DeepCell_demo_train_pred_nuclear.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "330e05e3",
   "metadata": {},
   "source": [
    "**Traing DeepCell Model and Make Predictions**<br>\n",
    "It trains a model with three heads: inner distance, outer distance, and fgbg. It works on tensorflow 2.7.1.<br>\n",
    "The demo datasets used in training is from the \"Intro to DeepCell\" notebook and are 256x256. In contrast, the pretrained model in NuclearApplication was trained on images of 512x512."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "05468ce3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.7.1\n"
     ]
    }
   ],
   "source": [
    "import syotil\n",
    "\n",
    "import numpy as np\n",
    "from skimage import io\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline\n",
    "from timeit import default_timer\n",
    "\n",
    "import tensorflow as tf\n",
    "from tensorflow.keras.optimizers import SGD, Adam\n",
    "from tensorflow.keras.losses import MSE\n",
    "\n",
    "import deepcell\n",
    "from deepcell.utils.train_utils import rate_scheduler, get_callbacks, count_gpus\n",
    "from deepcell.losses import weighted_categorical_crossentropy\n",
    "from deepcell.model_zoo.panopticnet import PanopticNet\n",
    "from deepcell_toolbox.deep_watershed import deep_watershed\n",
    "from deepcell import image_generators\n",
    "\n",
    "print(tf.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "262be798",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dict = np.load('/fh/fast/fong_y/tissuenet_v1.0/tissuenet_v1.0_train.npz')\n",
    "train_X, train_y = train_dict['X'], train_dict['y']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "593d8ef1",
   "metadata": {},
   "outputs": [],
   "source": [
    "base_url = ('https://deepcell-data.s3-us-west-1.amazonaws.com/'\n",
    "            'demos/janelia/hela_s3_{}_256.npz')\n",
    "\n",
    "nuclear = deepcell.datasets.Dataset(\n",
    "    path='nuclear.npz',\n",
    "    url=base_url.format('nuclear'),\n",
    "     file_hash=None,\n",
    "    metadata=None\n",
    ")\n",
    "\n",
    "fluo_cyto = deepcell.datasets.Dataset(\n",
    "    path='fluo_cyto.npz',\n",
    "    url=base_url.format('fluo_cyto'),\n",
    "     file_hash=None,\n",
    "    metadata=None\n",
    ")\n",
    "\n",
    "phase = deepcell.datasets.Dataset(\n",
    "    path='phase.npz',\n",
    "    url=base_url.format('phase'),\n",
    "     file_hash=None,\n",
    "    metadata=None\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "73d0c5d8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train.shape: (1728, 256, 256, 1)\n",
      "X_test.shape: (432, 256, 256, 1)\n",
      "WARNING:tensorflow:Removing 91 of 1728 images with fewer than 2 objects.\n",
      "WARNING:tensorflow:Removing 19 of 432 images with fewer than 2 objects.\n"
     ]
    }
   ],
   "source": [
    "dataset = nuclear  # nuclear, phase, fluo_cyto\n",
    "model_name = 'nuclear2_deep_watershed' \n",
    "\n",
    "min_objects = 2\n",
    "\n",
    "test_size = 0.2 # fraction of data saved as test\n",
    "seed = 0 # seed for random train-test split\n",
    "\n",
    "(X_train, y_train), (X_test, y_test) = dataset.load_data(\n",
    "    test_size=test_size, seed=seed)\n",
    "\n",
    "print('X_train.shape: {}\\nX_test.shape: {}'.format(\n",
    "    X_train.shape, X_test.shape))\n",
    "transforms = ['inner-distance', 'outer-distance', 'fgbg']\n",
    "transforms_kwargs = {'outer-distance': {'erosion_width': 0}}\n",
    "\n",
    "\n",
    "# use augmentation for training but not validation\n",
    "datagen = image_generators.SemanticDataGenerator(\n",
    "    rotation_range=180,\n",
    "    fill_mode='reflect',\n",
    "    zoom_range=(0.75, 1.25),\n",
    "    horizontal_flip=True,\n",
    "    vertical_flip=True)\n",
    "\n",
    "datagen_val = image_generators.SemanticDataGenerator()\n",
    "\n",
    "batch_size = 8\n",
    "\n",
    "train_data = datagen.flow(\n",
    "    {'X': X_train, 'y': y_train},\n",
    "    seed=seed,\n",
    "    transforms=transforms,\n",
    "    transforms_kwargs=transforms_kwargs,\n",
    "    min_objects=min_objects,\n",
    "    batch_size=batch_size)\n",
    "\n",
    "val_data = datagen_val.flow(\n",
    "    {'X': X_test, 'y': y_test},\n",
    "    seed=seed,\n",
    "    transforms=transforms,\n",
    "    transforms_kwargs=transforms_kwargs,\n",
    "    min_objects=min_objects,\n",
    "    batch_size=batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "0b79383a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1728, 256, 256, 1)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/app/software/scikit-image/0.19.1-foss-2021b/lib/python3.9/site-packages/skimage/io/_plugins/matplotlib_plugin.py:150: UserWarning: Low image data range; displaying image with stretched contrast.\n",
      "  lo, hi, cmap = _get_display_range(image)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(y_train.shape)\n",
    "\n",
    "i=5\n",
    "\n",
    "io.imshow(X_train[i,:,:,0])\n",
    "plt.show()\n",
    "#print(X_train[i,:,:,0])\n",
    "\n",
    "tmp = syotil.masks_to_outlines(y_train[i,:,:,0])\n",
    "io.imshow(tmp)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "1eb24799",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_path = '{}.h5'.format(model_name)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9591736f",
   "metadata": {},
   "source": [
    "**The two cells below start train the model and can be skipped if trained model will be loaded.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "7ebc07f0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-10-29 12:04:34.235126: W tensorflow/python/util/util.cc:368] Sets are not currently considered sequences, but this may change in the future, so consider avoiding using them.\n",
      "2022-10-29 12:04:37.605607: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 14657 MB memory:  -> device: 0, name: Tesla V100-PCIE-16GB, pci bus id: 0000:04:00.0, compute capability: 7.0\n",
      "2022-10-29 12:04:37.607163: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /job:localhost/replica:0/task:0/device:GPU:1 with 14657 MB memory:  -> device: 1, name: Tesla V100-PCIE-16GB, pci bus id: 0000:06:00.0, compute capability: 7.0\n",
