0.1210 - dice_coefficient: 0.9538\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "280/400 [====================>.........] - ETA: 3:30 - loss: 0.1209 - dice_coefficient: 0.9539\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "281/400 [====================>.........] - ETA: 3:28 - loss: 0.1206 - dice_coefficient: 0.9540\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "282/400 [====================>.........] - ETA: 3:27 - loss: 0.1203 - dice_coefficient: 0.9540\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "283/400 [====================>.........] - ETA: 3:25 - loss: 0.1200 - dice_coefficient: 0.9541\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "284/400 [====================>.........] - ETA: 3:23 - loss: 0.1196 - dice_coefficient: 0.9543\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "285/400 [====================>.........] - ETA: 3:21 - loss: 0.1193 - dice_coefficient: 0.9544\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "286/400 [====================>.........] - ETA: 3:20 - loss: 0.1190 - dice_coefficient: 0.9545\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "287/400 [====================>.........] - ETA: 3:18 - loss: 0.1189 - dice_coefficient: 0.9544\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "288/400 [====================>.........] - ETA: 3:16 - loss: 0.1187 - dice_coefficient: 0.9544\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "289/400 [====================>.........] - ETA: 3:14 - loss: 0.1184 - dice_coefficient: 0.9544\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "290/400 [====================>.........] - ETA: 3:13 - loss: 0.1181 - dice_coefficient: 0.9545\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "291/400 [====================>.........] - ETA: 3:11 - loss: 0.1180 - dice_coefficient: 0.9545\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "292/400 [====================>.........] - ETA: 3:09 - loss: 0.1177 - dice_coefficient: 0.9545\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "293/400 [====================>.........] - ETA: 3:07 - loss: 0.1174 - dice_coefficient: 0.9546\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "294/400 [=====================>........] - ETA: 3:05 - loss: 0.1171 - dice_coefficient: 0.9547\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "295/400 [=====================>........] - ETA: 3:04 - loss: 0.1169 - dice_coefficient: 0.9548\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "296/400 [=====================>........] - ETA: 3:02 - loss: 0.1166 - dice_coefficient: 0.9548\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "297/400 [=====================>........] - ETA: 3:00 - loss: 0.1163 - dice_coefficient: 0.9549\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "298/400 [=====================>........] - ETA: 2:58 - loss: 0.1160 - dice_coefficient: 0.9550\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "299/400 [=====================>........] - ETA: 2:57 - loss: 0.1157 - dice_coefficient: 0.9551\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "300/400 [=====================>........] - ETA: 2:55 - loss: 0.1156 - dice_coefficient: 0.9551\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "301/400 [=====================>........] - ETA: 2:53 - loss: 0.1154 - dice_coefficient: 0.9551\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "302/400 [=====================>........] - ETA: 2:51 - loss: 0.1151 - dice_coefficient: 0.9552\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "303/400 [=====================>........] - ETA: 2:50 - loss: 0.1150 - dice_coefficient: 0.9552\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "304/400 [=====================>........] - ETA: 2:48 - loss: 0.1148 - dice_coefficient: 0.9553\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "305/400 [=====================>........] - ETA: 2:46 - loss: 0.1145 - dice_coefficient: 0.9553\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "306/400 [=====================>........] - ETA: 2:44 - loss: 0.1142 - dice_coefficient: 0.9554\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "307/400 [======================>.......] - ETA: 2:43 - loss: 0.1141 - dice_coefficient: 0.9554\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "308/400 [======================>.......] - ETA: 2:41 - loss: 0.1138 - dice_coefficient: 0.9555\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "309/400 [======================>.......] - ETA: 2:39 - loss: 0.1135 - dice_coefficient: 0.9556\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "310/400 [======================>.......] - ETA: 2:37 - loss: 0.1135 - dice_coefficient: 0.9554\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "311/400 [======================>.......] - ETA: 2:36 - loss: 0.1134 - dice_coefficient: 0.9554\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "312/400 [======================>.......] - ETA: 2:34 - loss: 0.1131 - dice_coefficient: 0.9555\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "313/400 [======================>.......] - ETA: 2:32 - loss: 0.1128 - dice_coefficient: 0.9555\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "314/400 [======================>.......] - ETA: 2:30 - loss: 0.1125 - dice_coefficient: 0.9556\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "315/400 [======================>.......] - ETA: 2:29 - loss: 0.1124 - dice_coefficient: 0.9556\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "316/400 [======================>.......] - ETA: 2:27 - loss: 0.1123 - dice_coefficient: 0.9555\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "317/400 [======================>.......] - ETA: 2:25 - loss: 0.1121 - dice_coefficient: 0.9556\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "318/400 [======================>.......] - ETA: 2:23 - loss: 0.1120 - dice_coefficient: 0.9556\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "319/400 [======================>.......] - ETA: 2:22 - loss: 0.1119 - dice_coefficient: 0.9556\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "320/400 [=======================>......] - ETA: 2:20 - loss: 0.1117 - dice_coefficient: 0.9557\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "321/400 [=======================>......] - ETA: 2:18 - loss: 0.1116 - dice_coefficient: 0.9557\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "322/400 [=======================>......] - ETA: 2:16 - loss: 0.1114 - dice_coefficient: 0.9558\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "323/400 [=======================>......] - ETA: 2:15 - loss: 0.1112 - dice_coefficient: 0.9558\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "324/400 [=======================>......] - ETA: 2:13 - loss: 0.1112 - dice_coefficient: 0.9557\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "325/400 [=======================>......] - ETA: 2:11 - loss: 0.1109 - dice_coefficient: 0.9557\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "326/400 [=======================>......] - ETA: 2:09 - loss: 0.1108 - dice_coefficient: 0.9557\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "327/400 [=======================>......] - ETA: 2:08 - loss: 0.1106 - dice_coefficient: 0.9558\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "328/400 [=======================>......] - ETA: 2:06 - loss: 0.1104 - dice_coefficient: 0.9558\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "329/400 [=======================>......] - ETA: 2:04 - loss: 0.1101 - dice_coefficient: 0.9559\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "330/400 [=======================>......] - ETA: 2:02 - loss: 0.1099 - dice_coefficient: 0.9560\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "331/400 [=======================>......] - ETA: 2:01 - loss: 0.1098 - dice_coefficient: 0.9559\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "332/400 [=======================>......] - ETA: 1:59 - loss: 0.1098 - dice_coefficient: 0.9558\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "333/400 [=======================>......] - ETA: 1:57 - loss: 0.1095 - dice_coefficient: 0.9559\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "334/400 [========================>.....] - ETA: 1:55 - loss: 0.1093 - dice_coefficient: 0.9560\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "335/400 [========================>.....] - ETA: 1:54 - loss: 0.1091 - dice_coefficient: 0.9560\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "336/400 [========================>.....] - ETA: 1:52 - loss: 0.1090 - dice_coefficient: 0.9560\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "337/400 [========================>.....] - ETA: 1:50 - loss: 0.1088 - dice_coefficient: 0.9561\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "338/400 [========================>.....] - ETA: 1:48 - loss: 0.1087 - dice_coefficient: 0.9561\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "339/400 [========================>.....] - ETA: 1:47 - loss: 0.1086 - dice_coefficient: 0.9560\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "340/400 [========================>.....] - ETA: 1:45 - loss: 0.1084 - dice_coefficient: 0.9560\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "341/400 [========================>.....] - ETA: 1:43 - loss: 0.1084 - dice_coefficient: 0.9560\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "342/400 [========================>.....] - ETA: 1:41 - loss: 0.1082 - dice_coefficient: 0.9560\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "343/400 [========================>.....] - ETA: 1:40 - loss: 0.1081 - dice_coefficient: 0.9561\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "344/400 [========================>.....] - ETA: 1:38 - loss: 0.1078 - dice_coefficient: 0.9561\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "345/400 [========================>.....] - ETA: 1:36 - loss: 0.1076 - dice_coefficient: 0.9562\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "346/400 [========================>.....] - ETA: 1:34 - loss: 0.1074 - dice_coefficient: 0.9562\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "347/400 [=========================>....] - ETA: 1:33 - loss: 0.1073 - dice_coefficient: 0.9562\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "348/400 [=========================>....] - ETA: 1:31 - loss: 0.1070 - dice_coefficient: 0.9563\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "349/400 [=========================>....] - ETA: 1:29 - loss: 0.1068 - dice_coefficient: 0.9563\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "350/400 [=========================>....] - ETA: 1:27 - loss: 0.1067 - dice_coefficient: 0.9564\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "351/400 [=========================>....] - ETA: 1:26 - loss: 0.1065 - dice_coefficient: 0.9564\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "352/400 [=========================>....] - ETA: 1:24 - loss: 0.1065 - dice_coefficient: 0.9563\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "353/400 [=========================>....] - ETA: 1:22 - loss: 0.1063 - dice_coefficient: 0.9564\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "354/400 [=========================>....] - ETA: 1:20 - loss: 0.1062 - dice_coefficient: 0.9564\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "355/400 [=========================>....] - ETA: 1:18 - loss: 0.1061 - dice_coefficient: 0.9563\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "356/400 [=========================>....] - ETA: 1:17 - loss: 0.1058 - dice_coefficient: 0.9564\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "357/400 [=========================>....] - ETA: 1:15 - loss: 0.1056 - dice_coefficient: 0.9565\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "358/400 [=========================>....] - ETA: 1:13 - loss: 0.1055 - dice_coefficient: 0.9565\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "359/400 [=========================>....] - ETA: 1:11 - loss: 0.1054 - dice_coefficient: 0.9565\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "360/400 [==========================>...] - ETA: 1:10 - loss: 0.1053 - dice_coefficient: 0.9565\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "361/400 [==========================>...] - ETA: 1:08 - loss: 0.1051 - dice_coefficient: 0.9565\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "362/400 [==========================>...] - ETA: 1:06 - loss: 0.1050 - dice_coefficient: 0.9565\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "363/400 [==========================>...] - ETA: 1:04 - loss: 0.1049 - dice_coefficient: 0.9566\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "364/400 [==========================>...] - ETA: 1:03 - loss: 0.1048 - dice_coefficient: 0.9566\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "365/400 [==========================>...] - ETA: 1:01 - loss: 0.1048 - dice_coefficient: 0.9565\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "366/400 [==========================>...] - ETA: 59s - loss: 0.1047 - dice_coefficient: 0.9566 \n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "367/400 [==========================>...] - ETA: 57s - loss: 0.1045 - dice_coefficient: 0.9566\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "368/400 [==========================>...] - ETA: 56s - loss: 0.1044 - dice_coefficient: 0.9566\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "369/400 [==========================>...] - ETA: 54s - loss: 0.1041 - dice_coefficient: 0.9567\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "370/400 [==========================>...] - ETA: 52s - loss: 0.1040 - dice_coefficient: 0.9567\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "371/400 [==========================>...] - ETA: 50s - loss: 0.1039 - dice_coefficient: 0.9567\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "372/400 [==========================>...] - ETA: 49s - loss: 0.1038 - dice_coefficient: 0.9567\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "373/400 [==========================>...] - ETA: 47s - loss: 0.1036 - dice_coefficient: 0.9567\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "374/400 [===========================>..] - ETA: 45s - loss: 0.1035 - dice_coefficient: 0.9567\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "375/400 [===========================>..] - ETA: 43s - loss: 0.1033 - dice_coefficient: 0.9568\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "376/400 [===========================>..] - ETA: 42s - loss: 0.1032 - dice_coefficient: 0.9568\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "377/400 [===========================>..] - ETA: 40s - loss: 0.1030 - dice_coefficient: 0.9568\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "378/400 [===========================>..] - ETA: 38s - loss: 0.1028 - dice_coefficient: 0.9569\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "379/400 [===========================>..] - ETA: 36s - loss: 0.1027 - dice_coefficient: 0.9569\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "380/400 [===========================>..] - ETA: 35s - loss: 0.1026 - dice_coefficient: 0.9569\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "381/400 [===========================>..] - ETA: 33s - loss: 0.1023 - dice_coefficient: 0.9570\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "382/400 [===========================>..] - ETA: 31s - loss: 0.1022 - dice_coefficient: 0.9570\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "383/400 [===========================>..] - ETA: 29s - loss: 0.1021 - dice_coefficient: 0.9570\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "384/400 [===========================>..] - ETA: 28s - loss: 0.1020 - dice_coefficient: 0.9571\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "385/400 [===========================>..] - ETA: 26s - loss: 0.1018 - dice_coefficient: 0.9571\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "386/400 [===========================>..] - ETA: 24s - loss: 0.1017 - dice_coefficient: 0.9571\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "387/400 [============================>.] - ETA: 22s - loss: 0.1015 - dice_coefficient: 0.9571\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "388/400 [============================>.] - ETA: 21s - loss: 0.1014 - dice_coefficient: 0.9571\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "389/400 [============================>.] - ETA: 19s - loss: 0.1012 - dice_coefficient: 0.9572\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "390/400 [============================>.] - ETA: 17s - loss: 0.1011 - dice_coefficient: 0.9572\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "391/400 [============================>.] - ETA: 15s - loss: 0.1010 - dice_coefficient: 0.9572\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "392/400 [============================>.] - ETA: 14s - loss: 0.1008 - dice_coefficient: 0.9573\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "393/400 [============================>.] - ETA: 12s - loss: 0.1008 - dice_coefficient: 0.9573\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "394/400 [============================>.] - ETA: 10s - loss: 0.1006 - dice_coefficient: 0.9573\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "395/400 [============================>.] - ETA: 8s - loss: 0.1006 - dice_coefficient: 0.9574 \n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "396/400 [============================>.] - ETA: 7s - loss: 0.1004 - dice_coefficient: 0.9574\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "397/400 [============================>.] - ETA: 5s - loss: 0.1004 - dice_coefficient: 0.9573\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "398/400 [============================>.] - ETA: 3s - loss: 0.1003 - dice_coefficient: 0.9573\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "399/400 [============================>.] - ETA: 1s - loss: 0.1002 - dice_coefficient: 0.9573\n",
      "Epoch 00029: loss did not improve from 0.04222\n",
      "400/400 [==============================] - 706s 2s/step - loss: 0.1001 - dice_coefficient: 0.9573 - val_loss: 1.0979 - val_dice_coefficient: 0.5634\n",
      "Epoch 30/30\n",
      "\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "  1/400 [..............................] - ETA: 0s - loss: 0.0621 - dice_coefficient: 0.9726\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "  2/400 [..............................] - ETA: 5:47 - loss: 0.0542 - dice_coefficient: 0.9701\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "  3/400 [..............................] - ETA: 7:42 - loss: 0.0568 - dice_coefficient: 0.9646\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "  4/400 [..............................] - ETA: 8:37 - loss: 0.0870 - dice_coefficient: 0.9347\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "  5/400 [..............................] - ETA: 9:09 - loss: 0.0751 - dice_coefficient: 0.9446\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "  6/400 [..............................] - ETA: 9:30 - loss: 0.0694 - dice_coefficient: 0.9493\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "  7/400 [..............................] - ETA: 9:48 - loss: 0.0763 - dice_coefficient: 0.9417\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "  8/400 [..............................] - ETA: 9:58 - loss: 0.0774 - dice_coefficient: 0.9409\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "  9/400 [..............................] - ETA: 10:06 - loss: 0.0765 - dice_coefficient: 0.9413\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 10/400 [..............................] - ETA: 10:12 - loss: 0.0925 - dice_coefficient: 0.9350\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 11/400 [..............................] - ETA: 10:16 - loss: 0.0890 - dice_coefficient: 0.9375\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 12/400 [..............................] - ETA: 10:20 - loss: 0.0848 - dice_coefficient: 0.9404\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 13/400 [..............................] - ETA: 10:24 - loss: 0.0819 - dice_coefficient: 0.9420\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 14/400 [>.............................] - ETA: 10:26 - loss: 0.0789 - dice_coefficient: 0.9440\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 15/400 [>.............................] - ETA: 10:28 - loss: 0.0765 - dice_coefficient: 0.9450\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 16/400 [>.............................] - ETA: 10:30 - loss: 0.0747 - dice_coefficient: 0.9463\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 17/400 [>.............................] - ETA: 10:32 - loss: 0.0743 - dice_coefficient: 0.9467\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 18/400 [>.............................] - ETA: 10:33 - loss: 0.0753 - dice_coefficient: 0.9455\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 19/400 [>.............................] - ETA: 10:34 - loss: 0.0756 - dice_coefficient: 0.9454\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 20/400 [>.............................] - ETA: 10:34 - loss: 0.0750 - dice_coefficient: 0.9459\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 21/400 [>.............................] - ETA: 10:34 - loss: 0.0746 - dice_coefficient: 0.9473\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 22/400 [>.............................] - ETA: 10:33 - loss: 0.0737 - dice_coefficient: 0.9483\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 23/400 [>.............................] - ETA: 10:33 - loss: 0.0735 - dice_coefficient: 0.9482\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 24/400 [>.............................] - ETA: 10:32 - loss: 0.0734 - dice_coefficient: 0.9480\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 25/400 [>.............................] - ETA: 10:32 - loss: 0.0728 - dice_coefficient: 0.9481\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 26/400 [>.............................] - ETA: 10:31 - loss: 0.0848 - dice_coefficient: 0.9399\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 27/400 [=>............................] - ETA: 10:30 - loss: 0.0830 - dice_coefficient: 0.9412\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 28/400 [=>............................] - ETA: 10:30 - loss: 0.0812 - dice_coefficient: 0.9425\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 29/400 [=>............................] - ETA: 10:29 - loss: 0.0797 - dice_coefficient: 0.9434\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 30/400 [=>............................] - ETA: 10:28 - loss: 0.0799 - dice_coefficient: 0.9438\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 31/400 [=>............................] - ETA: 10:26 - loss: 0.0805 - dice_coefficient: 0.9435\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 32/400 [=>............................] - ETA: 10:25 - loss: 0.0804 - dice_coefficient: 0.9434\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 33/400 [=>............................] - ETA: 10:24 - loss: 0.0813 - dice_coefficient: 0.9433\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 34/400 [=>............................] - ETA: 10:23 - loss: 0.0821 - dice_coefficient: 0.9423\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 35/400 [=>............................] - ETA: 10:22 - loss: 0.1061 - dice_coefficient: 0.9369\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 36/400 [=>............................] - ETA: 10:21 - loss: 0.1063 - dice_coefficient: 0.9364\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 37/400 [=>............................] - ETA: 10:19 - loss: 0.1047 - dice_coefficient: 0.9369\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 38/400 [=>............................] - ETA: 10:18 - loss: 0.1033 - dice_coefficient: 0.9375\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 39/400 [=>............................] - ETA: 10:17 - loss: 0.1037 - dice_coefficient: 0.9378\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 40/400 [==>...........................] - ETA: 10:16 - loss: 0.1025 - dice_coefficient: 0.9385\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 41/400 [==>...........................] - ETA: 10:15 - loss: 0.1019 - dice_coefficient: 0.9386\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 42/400 [==>...........................] - ETA: 10:13 - loss: 0.1026 - dice_coefficient: 0.9377\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 43/400 [==>...........................] - ETA: 10:12 - loss: 0.1030 - dice_coefficient: 0.9380\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 44/400 [==>...........................] - ETA: 10:10 - loss: 0.1025 - dice_coefficient: 0.9378\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 45/400 [==>...........................] - ETA: 10:09 - loss: 0.1016 - dice_coefficient: 0.9384\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 46/400 [==>...........................] - ETA: 10:07 - loss: 0.1001 - dice_coefficient: 0.9394\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 47/400 [==>...........................] - ETA: 10:06 - loss: 0.0985 - dice_coefficient: 0.9402\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 48/400 [==>...........................] - ETA: 10:05 - loss: 0.1023 - dice_coefficient: 0.9370\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 49/400 [==>...........................] - ETA: 10:03 - loss: 0.1020 - dice_coefficient: 0.9372\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 50/400 [==>...........................] - ETA: 10:02 - loss: 0.1037 - dice_coefficient: 0.9361\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 51/400 [==>...........................] - ETA: 10:00 - loss: 0.1025 - dice_coefficient: 0.9369\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 52/400 [==>...........................] - ETA: 9:59 - loss: 0.1025 - dice_coefficient: 0.9367 \n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 53/400 [==>...........................] - ETA: 9:58 - loss: 0.1017 - dice_coefficient: 0.9376\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 54/400 [===>..........................] - ETA: 9:56 - loss: 0.1010 - dice_coefficient: 0.9378\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 55/400 [===>..........................] - ETA: 9:54 - loss: 0.1015 - dice_coefficient: 0.9374\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 56/400 [===>..........................] - ETA: 9:53 - loss: 0.1007 - dice_coefficient: 0.9382\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 57/400 [===>..........................] - ETA: 9:51 - loss: 0.1000 - dice_coefficient: 0.9387\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 58/400 [===>..........................] - ETA: 9:50 - loss: 0.0991 - dice_coefficient: 0.9390\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 59/400 [===>..........................] - ETA: 9:48 - loss: 0.0991 - dice_coefficient: 0.9389\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 60/400 [===>..........................] - ETA: 9:46 - loss: 0.0989 - dice_coefficient: 0.9390\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 61/400 [===>..........................] - ETA: 9:45 - loss: 0.0983 - dice_coefficient: 0.9392\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 62/400 [===>..........................] - ETA: 9:43 - loss: 0.0981 - dice_coefficient: 0.9389\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 63/400 [===>..........................] - ETA: 9:42 - loss: 0.0978 - dice_coefficient: 0.9390\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 64/400 [===>..........................] - ETA: 9:40 - loss: 0.0984 - dice_coefficient: 0.9386\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 65/400 [===>..........................] - ETA: 9:39 - loss: 0.0986 - dice_coefficient: 0.9386\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 66/400 [===>..........................] - ETA: 9:37 - loss: 0.0984 - dice_coefficient: 0.9390\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 67/400 [====>.........................] - ETA: 9:35 - loss: 0.0983 - dice_coefficient: 0.9388\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 68/400 [====>.........................] - ETA: 9:34 - loss: 0.0981 - dice_coefficient: 0.9391\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 69/400 [====>.........................] - ETA: 9:32 - loss: 0.0987 - dice_coefficient: 0.9383\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 70/400 [====>.........................] - ETA: 9:31 - loss: 0.0978 - dice_coefficient: 0.9387\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 71/400 [====>.........................] - ETA: 9:29 - loss: 0.0975 - dice_coefficient: 0.9388\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 72/400 [====>.........................] - ETA: 9:28 - loss: 0.0979 - dice_coefficient: 0.9387\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 73/400 [====>.........................] - ETA: 9:26 - loss: 0.0984 - dice_coefficient: 0.9388\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 74/400 [====>.........................] - ETA: 9:24 - loss: 0.0998 - dice_coefficient: 0.9384\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 75/400 [====>.........................] - ETA: 9:23 - loss: 0.1000 - dice_coefficient: 0.9384\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 76/400 [====>.........................] - ETA: 9:21 - loss: 0.0993 - dice_coefficient: 0.9387\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 77/400 [====>.........................] - ETA: 9:20 - loss: 0.0991 - dice_coefficient: 0.9389\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 78/400 [====>.........................] - ETA: 9:18 - loss: 0.0983 - dice_coefficient: 0.9394\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 79/400 [====>.........................] - ETA: 9:17 - loss: 0.0979 - dice_coefficient: 0.9398\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 80/400 [=====>........................] - ETA: 9:16 - loss: 0.0978 - dice_coefficient: 0.9397\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 81/400 [=====>........................] - ETA: 9:14 - loss: 0.0974 - dice_coefficient: 0.9399\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 82/400 [=====>........................] - ETA: 9:13 - loss: 0.0977 - dice_coefficient: 0.9399\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 83/400 [=====>........................] - ETA: 9:11 - loss: 0.0982 - dice_coefficient: 0.9394\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 84/400 [=====>........................] - ETA: 9:09 - loss: 0.0988 - dice_coefficient: 0.9392\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 85/400 [=====>........................] - ETA: 9:08 - loss: 0.0990 - dice_coefficient: 0.9388\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 86/400 [=====>........................] - ETA: 9:06 - loss: 0.0987 - dice_coefficient: 0.9390\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 87/400 [=====>........................] - ETA: 9:04 - loss: 0.0988 - dice_coefficient: 0.9388\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 88/400 [=====>........................] - ETA: 9:03 - loss: 0.0982 - dice_coefficient: 0.9392\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 89/400 [=====>........................] - ETA: 9:01 - loss: 0.0980 - dice_coefficient: 0.9391\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 90/400 [=====>........................] - ETA: 8:59 - loss: 0.0977 - dice_coefficient: 0.9393\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 91/400 [=====>........................] - ETA: 8:58 - loss: 0.0973 - dice_coefficient: 0.9394\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 92/400 [=====>........................] - ETA: 8:56 - loss: 0.0967 - dice_coefficient: 0.9397\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 93/400 [=====>........................] - ETA: 8:54 - loss: 0.0962 - dice_coefficient: 0.9398\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 94/400 [======>.......................] - ETA: 8:53 - loss: 0.0963 - dice_coefficient: 0.9399\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 95/400 [======>.......................] - ETA: 8:51 - loss: 0.0957 - dice_coefficient: 0.9401\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 96/400 [======>.......................] - ETA: 8:49 - loss: 0.0960 - dice_coefficient: 0.9402\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 97/400 [======>.......................] - ETA: 8:47 - loss: 0.0958 - dice_coefficient: 0.9402\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 98/400 [======>.......................] - ETA: 8:46 - loss: 0.0955 - dice_coefficient: 0.9403\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      " 99/400 [======>.......................] - ETA: 8:44 - loss: 0.0951 - dice_coefficient: 0.9404\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "100/400 [======>.......................] - ETA: 8:42 - loss: 0.0945 - dice_coefficient: 0.9407\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "101/400 [======>.......................] - ETA: 8:41 - loss: 0.0948 - dice_coefficient: 0.9402\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "102/400 [======>.......................] - ETA: 8:39 - loss: 0.0944 - dice_coefficient: 0.9404\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "103/400 [======>.......................] - ETA: 8:37 - loss: 0.0943 - dice_coefficient: 0.9406\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "104/400 [======>.......................] - ETA: 8:36 - loss: 0.0941 - dice_coefficient: 0.9407\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "105/400 [======>.......................] - ETA: 8:34 - loss: 0.0940 - dice_coefficient: 0.9408\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "106/400 [======>.......................] - ETA: 8:32 - loss: 0.0937 - dice_coefficient: 0.9410\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "107/400 [=======>......................] - ETA: 8:30 - loss: 0.0936 - dice_coefficient: 0.9410\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "108/400 [=======>......................] - ETA: 8:29 - loss: 0.0933 - dice_coefficient: 0.9413\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "109/400 [=======>......................] - ETA: 8:27 - loss: 0.0929 - dice_coefficient: 0.9415\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "110/400 [=======>......................] - ETA: 8:25 - loss: 0.0923 - dice_coefficient: 0.9418\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "111/400 [=======>......................] - ETA: 8:24 - loss: 0.0923 - dice_coefficient: 0.9420\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "112/400 [=======>......................] - ETA: 8:22 - loss: 0.0933 - dice_coefficient: 0.9420\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "113/400 [=======>......................] - ETA: 8:20 - loss: 0.0929 - dice_coefficient: 0.9422\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "114/400 [=======>......................] - ETA: 8:19 - loss: 0.0926 - dice_coefficient: 0.9424\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "115/400 [=======>......................] - ETA: 8:17 - loss: 0.0922 - dice_coefficient: 0.9426\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "116/400 [=======>......................] - ETA: 8:15 - loss: 0.0919 - dice_coefficient: 0.9428\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "117/400 [=======>......................] - ETA: 8:13 - loss: 0.0914 - dice_coefficient: 0.9431\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "118/400 [=======>......................] - ETA: 8:12 - loss: 0.0911 - dice_coefficient: 0.9433\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "119/400 [=======>......................] - ETA: 8:10 - loss: 0.0917 - dice_coefficient: 0.9426\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "120/400 [========>.....................] - ETA: 8:08 - loss: 0.0916 - dice_coefficient: 0.9427\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "121/400 [========>.....................] - ETA: 8:06 - loss: 0.0912 - dice_coefficient: 0.9430\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "122/400 [========>.....................] - ETA: 8:05 - loss: 0.0910 - dice_coefficient: 0.9430\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "123/400 [========>.....................] - ETA: 8:03 - loss: 0.0907 - dice_coefficient: 0.9430\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "124/400 [========>.....................] - ETA: 8:01 - loss: 0.0908 - dice_coefficient: 0.9428\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "125/400 [========>.....................] - ETA: 8:00 - loss: 0.0904 - dice_coefficient: 0.9430\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "126/400 [========>.....................] - ETA: 7:58 - loss: 0.0900 - dice_coefficient: 0.9432\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "127/400 [========>.....................] - ETA: 7:56 - loss: 0.0912 - dice_coefficient: 0.9422\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "128/400 [========>.....................] - ETA: 7:54 - loss: 0.0915 - dice_coefficient: 0.9423\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "129/400 [========>.....................] - ETA: 7:53 - loss: 0.0918 - dice_coefficient: 0.9424\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "130/400 [========>.....................] - ETA: 7:51 - loss: 0.0916 - dice_coefficient: 0.9425\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "131/400 [========>.....................] - ETA: 7:49 - loss: 0.0913 - dice_coefficient: 0.9426\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "132/400 [========>.....................] - ETA: 7:47 - loss: 0.0910 - dice_coefficient: 0.9428\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "133/400 [========>.....................] - ETA: 7:46 - loss: 0.0910 - dice_coefficient: 0.9426\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "134/400 [=========>....................] - ETA: 7:44 - loss: 0.0907 - dice_coefficient: 0.9429\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "135/400 [=========>....................] - ETA: 7:42 - loss: 0.0904 - dice_coefficient: 0.9429\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "136/400 [=========>....................] - ETA: 7:40 - loss: 0.0902 - dice_coefficient: 0.9431\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "137/400 [=========>....................] - ETA: 7:39 - loss: 0.0900 - dice_coefficient: 0.9433\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "138/400 [=========>....................] - ETA: 7:37 - loss: 0.0903 - dice_coefficient: 0.9432\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "139/400 [=========>....................] - ETA: 7:35 - loss: 0.0901 - dice_coefficient: 0.9433\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "140/400 [=========>....................] - ETA: 7:33 - loss: 0.0898 - dice_coefficient: 0.9435\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "141/400 [=========>....................] - ETA: 7:32 - loss: 0.0896 - dice_coefficient: 0.9436\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "142/400 [=========>....................] - ETA: 7:30 - loss: 0.0898 - dice_coefficient: 0.9435\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "143/400 [=========>....................] - ETA: 7:28 - loss: 0.0895 - dice_coefficient: 0.9437\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "144/400 [=========>....................] - ETA: 7:27 - loss: 0.0901 - dice_coefficient: 0.9433\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "145/400 [=========>....................] - ETA: 7:25 - loss: 0.0903 - dice_coefficient: 0.9432\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "146/400 [=========>....................] - ETA: 7:23 - loss: 0.0903 - dice_coefficient: 0.9433\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "147/400 [==========>...................] - ETA: 7:21 - loss: 0.0901 - dice_coefficient: 0.9434\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "148/400 [==========>...................] - ETA: 7:20 - loss: 0.0897 - dice_coefficient: 0.9436\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "149/400 [==========>...................] - ETA: 7:18 - loss: 0.0897 - dice_coefficient: 0.9436\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "150/400 [==========>...................] - ETA: 7:16 - loss: 0.0894 - dice_coefficient: 0.9439\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "151/400 [==========>...................] - ETA: 7:14 - loss: 0.0901 - dice_coefficient: 0.9430\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "152/400 [==========>...................] - ETA: 7:13 - loss: 0.0899 - dice_coefficient: 0.9432\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "153/400 [==========>...................] - ETA: 7:11 - loss: 0.0900 - dice_coefficient: 0.9431\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "154/400 [==========>...................] - ETA: 7:09 - loss: 0.0899 - dice_coefficient: 0.9432\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "155/400 [==========>...................] - ETA: 7:07 - loss: 0.0894 - dice_coefficient: 0.9434\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "156/400 [==========>...................] - ETA: 7:06 - loss: 0.0891 - dice_coefficient: 0.9436\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "157/400 [==========>...................] - ETA: 7:04 - loss: 0.0888 - dice_coefficient: 0.9438\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "158/400 [==========>...................] - ETA: 7:02 - loss: 0.0885 - dice_coefficient: 0.9440\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "159/400 [==========>...................] - ETA: 7:00 - loss: 0.0881 - dice_coefficient: 0.9442\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "160/400 [===========>..................] - ETA: 6:59 - loss: 0.0880 - dice_coefficient: 0.9443\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "161/400 [===========>..................] - ETA: 6:57 - loss: 0.0878 - dice_coefficient: 0.9445\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "162/400 [===========>..................] - ETA: 6:55 - loss: 0.0876 - dice_coefficient: 0.9446\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "163/400 [===========>..................] - ETA: 6:53 - loss: 0.0880 - dice_coefficient: 0.9442\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "164/400 [===========>..................] - ETA: 6:52 - loss: 0.0876 - dice_coefficient: 0.9444\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "165/400 [===========>..................] - ETA: 6:50 - loss: 0.0875 - dice_coefficient: 0.9445\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "166/400 [===========>..................] - ETA: 6:48 - loss: 0.0873 - dice_coefficient: 0.9447\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "167/400 [===========>..................] - ETA: 6:46 - loss: 0.0871 - dice_coefficient: 0.9448\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "168/400 [===========>..................] - ETA: 6:45 - loss: 0.0868 - dice_coefficient: 0.9450\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "169/400 [===========>..................] - ETA: 6:43 - loss: 0.0865 - dice_coefficient: 0.9452\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "170/400 [===========>..................] - ETA: 6:41 - loss: 0.1812 - dice_coefficient: 0.9396\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "171/400 [===========>..................] - ETA: 6:39 - loss: 0.1803 - dice_coefficient: 0.9399\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "172/400 [===========>..................] - ETA: 6:38 - loss: 0.1797 - dice_coefficient: 0.9399\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "173/400 [===========>..................] - ETA: 6:36 - loss: 0.1788 - dice_coefficient: 0.9401\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "174/400 [============>.................] - ETA: 6:34 - loss: 0.1781 - dice_coefficient: 0.9403\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "175/400 [============>.................] - ETA: 6:33 - loss: 0.1777 - dice_coefficient: 0.9403\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "176/400 [============>.................] - ETA: 6:31 - loss: 0.1770 - dice_coefficient: 0.9404\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "177/400 [============>.................] - ETA: 6:29 - loss: 0.1764 - dice_coefficient: 0.9405\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "178/400 [============>.................] - ETA: 6:27 - loss: 0.1758 - dice_coefficient: 0.9405\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "179/400 [============>.................] - ETA: 6:26 - loss: 0.1749 - dice_coefficient: 0.9407\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "180/400 [============>.................] - ETA: 6:24 - loss: 0.1743 - dice_coefficient: 0.9408\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "181/400 [============>.................] - ETA: 6:22 - loss: 0.1735 - dice_coefficient: 0.9410\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "182/400 [============>.................] - ETA: 6:20 - loss: 0.1729 - dice_coefficient: 0.9411\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "183/400 [============>.................] - ETA: 6:19 - loss: 0.1723 - dice_coefficient: 0.9412\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "184/400 [============>.................] - ETA: 6:17 - loss: 0.1716 - dice_coefficient: 0.9413\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "185/400 [============>.................] - ETA: 6:15 - loss: 0.1709 - dice_coefficient: 0.9415\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "186/400 [============>.................] - ETA: 6:14 - loss: 0.1703 - dice_coefficient: 0.9415\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "187/400 [=============>................] - ETA: 6:12 - loss: 0.1699 - dice_coefficient: 0.9416\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "188/400 [=============>................] - ETA: 6:10 - loss: 0.1693 - dice_coefficient: 0.9418\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "189/400 [=============>................] - ETA: 6:08 - loss: 0.1689 - dice_coefficient: 0.9417\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "190/400 [=============>................] - ETA: 6:07 - loss: 0.1681 - dice_coefficient: 0.9419\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "191/400 [=============>................] - ETA: 6:05 - loss: 0.1677 - dice_coefficient: 0.9419\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "192/400 [=============>................] - ETA: 6:03 - loss: 0.1673 - dice_coefficient: 0.9417\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "193/400 [=============>................] - ETA: 6:02 - loss: 0.1667 - dice_coefficient: 0.9419\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "194/400 [=============>................] - ETA: 6:00 - loss: 0.1661 - dice_coefficient: 0.9420\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "195/400 [=============>................] - ETA: 5:58 - loss: 0.1655 - dice_coefficient: 0.9422\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "196/400 [=============>................] - ETA: 5:56 - loss: 0.1649 - dice_coefficient: 0.9422\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "197/400 [=============>................] - ETA: 5:55 - loss: 0.1646 - dice_coefficient: 0.9422\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "198/400 [=============>................] - ETA: 5:53 - loss: 0.1640 - dice_coefficient: 0.9423\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "199/400 [=============>................] - ETA: 5:51 - loss: 0.1634 - dice_coefficient: 0.9424\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "200/400 [==============>...............] - ETA: 5:50 - loss: 0.1629 - dice_coefficient: 0.9426\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "201/400 [==============>...............] - ETA: 5:48 - loss: 0.1623 - dice_coefficient: 0.9427\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "202/400 [==============>...............] - ETA: 5:46 - loss: 0.1617 - dice_coefficient: 0.9429\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "203/400 [==============>...............] - ETA: 5:44 - loss: 0.1611 - dice_coefficient: 0.9431\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "204/400 [==============>...............] - ETA: 5:43 - loss: 0.1604 - dice_coefficient: 0.9433\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "205/400 [==============>...............] - ETA: 5:41 - loss: 0.1600 - dice_coefficient: 0.9434\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "206/400 [==============>...............] - ETA: 5:39 - loss: 0.1595 - dice_coefficient: 0.9433\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "207/400 [==============>...............] - ETA: 5:37 - loss: 0.1590 - dice_coefficient: 0.9435\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "208/400 [==============>...............] - ETA: 5:36 - loss: 0.1585 - dice_coefficient: 0.9436\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "209/400 [==============>...............] - ETA: 5:34 - loss: 0.1581 - dice_coefficient: 0.9435\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "210/400 [==============>...............] - ETA: 5:32 - loss: 0.1575 - dice_coefficient: 0.9437\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "211/400 [==============>...............] - ETA: 5:31 - loss: 0.1572 - dice_coefficient: 0.9436\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "212/400 [==============>...............] - ETA: 5:29 - loss: 0.1566 - dice_coefficient: 0.9438\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "213/400 [==============>...............] - ETA: 5:27 - loss: 0.1562 - dice_coefficient: 0.9438\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "214/400 [===============>..............] - ETA: 5:25 - loss: 0.1558 - dice_coefficient: 0.9439\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "215/400 [===============>..............] - ETA: 5:24 - loss: 0.1554 - dice_coefficient: 0.9438\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "216/400 [===============>..............] - ETA: 5:22 - loss: 0.1549 - dice_coefficient: 0.9440\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "217/400 [===============>..............] - ETA: 5:20 - loss: 0.1546 - dice_coefficient: 0.9439\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "218/400 [===============>..............] - ETA: 5:18 - loss: 0.1541 - dice_coefficient: 0.9440\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "219/400 [===============>..............] - ETA: 5:17 - loss: 0.1536 - dice_coefficient: 0.9442\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "220/400 [===============>..............] - ETA: 5:15 - loss: 0.1530 - dice_coefficient: 0.9443\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "221/400 [===============>..............] - ETA: 5:13 - loss: 0.1526 - dice_coefficient: 0.9444\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "222/400 [===============>..............] - ETA: 5:11 - loss: 0.1520 - dice_coefficient: 0.9446\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "223/400 [===============>..............] - ETA: 5:10 - loss: 0.1521 - dice_coefficient: 0.9441\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "224/400 [===============>..............] - ETA: 5:08 - loss: 0.1517 - dice_coefficient: 0.9442\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "225/400 [===============>..............] - ETA: 5:06 - loss: 0.1514 - dice_coefficient: 