      "2022-10-29 12:04:37.608455: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /job:localhost/replica:0/task:0/device:GPU:2 with 14657 MB memory:  -> device: 2, name: Tesla V100-PCIE-16GB, pci bus id: 0000:08:00.0, compute capability: 7.0\n",
      "/app/software/TensorFlow/2.7.1-foss-2021b-CUDA-11.4.1/lib/python3.9/site-packages/keras/optimizer_v2/adam.py:105: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.\n",
      "  super(Adam, self).__init__(name, **kwargs)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[('semantic_0', (None, 256, 256, 1)),\n",
       " ('semantic_1', (None, 256, 256, 1)),\n",
       " ('semantic_2', (None, 256, 256, 2))]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "semantic_classes = [1, 1, 2] # inner distance, outer distance, fgbg\n",
    "\n",
    "model = PanopticNet(\n",
    "    backbone='resnet50',\n",
    "    input_shape=X_train.shape[1:],\n",
    "    norm_method='whole_image',\n",
    "    num_semantic_classes=semantic_classes)\n",
    "\n",
    "lr = 1e-4\n",
    "optimizer = Adam(lr=lr, clipnorm=0.001)\n",
    "lr_sched = rate_scheduler(lr=lr, decay=0.99)\n",
    "\n",
    "# Create a dictionary of losses for each semantic head\n",
    "\n",
    "def semantic_loss(n_classes):\n",
    "    def _semantic_loss(y_pred, y_true):\n",
    "        if n_classes > 1:\n",
    "            return 0.01 * weighted_categorical_crossentropy(\n",
    "                y_pred, y_true, n_classes=n_classes)\n",
    "        return MSE(y_pred, y_true)\n",
    "    return _semantic_loss\n",
    "\n",
    "loss = {}\n",
    "\n",
    "# Give losses for all of the semantic heads\n",
    "for layer in model.layers:\n",
    "    if layer.name.startswith('semantic_'):\n",
    "        n_classes = layer.output_shape[-1]\n",
    "        loss[layer.name] = semantic_loss(n_classes)\n",
    "        \n",
    "model.compile(loss=loss, optimizer=optimizer)\n",
    "\n",
    "[(layer.name, layer.output_shape) for layer in filter(lambda x: x.name.startswith('semantic_'), model.layers)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b1fab4f5",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-10-29 12:04:44.033321: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /device:GPU:0 with 14657 MB memory:  -> device: 0, name: Tesla V100-PCIE-16GB, pci bus id: 0000:04:00.0, compute capability: 7.0\n",
      "2022-10-29 12:04:44.034349: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /device:GPU:1 with 14657 MB memory:  -> device: 1, name: Tesla V100-PCIE-16GB, pci bus id: 0000:06:00.0, compute capability: 7.0\n",
      "2022-10-29 12:04:44.035349: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1525] Created device /device:GPU:2 with 14657 MB memory:  -> device: 2, name: Tesla V100-PCIE-16GB, pci bus id: 0000:08:00.0, compute capability: 7.0\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training on 3 GPUs.\n",
      "Epoch 1/200\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-10-29 12:04:58.611601: I tensorflow/stream_executor/cuda/cuda_dnn.cc:366] Loaded cuDNN version 8202\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "204/204 [==============================] - ETA: 0s - loss: 0.0864 - semantic_0_loss: 0.0808 - semantic_1_loss: 0.0042 - semantic_2_loss: 0.0014\n",
      "Epoch 00001: val_loss improved from inf to 0.01764, saving model to nuclear2_deep_watershed.h5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/app/software/TensorFlow/2.7.1-foss-2021b-CUDA-11.4.1/lib/python3.9/site-packages/keras/engine/functional.py:1410: CustomMaskWarning: Custom mask layers require a config and must override get_config. When loading, the custom mask layer must be passed to the custom_objects argument.\n",
      "  layer_config = serialize_layer_fn(layer)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "204/204 [==============================] - 75s 287ms/step - loss: 0.0864 - semantic_0_loss: 0.0808 - semantic_1_loss: 0.0042 - semantic_2_loss: 0.0014 - val_loss: 0.0176 - val_semantic_0_loss: 0.0079 - val_semantic_1_loss: 0.0064 - val_semantic_2_loss: 0.0033 - lr: 1.0000e-04\n",
      "Epoch 2/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 0.0017 - semantic_0_loss: 5.8954e-04 - semantic_1_loss: 7.0628e-04 - semantic_2_loss: 3.7948e-04\n",
      "Epoch 00002: val_loss improved from 0.01764 to 0.00662, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 274ms/step - loss: 0.0017 - semantic_0_loss: 5.8954e-04 - semantic_1_loss: 7.0628e-04 - semantic_2_loss: 3.7948e-04 - val_loss: 0.0066 - val_semantic_0_loss: 0.0010 - val_semantic_1_loss: 0.0031 - val_semantic_2_loss: 0.0025 - lr: 9.9000e-05\n",
      "Epoch 3/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 0.0012 - semantic_0_loss: 4.1339e-04 - semantic_1_loss: 4.8734e-04 - semantic_2_loss: 2.6869e-04\n",
      "Epoch 00003: val_loss improved from 0.00662 to 0.00308, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 274ms/step - loss: 0.0012 - semantic_0_loss: 4.1339e-04 - semantic_1_loss: 4.8734e-04 - semantic_2_loss: 2.6869e-04 - val_loss: 0.0031 - val_semantic_0_loss: 5.7611e-04 - val_semantic_1_loss: 0.0015 - val_semantic_2_loss: 0.0010 - lr: 9.8010e-05\n",
      "Epoch 4/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 9.5244e-04 - semantic_0_loss: 3.5044e-04 - semantic_1_loss: 3.7558e-04 - semantic_2_loss: 2.2642e-04\n",
      "Epoch 00004: val_loss improved from 0.00308 to 0.00106, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 273ms/step - loss: 9.5244e-04 - semantic_0_loss: 3.5044e-04 - semantic_1_loss: 3.7558e-04 - semantic_2_loss: 2.2642e-04 - val_loss: 0.0011 - val_semantic_0_loss: 3.7881e-04 - val_semantic_1_loss: 4.2813e-04 - val_semantic_2_loss: 2.5198e-04 - lr: 9.7030e-05\n",
      "Epoch 5/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 9.1788e-04 - semantic_0_loss: 3.3381e-04 - semantic_1_loss: 3.6945e-04 - semantic_2_loss: 2.1462e-04\n",
      "Epoch 00005: val_loss improved from 0.00106 to 0.00090, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 273ms/step - loss: 9.1788e-04 - semantic_0_loss: 3.3381e-04 - semantic_1_loss: 3.6945e-04 - semantic_2_loss: 2.1462e-04 - val_loss: 8.9832e-04 - val_semantic_0_loss: 3.4735e-04 - val_semantic_1_loss: 3.3296e-04 - val_semantic_2_loss: 2.1801e-04 - lr: 9.6060e-05\n",
      "Epoch 6/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 9.2890e-04 - semantic_0_loss: 3.3582e-04 - semantic_1_loss: 3.8270e-04 - semantic_2_loss: 2.1038e-04\n",
      "Epoch 00006: val_loss improved from 0.00090 to 0.00082, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 274ms/step - loss: 9.2890e-04 - semantic_0_loss: 3.3582e-04 - semantic_1_loss: 3.8270e-04 - semantic_2_loss: 2.1038e-04 - val_loss: 8.2307e-04 - val_semantic_0_loss: 3.3462e-04 - val_semantic_1_loss: 3.1176e-04 - val_semantic_2_loss: 1.7669e-04 - lr: 9.5099e-05\n",
      "Epoch 7/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 8.7714e-04 - semantic_0_loss: 3.2493e-04 - semantic_1_loss: 3.4938e-04 - semantic_2_loss: 2.0284e-04\n",
      "Epoch 00007: val_loss did not improve from 0.00082\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 8.7714e-04 - semantic_0_loss: 3.2493e-04 - semantic_1_loss: 3.4938e-04 - semantic_2_loss: 2.0284e-04 - val_loss: 8.2456e-04 - val_semantic_0_loss: 3.4992e-04 - val_semantic_1_loss: 3.0539e-04 - val_semantic_2_loss: 1.6925e-04 - lr: 9.4148e-05\n",
      "Epoch 8/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 7.7590e-04 - semantic_0_loss: 2.9161e-04 - semantic_1_loss: 2.9211e-04 - semantic_2_loss: 1.9218e-04\n",
      "Epoch 00008: val_loss did not improve from 0.00082\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 7.7590e-04 - semantic_0_loss: 2.9161e-04 - semantic_1_loss: 2.9211e-04 - semantic_2_loss: 1.9218e-04 - val_loss: 8.4449e-04 - val_semantic_0_loss: 3.3311e-04 - val_semantic_1_loss: 3.3043e-04 - val_semantic_2_loss: 1.8095e-04 - lr: 9.3207e-05\n",
      "Epoch 9/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 7.3562e-04 - semantic_0_loss: 2.7676e-04 - semantic_1_loss: 2.7596e-04 - semantic_2_loss: 1.8290e-04\n",
      "Epoch 00009: val_loss did not improve from 0.00082\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 7.3562e-04 - semantic_0_loss: 2.7676e-04 - semantic_1_loss: 2.7596e-04 - semantic_2_loss: 1.8290e-04 - val_loss: 9.5233e-04 - val_semantic_0_loss: 3.5911e-04 - val_semantic_1_loss: 3.5438e-04 - val_semantic_2_loss: 2.3884e-04 - lr: 9.2274e-05\n",
      "Epoch 10/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 7.9919e-04 - semantic_0_loss: 2.8637e-04 - semantic_1_loss: 3.2073e-04 - semantic_2_loss: 1.9209e-04\n",
      "Epoch 00010: val_loss did not improve from 0.00082\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 7.9919e-04 - semantic_0_loss: 2.8637e-04 - semantic_1_loss: 3.2073e-04 - semantic_2_loss: 1.9209e-04 - val_loss: 8.2587e-04 - val_semantic_0_loss: 3.2902e-04 - val_semantic_1_loss: 3.2648e-04 - val_semantic_2_loss: 1.7036e-04 - lr: 9.1352e-05\n",
      "Epoch 11/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 7.3593e-04 - semantic_0_loss: 2.7706e-04 - semantic_1_loss: 2.7582e-04 - semantic_2_loss: 1.8305e-04\n",
      "Epoch 00011: val_loss did not improve from 0.00082\n",
      "204/204 [==============================] - 50s 244ms/step - loss: 7.3593e-04 - semantic_0_loss: 2.7706e-04 - semantic_1_loss: 2.7582e-04 - semantic_2_loss: 1.8305e-04 - val_loss: 0.0015 - val_semantic_0_loss: 0.0011 - val_semantic_1_loss: 3.1537e-04 - val_semantic_2_loss: 1.5749e-04 - lr: 9.0438e-05\n",
      "Epoch 12/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 6.9907e-04 - semantic_0_loss: 2.6517e-04 - semantic_1_loss: 2.5767e-04 - semantic_2_loss: 1.7623e-04\n",
      "Epoch 00012: val_loss improved from 0.00082 to 0.00079, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 276ms/step - loss: 6.9907e-04 - semantic_0_loss: 2.6517e-04 - semantic_1_loss: 2.5767e-04 - semantic_2_loss: 1.7623e-04 - val_loss: 7.9080e-04 - val_semantic_0_loss: 3.1844e-04 - val_semantic_1_loss: 2.9112e-04 - val_semantic_2_loss: 1.8124e-04 - lr: 8.9534e-05\n",
      "Epoch 13/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 6.6622e-04 - semantic_0_loss: 2.5695e-04 - semantic_1_loss: 2.3843e-04 - semantic_2_loss: 1.7084e-04\n",
      "Epoch 00013: val_loss improved from 0.00079 to 0.00076, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 272ms/step - loss: 6.6622e-04 - semantic_0_loss: 2.5695e-04 - semantic_1_loss: 2.3843e-04 - semantic_2_loss: 1.7084e-04 - val_loss: 7.5816e-04 - val_semantic_0_loss: 3.1579e-04 - val_semantic_1_loss: 2.9051e-04 - val_semantic_2_loss: 1.5186e-04 - lr: 8.8638e-05\n",
      "Epoch 14/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 6.4941e-04 - semantic_0_loss: 2.4683e-04 - semantic_1_loss: 2.3277e-04 - semantic_2_loss: 1.6982e-04\n",
      "Epoch 00014: val_loss did not improve from 0.00076\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 6.4941e-04 - semantic_0_loss: 2.4683e-04 - semantic_1_loss: 2.3277e-04 - semantic_2_loss: 1.6982e-04 - val_loss: 0.0012 - val_semantic_0_loss: 2.9648e-04 - val_semantic_1_loss: 3.0326e-04 - val_semantic_2_loss: 6.4989e-04 - lr: 8.7752e-05\n",
      "Epoch 15/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 6.4539e-04 - semantic_0_loss: 2.4506e-04 - semantic_1_loss: 2.3023e-04 - semantic_2_loss: 1.7010e-04\n",
      "Epoch 00015: val_loss improved from 0.00076 to 0.00071, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 274ms/step - loss: 6.4539e-04 - semantic_0_loss: 2.4506e-04 - semantic_1_loss: 2.3023e-04 - semantic_2_loss: 1.7010e-04 - val_loss: 7.1457e-04 - val_semantic_0_loss: 2.9968e-04 - val_semantic_1_loss: 2.6069e-04 - val_semantic_2_loss: 1.5420e-04 - lr: 8.6875e-05\n",
      "Epoch 16/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 6.4380e-04 - semantic_0_loss: 2.4555e-04 - semantic_1_loss: 2.2779e-04 - semantic_2_loss: 1.7047e-04\n",
      "Epoch 00016: val_loss did not improve from 0.00071\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 6.4380e-04 - semantic_0_loss: 2.4555e-04 - semantic_1_loss: 2.2779e-04 - semantic_2_loss: 1.7047e-04 - val_loss: 0.0023 - val_semantic_0_loss: 0.0019 - val_semantic_1_loss: 2.5495e-04 - val_semantic_2_loss: 1.4542e-04 - lr: 8.6006e-05\n",