0.9443\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "226/400 [===============>..............] - ETA: 5:04 - loss: 0.1509 - dice_coefficient: 0.9444\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "227/400 [================>.............] - ETA: 5:03 - loss: 0.1504 - dice_coefficient: 0.9445\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "228/400 [================>.............] - ETA: 5:01 - loss: 0.1500 - dice_coefficient: 0.9446\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "229/400 [================>.............] - ETA: 4:59 - loss: 0.1499 - dice_coefficient: 0.9443\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "230/400 [================>.............] - ETA: 4:57 - loss: 0.1495 - dice_coefficient: 0.9444\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "231/400 [================>.............] - ETA: 4:56 - loss: 0.1495 - dice_coefficient: 0.9443\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "232/400 [================>.............] - ETA: 4:54 - loss: 0.1491 - dice_coefficient: 0.9443\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "233/400 [================>.............] - ETA: 4:52 - loss: 0.1486 - dice_coefficient: 0.9445\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "234/400 [================>.............] - ETA: 4:50 - loss: 0.1482 - dice_coefficient: 0.9445\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "235/400 [================>.............] - ETA: 4:49 - loss: 0.1478 - dice_coefficient: 0.9446\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "236/400 [================>.............] - ETA: 4:47 - loss: 0.1473 - dice_coefficient: 0.9448\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "237/400 [================>.............] - ETA: 4:45 - loss: 0.1468 - dice_coefficient: 0.9449\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "238/400 [================>.............] - ETA: 4:43 - loss: 0.1463 - dice_coefficient: 0.9450\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "239/400 [================>.............] - ETA: 4:42 - loss: 0.1459 - dice_coefficient: 0.9451\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "240/400 [=================>............] - ETA: 4:40 - loss: 0.1456 - dice_coefficient: 0.9451\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "241/400 [=================>............] - ETA: 4:38 - loss: 0.1452 - dice_coefficient: 0.9452\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "242/400 [=================>............] - ETA: 4:36 - loss: 0.1448 - dice_coefficient: 0.9453\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "243/400 [=================>............] - ETA: 4:35 - loss: 0.1444 - dice_coefficient: 0.9454\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "244/400 [=================>............] - ETA: 4:33 - loss: 0.1440 - dice_coefficient: 0.9455\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "245/400 [=================>............] - ETA: 4:31 - loss: 0.1436 - dice_coefficient: 0.9457\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "246/400 [=================>............] - ETA: 4:30 - loss: 0.1432 - dice_coefficient: 0.9457\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "247/400 [=================>............] - ETA: 4:28 - loss: 0.1428 - dice_coefficient: 0.9458\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "248/400 [=================>............] - ETA: 4:26 - loss: 0.1425 - dice_coefficient: 0.9459\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "249/400 [=================>............] - ETA: 4:24 - loss: 0.1421 - dice_coefficient: 0.9460\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "250/400 [=================>............] - ETA: 4:23 - loss: 0.1420 - dice_coefficient: 0.9458\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "251/400 [=================>............] - ETA: 4:21 - loss: 0.1416 - dice_coefficient: 0.9459\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "252/400 [=================>............] - ETA: 4:19 - loss: 0.1414 - dice_coefficient: 0.9458\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "253/400 [=================>............] - ETA: 4:17 - loss: 0.1411 - dice_coefficient: 0.9458\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "254/400 [==================>...........] - ETA: 4:16 - loss: 0.1406 - dice_coefficient: 0.9460\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "255/400 [==================>...........] - ETA: 4:14 - loss: 0.1402 - dice_coefficient: 0.9461\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "256/400 [==================>...........] - ETA: 4:12 - loss: 0.1399 - dice_coefficient: 0.9462\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "257/400 [==================>...........] - ETA: 4:10 - loss: 0.1397 - dice_coefficient: 0.9461\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "258/400 [==================>...........] - ETA: 4:09 - loss: 0.1394 - dice_coefficient: 0.9462\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "259/400 [==================>...........] - ETA: 4:07 - loss: 0.1392 - dice_coefficient: 0.9461\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "260/400 [==================>...........] - ETA: 4:05 - loss: 0.1389 - dice_coefficient: 0.9461\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "261/400 [==================>...........] - ETA: 4:03 - loss: 0.1387 - dice_coefficient: 0.9462\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "262/400 [==================>...........] - ETA: 4:02 - loss: 0.1386 - dice_coefficient: 0.9461\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "263/400 [==================>...........] - ETA: 4:00 - loss: 0.1384 - dice_coefficient: 0.9461\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "264/400 [==================>...........] - ETA: 3:58 - loss: 0.1380 - dice_coefficient: 0.9462\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "265/400 [==================>...........] - ETA: 3:56 - loss: 0.1376 - dice_coefficient: 0.9464\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "266/400 [==================>...........] - ETA: 3:55 - loss: 0.1371 - dice_coefficient: 0.9466\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "267/400 [===================>..........] - ETA: 3:53 - loss: 0.1368 - dice_coefficient: 0.9467\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "268/400 [===================>..........] - ETA: 3:51 - loss: 0.1366 - dice_coefficient: 0.9467\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "269/400 [===================>..........] - ETA: 3:49 - loss: 0.1365 - dice_coefficient: 0.9466\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "270/400 [===================>..........] - ETA: 3:48 - loss: 0.1362 - dice_coefficient: 0.9467\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "271/400 [===================>..........] - ETA: 3:46 - loss: 0.1359 - dice_coefficient: 0.9468\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "272/400 [===================>..........] - ETA: 3:44 - loss: 0.1356 - dice_coefficient: 0.9469\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "273/400 [===================>..........] - ETA: 3:42 - loss: 0.1354 - dice_coefficient: 0.9469\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "274/400 [===================>..........] - ETA: 3:41 - loss: 0.1353 - dice_coefficient: 0.9467\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "275/400 [===================>..........] - ETA: 3:39 - loss: 0.1352 - dice_coefficient: 0.9466\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "276/400 [===================>..........] - ETA: 3:37 - loss: 0.1350 - dice_coefficient: 0.9465\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "277/400 [===================>..........] - ETA: 3:35 - loss: 0.1346 - dice_coefficient: 0.9467\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "278/400 [===================>..........] - ETA: 3:34 - loss: 0.1345 - dice_coefficient: 0.9467\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "279/400 [===================>..........] - ETA: 3:32 - loss: 0.1342 - dice_coefficient: 0.9468\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "280/400 [====================>.........] - ETA: 3:30 - loss: 0.1341 - dice_coefficient: 0.9467\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "281/400 [====================>.........] - ETA: 3:28 - loss: 0.1338 - dice_coefficient: 0.9468\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "282/400 [====================>.........] - ETA: 3:27 - loss: 0.1336 - dice_coefficient: 0.9468\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "283/400 [====================>.........] - ETA: 3:25 - loss: 0.1332 - dice_coefficient: 0.9470\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "284/400 [====================>.........] - ETA: 3:23 - loss: 0.1330 - dice_coefficient: 0.9469\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "285/400 [====================>.........] - ETA: 3:21 - loss: 0.1328 - dice_coefficient: 0.9470\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "286/400 [====================>.........] - ETA: 3:20 - loss: 0.1325 - dice_coefficient: 0.9471\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "287/400 [====================>.........] - ETA: 3:18 - loss: 0.1326 - dice_coefficient: 0.9467\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "288/400 [====================>.........] - ETA: 3:16 - loss: 0.1324 - dice_coefficient: 0.9468\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "289/400 [====================>.........] - ETA: 3:14 - loss: 0.1322 - dice_coefficient: 0.9468\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "290/400 [====================>.........] - ETA: 3:13 - loss: 0.1320 - dice_coefficient: 0.9469\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "291/400 [====================>.........] - ETA: 3:11 - loss: 0.1316 - dice_coefficient: 0.9470\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "292/400 [====================>.........] - ETA: 3:09 - loss: 0.1313 - dice_coefficient: 0.9470\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "293/400 [====================>.........] - ETA: 3:07 - loss: 0.1310 - dice_coefficient: 0.9472\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "294/400 [=====================>........] - ETA: 3:06 - loss: 0.1307 - dice_coefficient: 0.9472\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "295/400 [=====================>........] - ETA: 3:04 - loss: 0.1303 - dice_coefficient: 0.9473\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "296/400 [=====================>........] - ETA: 3:02 - loss: 0.1300 - dice_coefficient: 0.9474\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "297/400 [=====================>........] - ETA: 3:00 - loss: 0.1297 - dice_coefficient: 0.9475\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "298/400 [=====================>........] - ETA: 2:59 - loss: 0.1294 - dice_coefficient: 0.9476\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "299/400 [=====================>........] - ETA: 2:57 - loss: 0.1291 - dice_coefficient: 0.9476\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "300/400 [=====================>........] - ETA: 2:55 - loss: 0.1288 - dice_coefficient: 0.9477\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "301/400 [=====================>........] - ETA: 2:53 - loss: 0.1285 - dice_coefficient: 0.9478\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "302/400 [=====================>........] - ETA: 2:52 - loss: 0.1283 - dice_coefficient: 0.9478\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "303/400 [=====================>........] - ETA: 2:50 - loss: 0.1281 - dice_coefficient: 0.9478\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "304/400 [=====================>........] - ETA: 2:48 - loss: 0.1280 - dice_coefficient: 0.9479\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "305/400 [=====================>........] - ETA: 2:46 - loss: 0.1277 - dice_coefficient: 0.9479\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "306/400 [=====================>........] - ETA: 2:45 - loss: 0.1275 - dice_coefficient: 0.9479\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "307/400 [======================>.......] - ETA: 2:43 - loss: 0.1273 - dice_coefficient: 0.9480\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "308/400 [======================>.......] - ETA: 2:41 - loss: 0.1270 - dice_coefficient: 0.9481\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "309/400 [======================>.......] - ETA: 2:39 - loss: 0.1266 - dice_coefficient: 0.9482\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "310/400 [======================>.......] - ETA: 2:38 - loss: 0.1264 - dice_coefficient: 0.9483\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "311/400 [======================>.......] - ETA: 2:36 - loss: 0.1261 - dice_coefficient: 0.9483\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "312/400 [======================>.......] - ETA: 2:34 - loss: 0.1258 - dice_coefficient: 0.9484\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "313/400 [======================>.......] - ETA: 2:32 - loss: 0.1259 - dice_coefficient: 0.9481\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "314/400 [======================>.......] - ETA: 2:31 - loss: 0.1256 - dice_coefficient: 0.9482\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "315/400 [======================>.......] - ETA: 2:29 - loss: 0.1255 - dice_coefficient: 0.9482\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "316/400 [======================>.......] - ETA: 2:27 - loss: 0.1253 - dice_coefficient: 0.9483\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "317/400 [======================>.......] - ETA: 2:25 - loss: 0.1250 - dice_coefficient: 0.9484\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "318/400 [======================>.......] - ETA: 2:24 - loss: 0.1248 - dice_coefficient: 0.9484\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "319/400 [======================>.......] - ETA: 2:22 - loss: 0.1245 - dice_coefficient: 0.9485\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "320/400 [=======================>......] - ETA: 2:20 - loss: 0.1242 - dice_coefficient: 0.9486\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "321/400 [=======================>......] - ETA: 2:18 - loss: 0.1241 - dice_coefficient: 0.9486\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "322/400 [=======================>......] - ETA: 2:17 - loss: 0.1238 - dice_coefficient: 0.9486\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "323/400 [=======================>......] - ETA: 2:15 - loss: 0.1235 - dice_coefficient: 0.9487\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "324/400 [=======================>......] - ETA: 2:13 - loss: 0.1234 - dice_coefficient: 0.9487\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "325/400 [=======================>......] - ETA: 2:11 - loss: 0.1233 - dice_coefficient: 0.9487\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "326/400 [=======================>......] - ETA: 2:10 - loss: 0.1232 - dice_coefficient: 0.9488\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "327/400 [=======================>......] - ETA: 2:08 - loss: 0.1229 - dice_coefficient: 0.9488\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "328/400 [=======================>......] - ETA: 2:06 - loss: 0.1227 - dice_coefficient: 0.9489\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "329/400 [=======================>......] - ETA: 2:04 - loss: 0.1225 - dice_coefficient: 0.9489\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "330/400 [=======================>......] - ETA: 2:03 - loss: 0.1224 - dice_coefficient: 0.9489\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "331/400 [=======================>......] - ETA: 2:01 - loss: 0.1222 - dice_coefficient: 0.9490\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "332/400 [=======================>......] - ETA: 1:59 - loss: 0.1220 - dice_coefficient: 0.9490\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "333/400 [=======================>......] - ETA: 1:57 - loss: 0.1218 - dice_coefficient: 0.9490\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "334/400 [========================>.....] - ETA: 1:55 - loss: 0.1216 - dice_coefficient: 0.9491\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "335/400 [========================>.....] - ETA: 1:54 - loss: 0.1213 - dice_coefficient: 0.9492\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "336/400 [========================>.....] - ETA: 1:52 - loss: 0.1212 - dice_coefficient: 0.9492\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "337/400 [========================>.....] - ETA: 1:50 - loss: 0.1210 - dice_coefficient: 0.9492\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "338/400 [========================>.....] - ETA: 1:48 - loss: 0.1208 - dice_coefficient: 0.9493\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "339/400 [========================>.....] - ETA: 1:47 - loss: 0.1205 - dice_coefficient: 0.9494\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "340/400 [========================>.....] - ETA: 1:45 - loss: 0.1203 - dice_coefficient: 0.9495\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "341/400 [========================>.....] - ETA: 1:43 - loss: 0.1201 - dice_coefficient: 0.9495\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "342/400 [========================>.....] - ETA: 1:41 - loss: 0.1199 - dice_coefficient: 0.9496\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "343/400 [========================>.....] - ETA: 1:40 - loss: 0.1196 - dice_coefficient: 0.9497\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "344/400 [========================>.....] - ETA: 1:38 - loss: 0.1193 - dice_coefficient: 0.9498\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "345/400 [========================>.....] - ETA: 1:36 - loss: 0.1190 - dice_coefficient: 0.9499\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "346/400 [========================>.....] - ETA: 1:34 - loss: 0.1188 - dice_coefficient: 0.9499\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "347/400 [=========================>....] - ETA: 1:33 - loss: 0.1187 - dice_coefficient: 0.9499\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "348/400 [=========================>....] - ETA: 1:31 - loss: 0.1184 - dice_coefficient: 0.9500\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "349/400 [=========================>....] - ETA: 1:29 - loss: 0.1181 - dice_coefficient: 0.9501\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "350/400 [=========================>....] - ETA: 1:27 - loss: 0.1180 - dice_coefficient: 0.9501\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "351/400 [=========================>....] - ETA: 1:26 - loss: 0.1177 - dice_coefficient: 0.9502\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "352/400 [=========================>....] - ETA: 1:24 - loss: 0.1175 - dice_coefficient: 0.9503\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "353/400 [=========================>....] - ETA: 1:22 - loss: 0.1174 - dice_coefficient: 0.9503\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "354/400 [=========================>....] - ETA: 1:20 - loss: 0.1171 - dice_coefficient: 0.9504\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "355/400 [=========================>....] - ETA: 1:19 - loss: 0.1169 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "356/400 [=========================>....] - ETA: 1:17 - loss: 0.1168 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "357/400 [=========================>....] - ETA: 1:15 - loss: 0.1167 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "358/400 [=========================>....] - ETA: 1:13 - loss: 0.1167 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "359/400 [=========================>....] - ETA: 1:12 - loss: 0.1165 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "360/400 [==========================>...] - ETA: 1:10 - loss: 0.1163 - dice_coefficient: 0.9506\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "361/400 [==========================>...] - ETA: 1:08 - loss: 0.1161 - dice_coefficient: 0.9506\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "362/400 [==========================>...] - ETA: 1:06 - loss: 0.1161 - dice_coefficient: 0.9506\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "363/400 [==========================>...] - ETA: 1:05 - loss: 0.1158 - dice_coefficient: 0.9507\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "364/400 [==========================>...] - ETA: 1:03 - loss: 0.1159 - dice_coefficient: 0.9506\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "365/400 [==========================>...] - ETA: 1:01 - loss: 0.1159 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "366/400 [==========================>...] - ETA: 59s - loss: 0.1157 - dice_coefficient: 0.9505 \n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "367/400 [==========================>...] - ETA: 57s - loss: 0.1156 - dice_coefficient: 0.9506\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "368/400 [==========================>...] - ETA: 56s - loss: 0.1153 - dice_coefficient: 0.9507\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "369/400 [==========================>...] - ETA: 54s - loss: 0.1154 - dice_coefficient: 0.9504\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "370/400 [==========================>...] - ETA: 52s - loss: 0.1152 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "371/400 [==========================>...] - ETA: 50s - loss: 0.1151 - dice_coefficient: 0.9506\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "372/400 [==========================>...] - ETA: 49s - loss: 0.1150 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "373/400 [==========================>...] - ETA: 47s - loss: 0.1150 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "374/400 [===========================>..] - ETA: 45s - loss: 0.1150 - dice_coefficient: 0.9504\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "375/400 [===========================>..] - ETA: 43s - loss: 0.1149 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "376/400 [===========================>..] - ETA: 42s - loss: 0.1147 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "377/400 [===========================>..] - ETA: 40s - loss: 0.1146 - dice_coefficient: 0.9505\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "378/400 [===========================>..] - ETA: 38s - loss: 0.1144 - dice_coefficient: 0.9506\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "379/400 [===========================>..] - ETA: 36s - loss: 0.1142 - dice_coefficient: 0.9506\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "380/400 [===========================>..] - ETA: 35s - loss: 0.1141 - dice_coefficient: 0.9506\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "381/400 [===========================>..] - ETA: 33s - loss: 0.1140 - dice_coefficient: 0.9506\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "382/400 [===========================>..] - ETA: 31s - loss: 0.1139 - dice_coefficient: 0.9507\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "383/400 [===========================>..] - ETA: 29s - loss: 0.1137 - dice_coefficient: 0.9508\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "384/400 [===========================>..] - ETA: 28s - loss: 0.1136 - dice_coefficient: 0.9507\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "385/400 [===========================>..] - ETA: 26s - loss: 0.1134 - dice_coefficient: 0.9507\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "386/400 [===========================>..] - ETA: 24s - loss: 0.1133 - dice_coefficient: 0.9507\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "387/400 [============================>.] - ETA: 22s - loss: 0.1131 - dice_coefficient: 0.9508\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "388/400 [============================>.] - ETA: 21s - loss: 0.1129 - dice_coefficient: 0.9509\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "389/400 [============================>.] - ETA: 19s - loss: 0.1127 - dice_coefficient: 0.9509\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "390/400 [============================>.] - ETA: 17s - loss: 0.1125 - dice_coefficient: 0.9510\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "391/400 [============================>.] - ETA: 15s - loss: 0.1123 - dice_coefficient: 0.9510\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "392/400 [============================>.] - ETA: 14s - loss: 0.1120 - dice_coefficient: 0.9511\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "393/400 [============================>.] - ETA: 12s - loss: 0.1119 - dice_coefficient: 0.9511\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "394/400 [============================>.] - ETA: 10s - loss: 0.1122 - dice_coefficient: 0.9508\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "395/400 [============================>.] - ETA: 8s - loss: 0.1120 - dice_coefficient: 0.9509 \n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "396/400 [============================>.] - ETA: 7s - loss: 0.1119 - dice_coefficient: 0.9509\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "397/400 [============================>.] - ETA: 5s - loss: 0.1117 - dice_coefficient: 0.9510\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "398/400 [============================>.] - ETA: 3s - loss: 0.1116 - dice_coefficient: 0.9510\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "399/400 [============================>.] - ETA: 1s - loss: 0.1114 - dice_coefficient: 0.9511\n",
      "Epoch 00030: loss did not improve from 0.04222\n",
      "400/400 [==============================] - 707s 2s/step - loss: 0.1112 - dice_coefficient: 0.9511 - val_loss: 1.3410 - val_dice_coefficient: 0.5403\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x7f2b28b0ad30>"
      ]
     },
     "execution_count": 25,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, y_train, epochs = 30, batch_size = 1, callbacks = [checkpoint, reduce_lr, stop], validation_data = (X_test, y_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "jdkEeSsnejc9",
    "outputId": "749410c0-bcd1-4bc3-9658-134e7d085224"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 1ms/step - loss: 1.0759 - dice_coefficient: 0.5267\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[1.0759388208389282, 0.5267057418823242]"
      ]
     },
     "execution_count": 26,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(X_test, y_test, verbose = 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "z1UpPBF97x7b"
   },
   "source": [
    "#### Get the predicted mask for a sample image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "rmwAYQB-7s0M",
    "outputId": "e89e91d8-d819-4fc3-a9cd-41608bdbdd1a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 2ms/step\n"
     ]
    }
   ],
   "source": [
    "# Load previous model weight\n",
    "WEIGHTS_FILE = \"model_0.04.h5\"\n",
    "learned_model = create_model()\n",
    "learned_model.load_weights(WEIGHTS_FILE)\n",
    "y_pred = learned_model.predict(X_test, verbose = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 457
    },
    "id": "Bmt64B0t8Tqh",
    "outputId": "8a13801e-ca79-44e0-a37b-65c475b4c94c"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f2b1c2f3780>"
      ]
     },
     "execution_count": 35,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x518.4 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# For a sample image\n",
    "n = 5\n",
    "image = cv2.resize(X_test[n], dsize = (IMAGE_HEIGHT, IMAGE_WIDTH), interpolation = cv2.INTER_CUBIC)\n",
    "pred_mask = cv2.resize(1.0 * (y_pred[n] > 0.1), (IMAGE_WIDTH, IMAGE_HEIGHT))\n",
    "\n",
    "image2 = image\n",
    "image2[:, :, 0] = pred_mask * image[:, :, 0]\n",
    "image2[:, :, 1] = pred_mask * image[:, :, 1]\n",
    "image2[:, :, 2] = pred_mask * image[:, :, 2]\n",
    "out_image = image2\n",
    "\n",
    "fig = plt.figure(figsize = (15, 7.2))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "plt.axis('off')\n",
    "plt.imshow(out_image)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 439
    },
    "id": "DWU-6P-_9qth",
    "outputId": "55976318-364c-4a78-b29b-f44cc3923672"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f2b1c3f5630>"
      ]
     },
     "execution_count": 36,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x518.4 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize = (15, 7.2))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "plt.axis('off')\n",
    "plt.imshow(pred_mask, alpha = 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Z7gL80hK-K-i"
   },
   "source": [
    "#### Impose the mask on the image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 439
    },
    "id": "gS3kgVDl-DPb",
    "outputId": "2873c4e9-8a0a-4e8a-9b3f-3ec014f54c64"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x518.4 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize = (15, 7.2))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "plt.axis('off')\n",
    "plt.imshow(X_test[n])\n",
    "plt.savefig('image.jpg', bbox_inches = 'tight', pad_inches = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 422
    },
    "id": "3o4KVLRJ-UdM",
    "outputId": "ea61d801-7998-4b3b-b274-cbce5e9d5edc"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x518.4 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize = (15, 7.2))\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "plt.axis('off')\n",
    "plt.imshow(y_pred[n], alpha = 0.8)\n",
    "plt.savefig('mask.jpg', bbox_inches = 'tight', pad_inches = 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Thus, from this exercise we were able to decipher the following points:\n",
    "- The idea behind the designing model in order to predict the boundaries (mask) around a face was **how to use Transfer Learning**. Here, we used a pretrained MobileNet and on top it we added the UNET layers to train, fit and evaluate the model\n",
    "- Here our model was compiled using **binary cross entropy** as loss, **Adam Optimizer** as our optimizer and **Dice Coefficient** as metric\n",
    "- For callbacks we used **Model Checkpoint**, **Early Stopping** and **ReduceLROnPlateau**\n",
    "- We had split the given WIDER dataset (a subset of the original one) into two where we had **400 images in training set** and **9 images in test set**\n",
    "- After training the model for 30 epochs on training set, the best loss that I had acheived was **0.11 on loss and a dice coefficient of 0.95**\n",
    "- Furthermore, we used model weights (in this case - model_0.04.h5) on our test data to predict the mask around a face\n",
    "- As seen in the above example (for image = 5 of our test data), our model does a good job in predicting the masks."
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "Project 8- CV_Face_Mask_Segmentation_Raman.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}