      "Epoch 17/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 6.3147e-04 - semantic_0_loss: 2.3746e-04 - semantic_1_loss: 2.2528e-04 - semantic_2_loss: 1.6873e-04\n",
      "Epoch 00017: val_loss did not improve from 0.00071\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 6.3147e-04 - semantic_0_loss: 2.3746e-04 - semantic_1_loss: 2.2528e-04 - semantic_2_loss: 1.6873e-04 - val_loss: 7.8511e-04 - val_semantic_0_loss: 3.1246e-04 - val_semantic_1_loss: 3.1471e-04 - val_semantic_2_loss: 1.5794e-04 - lr: 8.5146e-05\n",
      "Epoch 18/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 6.2325e-04 - semantic_0_loss: 2.3839e-04 - semantic_1_loss: 2.2003e-04 - semantic_2_loss: 1.6483e-04\n",
      "Epoch 00018: val_loss did not improve from 0.00071\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 6.2325e-04 - semantic_0_loss: 2.3839e-04 - semantic_1_loss: 2.2003e-04 - semantic_2_loss: 1.6483e-04 - val_loss: 7.9086e-04 - val_semantic_0_loss: 2.9547e-04 - val_semantic_1_loss: 3.5399e-04 - val_semantic_2_loss: 1.4141e-04 - lr: 8.4294e-05\n",
      "Epoch 19/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 6.0029e-04 - semantic_0_loss: 2.2667e-04 - semantic_1_loss: 2.1115e-04 - semantic_2_loss: 1.6247e-04\n",
      "Epoch 00019: val_loss did not improve from 0.00071\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 6.0029e-04 - semantic_0_loss: 2.2667e-04 - semantic_1_loss: 2.1115e-04 - semantic_2_loss: 1.6247e-04 - val_loss: 7.2137e-04 - val_semantic_0_loss: 3.3279e-04 - val_semantic_1_loss: 2.4808e-04 - val_semantic_2_loss: 1.4050e-04 - lr: 8.3451e-05\n",
      "Epoch 20/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 6.0062e-04 - semantic_0_loss: 2.2545e-04 - semantic_1_loss: 2.1048e-04 - semantic_2_loss: 1.6469e-04\n",
      "Epoch 00020: val_loss did not improve from 0.00071\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 6.0062e-04 - semantic_0_loss: 2.2545e-04 - semantic_1_loss: 2.1048e-04 - semantic_2_loss: 1.6469e-04 - val_loss: 7.3566e-04 - val_semantic_0_loss: 3.1759e-04 - val_semantic_1_loss: 2.7808e-04 - val_semantic_2_loss: 1.4000e-04 - lr: 8.2617e-05\n",
      "Epoch 21/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 6.1698e-04 - semantic_0_loss: 2.3056e-04 - semantic_1_loss: 2.1620e-04 - semantic_2_loss: 1.7022e-04\n",
      "Epoch 00021: val_loss did not improve from 0.00071\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 6.1698e-04 - semantic_0_loss: 2.3056e-04 - semantic_1_loss: 2.1620e-04 - semantic_2_loss: 1.7022e-04 - val_loss: 8.6585e-04 - val_semantic_0_loss: 3.0517e-04 - val_semantic_1_loss: 2.9401e-04 - val_semantic_2_loss: 2.6667e-04 - lr: 8.1791e-05\n",
      "Epoch 22/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 5.6617e-04 - semantic_0_loss: 2.1845e-04 - semantic_1_loss: 1.8905e-04 - semantic_2_loss: 1.5868e-04\n",
      "Epoch 00022: val_loss improved from 0.00071 to 0.00071, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 273ms/step - loss: 5.6617e-04 - semantic_0_loss: 2.1845e-04 - semantic_1_loss: 1.8905e-04 - semantic_2_loss: 1.5868e-04 - val_loss: 7.0658e-04 - val_semantic_0_loss: 3.1163e-04 - val_semantic_1_loss: 2.4029e-04 - val_semantic_2_loss: 1.5467e-04 - lr: 8.0973e-05\n",
      "Epoch 23/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 5.6878e-04 - semantic_0_loss: 2.1642e-04 - semantic_1_loss: 1.9453e-04 - semantic_2_loss: 1.5783e-04\n",
      "Epoch 00023: val_loss improved from 0.00071 to 0.00070, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 276ms/step - loss: 5.6878e-04 - semantic_0_loss: 2.1642e-04 - semantic_1_loss: 1.9453e-04 - semantic_2_loss: 1.5783e-04 - val_loss: 7.0202e-04 - val_semantic_0_loss: 3.1846e-04 - val_semantic_1_loss: 2.4602e-04 - val_semantic_2_loss: 1.3754e-04 - lr: 8.0163e-05\n",
      "Epoch 24/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 5.6174e-04 - semantic_0_loss: 2.1395e-04 - semantic_1_loss: 1.9103e-04 - semantic_2_loss: 1.5675e-04\n",
      "Epoch 00024: val_loss did not improve from 0.00070\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 5.6174e-04 - semantic_0_loss: 2.1395e-04 - semantic_1_loss: 1.9103e-04 - semantic_2_loss: 1.5675e-04 - val_loss: 7.0954e-04 - val_semantic_0_loss: 3.1842e-04 - val_semantic_1_loss: 2.5789e-04 - val_semantic_2_loss: 1.3322e-04 - lr: 7.9361e-05\n",
      "Epoch 25/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 5.4241e-04 - semantic_0_loss: 2.0869e-04 - semantic_1_loss: 1.7925e-04 - semantic_2_loss: 1.5446e-04\n",
      "Epoch 00025: val_loss did not improve from 0.00070\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 5.4241e-04 - semantic_0_loss: 2.0869e-04 - semantic_1_loss: 1.7925e-04 - semantic_2_loss: 1.5446e-04 - val_loss: 7.4559e-04 - val_semantic_0_loss: 3.0947e-04 - val_semantic_1_loss: 2.4436e-04 - val_semantic_2_loss: 1.9176e-04 - lr: 7.8568e-05\n",
      "Epoch 26/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 5.4970e-04 - semantic_0_loss: 2.0911e-04 - semantic_1_loss: 1.8913e-04 - semantic_2_loss: 1.5146e-04\n",
      "Epoch 00026: val_loss improved from 0.00070 to 0.00068, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 273ms/step - loss: 5.4970e-04 - semantic_0_loss: 2.0911e-04 - semantic_1_loss: 1.8913e-04 - semantic_2_loss: 1.5146e-04 - val_loss: 6.7560e-04 - val_semantic_0_loss: 2.9042e-04 - val_semantic_1_loss: 2.5404e-04 - val_semantic_2_loss: 1.3114e-04 - lr: 7.7782e-05\n",
      "Epoch 27/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 5.3641e-04 - semantic_0_loss: 2.0693e-04 - semantic_1_loss: 1.7861e-04 - semantic_2_loss: 1.5086e-04\n",
      "Epoch 00027: val_loss improved from 0.00068 to 0.00067, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 275ms/step - loss: 5.3641e-04 - semantic_0_loss: 2.0693e-04 - semantic_1_loss: 1.7861e-04 - semantic_2_loss: 1.5086e-04 - val_loss: 6.7105e-04 - val_semantic_0_loss: 3.0141e-04 - val_semantic_1_loss: 2.3446e-04 - val_semantic_2_loss: 1.3518e-04 - lr: 7.7004e-05\n",
      "Epoch 28/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 5.1374e-04 - semantic_0_loss: 1.9493e-04 - semantic_1_loss: 1.6956e-04 - semantic_2_loss: 1.4925e-04\n",
      "Epoch 00028: val_loss did not improve from 0.00067\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 5.1374e-04 - semantic_0_loss: 1.9493e-04 - semantic_1_loss: 1.6956e-04 - semantic_2_loss: 1.4925e-04 - val_loss: 6.8244e-04 - val_semantic_0_loss: 3.0896e-04 - val_semantic_1_loss: 2.3853e-04 - val_semantic_2_loss: 1.3495e-04 - lr: 7.6234e-05\n",
      "Epoch 29/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 5.2626e-04 - semantic_0_loss: 2.0043e-04 - semantic_1_loss: 1.7471e-04 - semantic_2_loss: 1.5112e-04\n",
      "Epoch 00029: val_loss did not improve from 0.00067\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 5.2626e-04 - semantic_0_loss: 2.0043e-04 - semantic_1_loss: 1.7471e-04 - semantic_2_loss: 1.5112e-04 - val_loss: 6.8991e-04 - val_semantic_0_loss: 3.1912e-04 - val_semantic_1_loss: 2.3965e-04 - val_semantic_2_loss: 1.3114e-04 - lr: 7.5472e-05\n",
      "Epoch 30/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 5.2897e-04 - semantic_0_loss: 2.0258e-04 - semantic_1_loss: 1.7644e-04 - semantic_2_loss: 1.4994e-04\n",
      "Epoch 00030: val_loss did not improve from 0.00067\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 5.2897e-04 - semantic_0_loss: 2.0258e-04 - semantic_1_loss: 1.7644e-04 - semantic_2_loss: 1.4994e-04 - val_loss: 6.7700e-04 - val_semantic_0_loss: 3.0666e-04 - val_semantic_1_loss: 2.3795e-04 - val_semantic_2_loss: 1.3239e-04 - lr: 7.4717e-05\n",
      "Epoch 31/200\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "204/204 [==============================] - ETA: 0s - loss: 5.0417e-04 - semantic_0_loss: 1.9399e-04 - semantic_1_loss: 1.6356e-04 - semantic_2_loss: 1.4661e-04\n",
      "Epoch 00031: val_loss improved from 0.00067 to 0.00067, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 273ms/step - loss: 5.0417e-04 - semantic_0_loss: 1.9399e-04 - semantic_1_loss: 1.6356e-04 - semantic_2_loss: 1.4661e-04 - val_loss: 6.7097e-04 - val_semantic_0_loss: 2.9143e-04 - val_semantic_1_loss: 2.4514e-04 - val_semantic_2_loss: 1.3439e-04 - lr: 7.3970e-05\n",
      "Epoch 32/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.8889e-04 - semantic_0_loss: 1.8822e-04 - semantic_1_loss: 1.5522e-04 - semantic_2_loss: 1.4545e-04\n",
      "Epoch 00032: val_loss improved from 0.00067 to 0.00067, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 273ms/step - loss: 4.8889e-04 - semantic_0_loss: 1.8822e-04 - semantic_1_loss: 1.5522e-04 - semantic_2_loss: 1.4545e-04 - val_loss: 6.6889e-04 - val_semantic_0_loss: 3.0579e-04 - val_semantic_1_loss: 2.2785e-04 - val_semantic_2_loss: 1.3525e-04 - lr: 7.3230e-05\n",
      "Epoch 33/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.8927e-04 - semantic_0_loss: 1.8352e-04 - semantic_1_loss: 1.6082e-04 - semantic_2_loss: 1.4493e-04\n",
      "Epoch 00033: val_loss did not improve from 0.00067\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 4.8927e-04 - semantic_0_loss: 1.8352e-04 - semantic_1_loss: 1.6082e-04 - semantic_2_loss: 1.4493e-04 - val_loss: 6.7691e-04 - val_semantic_0_loss: 2.9857e-04 - val_semantic_1_loss: 2.3528e-04 - val_semantic_2_loss: 1.4306e-04 - lr: 7.2498e-05\n",
      "Epoch 34/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.9240e-04 - semantic_0_loss: 1.8600e-04 - semantic_1_loss: 1.6190e-04 - semantic_2_loss: 1.4450e-04\n",
      "Epoch 00034: val_loss improved from 0.00067 to 0.00065, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 273ms/step - loss: 4.9240e-04 - semantic_0_loss: 1.8600e-04 - semantic_1_loss: 1.6190e-04 - semantic_2_loss: 1.4450e-04 - val_loss: 6.5284e-04 - val_semantic_0_loss: 3.0041e-04 - val_semantic_1_loss: 2.2658e-04 - val_semantic_2_loss: 1.2584e-04 - lr: 7.1773e-05\n",
      "Epoch 35/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.8665e-04 - semantic_0_loss: 1.8199e-04 - semantic_1_loss: 1.5931e-04 - semantic_2_loss: 1.4535e-04\n",
      "Epoch 00035: val_loss did not improve from 0.00065\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 4.8665e-04 - semantic_0_loss: 1.8199e-04 - semantic_1_loss: 1.5931e-04 - semantic_2_loss: 1.4535e-04 - val_loss: 6.7691e-04 - val_semantic_0_loss: 3.0593e-04 - val_semantic_1_loss: 2.3601e-04 - val_semantic_2_loss: 1.3496e-04 - lr: 7.1055e-05\n",
      "Epoch 36/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.7152e-04 - semantic_0_loss: 1.7676e-04 - semantic_1_loss: 1.5103e-04 - semantic_2_loss: 1.4373e-04\n",
      "Epoch 00036: val_loss improved from 0.00065 to 0.00064, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 57s 278ms/step - loss: 4.7152e-04 - semantic_0_loss: 1.7676e-04 - semantic_1_loss: 1.5103e-04 - semantic_2_loss: 1.4373e-04 - val_loss: 6.4385e-04 - val_semantic_0_loss: 2.9127e-04 - val_semantic_1_loss: 2.2931e-04 - val_semantic_2_loss: 1.2327e-04 - lr: 7.0345e-05\n",
      "Epoch 37/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.7783e-04 - semantic_0_loss: 1.7908e-04 - semantic_1_loss: 1.5530e-04 - semantic_2_loss: 1.4346e-04\n",
      "Epoch 00037: val_loss did not improve from 0.00064\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 4.7783e-04 - semantic_0_loss: 1.7908e-04 - semantic_1_loss: 1.5530e-04 - semantic_2_loss: 1.4346e-04 - val_loss: 6.6230e-04 - val_semantic_0_loss: 3.1876e-04 - val_semantic_1_loss: 2.1605e-04 - val_semantic_2_loss: 1.2749e-04 - lr: 6.9641e-05\n",
      "Epoch 38/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.7628e-04 - semantic_0_loss: 1.7668e-04 - semantic_1_loss: 1.5733e-04 - semantic_2_loss: 1.4227e-04\n",
      "Epoch 00038: val_loss did not improve from 0.00064\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 4.7628e-04 - semantic_0_loss: 1.7668e-04 - semantic_1_loss: 1.5733e-04 - semantic_2_loss: 1.4227e-04 - val_loss: 6.5677e-04 - val_semantic_0_loss: 3.0664e-04 - val_semantic_1_loss: 2.2822e-04 - val_semantic_2_loss: 1.2190e-04 - lr: 6.8945e-05\n",
      "Epoch 39/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.7901e-04 - semantic_0_loss: 1.7842e-04 - semantic_1_loss: 1.5662e-04 - semantic_2_loss: 1.4398e-04\n",
      "Epoch 00039: val_loss did not improve from 0.00064\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 4.7901e-04 - semantic_0_loss: 1.7842e-04 - semantic_1_loss: 1.5662e-04 - semantic_2_loss: 1.4398e-04 - val_loss: 6.5481e-04 - val_semantic_0_loss: 3.0334e-04 - val_semantic_1_loss: 2.2532e-04 - val_semantic_2_loss: 1.2615e-04 - lr: 6.8255e-05\n",
      "Epoch 40/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.6151e-04 - semantic_0_loss: 1.6762e-04 - semantic_1_loss: 1.4966e-04 - semantic_2_loss: 1.4422e-04\n",
      "Epoch 00040: val_loss improved from 0.00064 to 0.00062, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 274ms/step - loss: 4.6151e-04 - semantic_0_loss: 1.6762e-04 - semantic_1_loss: 1.4966e-04 - semantic_2_loss: 1.4422e-04 - val_loss: 6.2476e-04 - val_semantic_0_loss: 2.9236e-04 - val_semantic_1_loss: 2.1035e-04 - val_semantic_2_loss: 1.2205e-04 - lr: 6.7573e-05\n",
      "Epoch 41/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.5472e-04 - semantic_0_loss: 1.6967e-04 - semantic_1_loss: 1.4501e-04 - semantic_2_loss: 1.4004e-04\n",
      "Epoch 00041: val_loss improved from 0.00062 to 0.00062, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 57s 277ms/step - loss: 4.5472e-04 - semantic_0_loss: 1.6967e-04 - semantic_1_loss: 1.4501e-04 - semantic_2_loss: 1.4004e-04 - val_loss: 6.1955e-04 - val_semantic_0_loss: 2.7852e-04 - val_semantic_1_loss: 2.1819e-04 - val_semantic_2_loss: 1.2285e-04 - lr: 6.6897e-05\n",
      "Epoch 42/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.4477e-04 - semantic_0_loss: 1.6415e-04 - semantic_1_loss: 1.4221e-04 - semantic_2_loss: 1.3841e-04\n",
      "Epoch 00042: val_loss improved from 0.00062 to 0.00062, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 57s 277ms/step - loss: 4.4477e-04 - semantic_0_loss: 1.6415e-04 - semantic_1_loss: 1.4221e-04 - semantic_2_loss: 1.3841e-04 - val_loss: 6.1605e-04 - val_semantic_0_loss: 2.9103e-04 - val_semantic_1_loss: 2.1057e-04 - val_semantic_2_loss: 1.1445e-04 - lr: 6.6228e-05\n",
      "Epoch 43/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.5506e-04 - semantic_0_loss: 1.7024e-04 - semantic_1_loss: 1.4457e-04 - semantic_2_loss: 1.4026e-04\n",
      "Epoch 00043: val_loss improved from 0.00062 to 0.00060, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 272ms/step - loss: 4.5506e-04 - semantic_0_loss: 1.7024e-04 - semantic_1_loss: 1.4457e-04 - semantic_2_loss: 1.4026e-04 - val_loss: 5.9726e-04 - val_semantic_0_loss: 2.7595e-04 - val_semantic_1_loss: 2.0439e-04 - val_semantic_2_loss: 1.1692e-04 - lr: 6.5566e-05\n",
      "Epoch 44/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.5551e-04 - semantic_0_loss: 1.7073e-04 - semantic_1_loss: 1.4536e-04 - semantic_2_loss: 1.3942e-04\n",
      "Epoch 00044: val_loss did not improve from 0.00060\n",
      "204/204 [==============================] - 51s 250ms/step - loss: 4.5551e-04 - semantic_0_loss: 1.7073e-04 - semantic_1_loss: 1.4536e-04 - semantic_2_loss: 1.3942e-04 - val_loss: 6.3128e-04 - val_semantic_0_loss: 2.9218e-04 - val_semantic_1_loss: 2.1807e-04 - val_semantic_2_loss: 1.2103e-04 - lr: 6.4910e-05\n",
      "Epoch 45/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.4179e-04 - semantic_0_loss: 1.6152e-04 - semantic_1_loss: 1.4183e-04 - semantic_2_loss: 1.3844e-04\n",
      "Epoch 00045: val_loss did not improve from 0.00060\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 4.4179e-04 - semantic_0_loss: 1.6152e-04 - semantic_1_loss: 1.4183e-04 - semantic_2_loss: 1.3844e-04 - val_loss: 6.2705e-04 - val_semantic_0_loss: 2.9578e-04 - val_semantic_1_loss: 2.1270e-04 - val_semantic_2_loss: 1.1856e-04 - lr: 6.4261e-05\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 46/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.3875e-04 - semantic_0_loss: 1.5984e-04 - semantic_1_loss: 1.3917e-04 - semantic_2_loss: 1.3974e-04\n",
      "Epoch 00046: val_loss did not improve from 0.00060\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 4.3875e-04 - semantic_0_loss: 1.5984e-04 - semantic_1_loss: 1.3917e-04 - semantic_2_loss: 1.3974e-04 - val_loss: 6.0717e-04 - val_semantic_0_loss: 2.7065e-04 - val_semantic_1_loss: 2.1442e-04 - val_semantic_2_loss: 1.2211e-04 - lr: 6.3619e-05\n",
      "Epoch 47/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.3684e-04 - semantic_0_loss: 1.6083e-04 - semantic_1_loss: 1.3922e-04 - semantic_2_loss: 1.3679e-04\n",
      "Epoch 00047: val_loss did not improve from 0.00060\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 4.3684e-04 - semantic_0_loss: 1.6083e-04 - semantic_1_loss: 1.3922e-04 - semantic_2_loss: 1.3679e-04 - val_loss: 6.2405e-04 - val_semantic_0_loss: 2.8979e-04 - val_semantic_1_loss: 2.1489e-04 - val_semantic_2_loss: 1.1937e-04 - lr: 6.2982e-05\n",
      "Epoch 48/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.3175e-04 - semantic_0_loss: 1.5507e-04 - semantic_1_loss: 1.3986e-04 - semantic_2_loss: 1.3682e-04\n",
      "Epoch 00048: val_loss did not improve from 0.00060\n",
      "204/204 [==============================] - 50s 244ms/step - loss: 4.3175e-04 - semantic_0_loss: 1.5507e-04 - semantic_1_loss: 1.3986e-04 - semantic_2_loss: 1.3682e-04 - val_loss: 6.6090e-04 - val_semantic_0_loss: 2.8845e-04 - val_semantic_1_loss: 2.3303e-04 - val_semantic_2_loss: 1.3942e-04 - lr: 6.2353e-05\n",
      "Epoch 49/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.2258e-04 - semantic_0_loss: 1.5213e-04 - semantic_1_loss: 1.3356e-04 - semantic_2_loss: 1.3689e-04\n",
      "Epoch 00049: val_loss did not improve from 0.00060\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 4.2258e-04 - semantic_0_loss: 1.5213e-04 - semantic_1_loss: 1.3356e-04 - semantic_2_loss: 1.3689e-04 - val_loss: 6.1131e-04 - val_semantic_0_loss: 2.8388e-04 - val_semantic_1_loss: 2.1275e-04 - val_semantic_2_loss: 1.1468e-04 - lr: 6.1729e-05\n",
      "Epoch 50/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.2703e-04 - semantic_0_loss: 1.5350e-04 - semantic_1_loss: 1.3583e-04 - semantic_2_loss: 1.3770e-04\n",
      "Epoch 00050: val_loss did not improve from 0.00060\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 4.2703e-04 - semantic_0_loss: 1.5350e-04 - semantic_1_loss: 1.3583e-04 - semantic_2_loss: 1.3770e-04 - val_loss: 6.1281e-04 - val_semantic_0_loss: 2.8608e-04 - val_semantic_1_loss: 2.0890e-04 - val_semantic_2_loss: 1.1783e-04 - lr: 6.1112e-05\n",
      "Epoch 51/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.0505e-04 - semantic_0_loss: 1.4511e-04 - semantic_1_loss: 1.2670e-04 - semantic_2_loss: 1.3324e-04\n",
      "Epoch 00051: val_loss improved from 0.00060 to 0.00059, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 276ms/step - loss: 4.0505e-04 - semantic_0_loss: 1.4511e-04 - semantic_1_loss: 1.2670e-04 - semantic_2_loss: 1.3324e-04 - val_loss: 5.8988e-04 - val_semantic_0_loss: 2.7160e-04 - val_semantic_1_loss: 2.0437e-04 - val_semantic_2_loss: 1.1391e-04 - lr: 6.0501e-05\n",
      "Epoch 52/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.0834e-04 - semantic_0_loss: 1.4193e-04 - semantic_1_loss: 1.3229e-04 - semantic_2_loss: 1.3412e-04\n",
      "Epoch 00052: val_loss did not improve from 0.00059\n",
      "204/204 [==============================] - 51s 250ms/step - loss: 4.0834e-04 - semantic_0_loss: 1.4193e-04 - semantic_1_loss: 1.3229e-04 - semantic_2_loss: 1.3412e-04 - val_loss: 6.1223e-04 - val_semantic_0_loss: 2.8074e-04 - val_semantic_1_loss: 2.1513e-04 - val_semantic_2_loss: 1.1635e-04 - lr: 5.9896e-05\n",
      "Epoch 53/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.0942e-04 - semantic_0_loss: 1.4141e-04 - semantic_1_loss: 1.3271e-04 - semantic_2_loss: 1.3530e-04\n",
      "Epoch 00053: val_loss improved from 0.00059 to 0.00058, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 276ms/step - loss: 4.0942e-04 - semantic_0_loss: 1.4141e-04 - semantic_1_loss: 1.3271e-04 - semantic_2_loss: 1.3530e-04 - val_loss: 5.8137e-04 - val_semantic_0_loss: 2.6795e-04 - val_semantic_1_loss: 2.0119e-04 - val_semantic_2_loss: 1.1223e-04 - lr: 5.9297e-05\n",
      "Epoch 54/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.0711e-04 - semantic_0_loss: 1.4517e-04 - semantic_1_loss: 1.2906e-04 - semantic_2_loss: 1.3288e-04\n",
      "Epoch 00054: val_loss did not improve from 0.00058\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 4.0711e-04 - semantic_0_loss: 1.4517e-04 - semantic_1_loss: 1.2906e-04 - semantic_2_loss: 1.3288e-04 - val_loss: 5.8244e-04 - val_semantic_0_loss: 2.6316e-04 - val_semantic_1_loss: 2.0464e-04 - val_semantic_2_loss: 1.1464e-04 - lr: 5.8704e-05\n",
      "Epoch 55/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.0007e-04 - semantic_0_loss: 1.3779e-04 - semantic_1_loss: 1.2858e-04 - semantic_2_loss: 1.3370e-04\n",
      "Epoch 00055: val_loss did not improve from 0.00058\n",
      "204/204 [==============================] - 51s 251ms/step - loss: 4.0007e-04 - semantic_0_loss: 1.3779e-04 - semantic_1_loss: 1.2858e-04 - semantic_2_loss: 1.3370e-04 - val_loss: 6.1847e-04 - val_semantic_0_loss: 2.9845e-04 - val_semantic_1_loss: 2.0671e-04 - val_semantic_2_loss: 1.1331e-04 - lr: 5.8117e-05\n",
      "Epoch 56/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.9856e-04 - semantic_0_loss: 1.4140e-04 - semantic_1_loss: 1.2541e-04 - semantic_2_loss: 1.3174e-04\n",
      "Epoch 00056: val_loss did not improve from 0.00058\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.9856e-04 - semantic_0_loss: 1.4140e-04 - semantic_1_loss: 1.2541e-04 - semantic_2_loss: 1.3174e-04 - val_loss: 6.2523e-04 - val_semantic_0_loss: 3.0281e-04 - val_semantic_1_loss: 2.0645e-04 - val_semantic_2_loss: 1.1598e-04 - lr: 5.7535e-05\n",
      "Epoch 57/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.0264e-04 - semantic_0_loss: 1.4171e-04 - semantic_1_loss: 1.2859e-04 - semantic_2_loss: 1.3234e-04\n",
      "Epoch 00057: val_loss improved from 0.00058 to 0.00057, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 273ms/step - loss: 4.0264e-04 - semantic_0_loss: 1.4171e-04 - semantic_1_loss: 1.2859e-04 - semantic_2_loss: 1.3234e-04 - val_loss: 5.6965e-04 - val_semantic_0_loss: 2.6157e-04 - val_semantic_1_loss: 1.9055e-04 - val_semantic_2_loss: 1.1753e-04 - lr: 5.6960e-05\n",
      "Epoch 58/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 4.0269e-04 - semantic_0_loss: 1.4182e-04 - semantic_1_loss: 1.2776e-04 - semantic_2_loss: 1.3311e-04\n",
      "Epoch 00058: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 4.0269e-04 - semantic_0_loss: 1.4182e-04 - semantic_1_loss: 1.2776e-04 - semantic_2_loss: 1.3311e-04 - val_loss: 5.8283e-04 - val_semantic_0_loss: 2.7092e-04 - val_semantic_1_loss: 1.9555e-04 - val_semantic_2_loss: 1.1636e-04 - lr: 5.6391e-05\n",
      "Epoch 59/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.9326e-04 - semantic_0_loss: 1.3852e-04 - semantic_1_loss: 1.2516e-04 - semantic_2_loss: 1.2958e-04\n",
      "Epoch 00059: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 250ms/step - loss: 3.9326e-04 - semantic_0_loss: 1.3852e-04 - semantic_1_loss: 1.2516e-04 - semantic_2_loss: 1.2958e-04 - val_loss: 5.9422e-04 - val_semantic_0_loss: 2.6614e-04 - val_semantic_1_loss: 2.0997e-04 - val_semantic_2_loss: 1.1811e-04 - lr: 5.5827e-05\n",
      "Epoch 60/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.8305e-04 - semantic_0_loss: 1.3319e-04 - semantic_1_loss: 1.2108e-04 - semantic_2_loss: 1.2877e-04\n",
      "Epoch 00060: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.8305e-04 - semantic_0_loss: 1.3319e-04 - semantic_1_loss: 1.2108e-04 - semantic_2_loss: 1.2877e-04 - val_loss: 6.1529e-04 - val_semantic_0_loss: 2.9196e-04 - val_semantic_1_loss: 2.0488e-04 - val_semantic_2_loss: 1.1845e-04 - lr: 5.5268e-05\n",
      "Epoch 61/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.9138e-04 - semantic_0_loss: 1.3254e-04 - semantic_1_loss: 1.2815e-04 - semantic_2_loss: 1.3069e-04\n",
      "Epoch 00061: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.9138e-04 - semantic_0_loss: 1.3254e-04 - semantic_1_loss: 1.2815e-04 - semantic_2_loss: 1.3069e-04 - val_loss: 6.0653e-04 - val_semantic_0_loss: 2.9437e-04 - val_semantic_1_loss: 1.9848e-04 - val_semantic_2_loss: 1.1368e-04 - lr: 5.4716e-05\n",
      "Epoch 62/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.8674e-04 - semantic_0_loss: 1.3317e-04 - semantic_1_loss: 1.2313e-04 - semantic_2_loss: 1.3044e-04\n",
      "Epoch 00062: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.8674e-04 - semantic_0_loss: 1.3317e-04 - semantic_1_loss: 1.2313e-04 - semantic_2_loss: 1.3044e-04 - val_loss: 5.8807e-04 - val_semantic_0_loss: 2.6703e-04 - val_semantic_1_loss: 2.0527e-04 - val_semantic_2_loss: 1.1577e-04 - lr: 5.4169e-05\n",
      "Epoch 63/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.7431e-04 - semantic_0_loss: 1.2781e-04 - semantic_1_loss: 1.1711e-04 - semantic_2_loss: 1.2939e-04\n",
      "Epoch 00063: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.7431e-04 - semantic_0_loss: 1.2781e-04 - semantic_1_loss: 1.1711e-04 - semantic_2_loss: 1.2939e-04 - val_loss: 5.8428e-04 - val_semantic_0_loss: 2.7068e-04 - val_semantic_1_loss: 1.9998e-04 - val_semantic_2_loss: 1.1362e-04 - lr: 5.3627e-05\n",
      "Epoch 64/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.7272e-04 - semantic_0_loss: 1.2467e-04 - semantic_1_loss: 1.1888e-04 - semantic_2_loss: 1.2917e-04\n",
      "Epoch 00064: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 3.7272e-04 - semantic_0_loss: 1.2467e-04 - semantic_1_loss: 1.1888e-04 - semantic_2_loss: 1.2917e-04 - val_loss: 6.2092e-04 - val_semantic_0_loss: 3.0026e-04 - val_semantic_1_loss: 2.0370e-04 - val_semantic_2_loss: 1.1696e-04 - lr: 5.3091e-05\n",
      "Epoch 65/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.8331e-04 - semantic_0_loss: 1.2918e-04 - semantic_1_loss: 1.2449e-04 - semantic_2_loss: 1.2964e-04\n",
      "Epoch 00065: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.8331e-04 - semantic_0_loss: 1.2918e-04 - semantic_1_loss: 1.2449e-04 - semantic_2_loss: 1.2964e-04 - val_loss: 6.0503e-04 - val_semantic_0_loss: 2.7867e-04 - val_semantic_1_loss: 2.1022e-04 - val_semantic_2_loss: 1.1614e-04 - lr: 5.2560e-05\n",
      "Epoch 66/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.6666e-04 - semantic_0_loss: 1.2159e-04 - semantic_1_loss: 1.1715e-04 - semantic_2_loss: 1.2791e-04\n",
      "Epoch 00066: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 3.6666e-04 - semantic_0_loss: 1.2159e-04 - semantic_1_loss: 1.1715e-04 - semantic_2_loss: 1.2791e-04 - val_loss: 5.7602e-04 - val_semantic_0_loss: 2.6889e-04 - val_semantic_1_loss: 1.9584e-04 - val_semantic_2_loss: 1.1129e-04 - lr: 5.2034e-05\n",
      "Epoch 67/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.6677e-04 - semantic_0_loss: 1.2036e-04 - semantic_1_loss: 1.1777e-04 - semantic_2_loss: 1.2864e-04\n",
      "Epoch 00067: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.6677e-04 - semantic_0_loss: 1.2036e-04 - semantic_1_loss: 1.1777e-04 - semantic_2_loss: 1.2864e-04 - val_loss: 6.0053e-04 - val_semantic_0_loss: 2.8587e-04 - val_semantic_1_loss: 1.9983e-04 - val_semantic_2_loss: 1.1483e-04 - lr: 5.1514e-05\n",
      "Epoch 68/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.8747e-04 - semantic_0_loss: 1.2733e-04 - semantic_1_loss: 1.2349e-04 - semantic_2_loss: 1.3664e-04\n",
      "Epoch 00068: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.8747e-04 - semantic_0_loss: 1.2733e-04 - semantic_1_loss: 1.2349e-04 - semantic_2_loss: 1.3664e-04 - val_loss: 5.9817e-04 - val_semantic_0_loss: 2.7987e-04 - val_semantic_1_loss: 2.0188e-04 - val_semantic_2_loss: 1.1642e-04 - lr: 5.0999e-05\n",
      "Epoch 69/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.8012e-04 - semantic_0_loss: 1.2532e-04 - semantic_1_loss: 1.2412e-04 - semantic_2_loss: 1.3068e-04\n",
      "Epoch 00069: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 3.8012e-04 - semantic_0_loss: 1.2532e-04 - semantic_1_loss: 1.2412e-04 - semantic_2_loss: 1.3068e-04 - val_loss: 6.4163e-04 - val_semantic_0_loss: 3.3046e-04 - val_semantic_1_loss: 1.9946e-04 - val_semantic_2_loss: 1.1171e-04 - lr: 5.0489e-05\n",
      "Epoch 70/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.6854e-04 - semantic_0_loss: 1.1967e-04 - semantic_1_loss: 1.2092e-04 - semantic_2_loss: 1.2795e-04\n",
      "Epoch 00070: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 250ms/step - loss: 3.6854e-04 - semantic_0_loss: 1.1967e-04 - semantic_1_loss: 1.2092e-04 - semantic_2_loss: 1.2795e-04 - val_loss: 6.2048e-04 - val_semantic_0_loss: 2.9764e-04 - val_semantic_1_loss: 2.0770e-04 - val_semantic_2_loss: 1.1514e-04 - lr: 4.9984e-05\n",
      "Epoch 71/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.6891e-04 - semantic_0_loss: 1.1829e-04 - semantic_1_loss: 1.2189e-04 - semantic_2_loss: 1.2874e-04\n",
      "Epoch 00071: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.6891e-04 - semantic_0_loss: 1.1829e-04 - semantic_1_loss: 1.2189e-04 - semantic_2_loss: 1.2874e-04 - val_loss: 6.1436e-04 - val_semantic_0_loss: 3.0481e-04 - val_semantic_1_loss: 1.9558e-04 - val_semantic_2_loss: 1.1397e-04 - lr: 4.9484e-05\n",
      "Epoch 72/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.6003e-04 - semantic_0_loss: 1.1780e-04 - semantic_1_loss: 1.1509e-04 - semantic_2_loss: 1.2714e-04\n",
      "Epoch 00072: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.6003e-04 - semantic_0_loss: 1.1780e-04 - semantic_1_loss: 1.1509e-04 - semantic_2_loss: 1.2714e-04 - val_loss: 6.2286e-04 - val_semantic_0_loss: 3.0013e-04 - val_semantic_1_loss: 2.0934e-04 - val_semantic_2_loss: 1.1339e-04 - lr: 4.8989e-05\n",
      "Epoch 73/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.5842e-04 - semantic_0_loss: 1.1800e-04 - semantic_1_loss: 1.1465e-04 - semantic_2_loss: 1.2577e-04\n",
      "Epoch 00073: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 3.5842e-04 - semantic_0_loss: 1.1800e-04 - semantic_1_loss: 1.1465e-04 - semantic_2_loss: 1.2577e-04 - val_loss: 6.1809e-04 - val_semantic_0_loss: 2.8642e-04 - val_semantic_1_loss: 2.1047e-04 - val_semantic_2_loss: 1.2119e-04 - lr: 4.8499e-05\n",
      "Epoch 74/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.5636e-04 - semantic_0_loss: 1.1537e-04 - semantic_1_loss: 1.1507e-04 - semantic_2_loss: 1.2591e-04\n",
      "Epoch 00074: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.5636e-04 - semantic_0_loss: 1.1537e-04 - semantic_1_loss: 1.1507e-04 - semantic_2_loss: 1.2591e-04 - val_loss: 6.0685e-04 - val_semantic_0_loss: 2.8230e-04 - val_semantic_1_loss: 2.0579e-04 - val_semantic_2_loss: 1.1875e-04 - lr: 4.8014e-05\n",
      "Epoch 75/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.4663e-04 - semantic_0_loss: 1.1083e-04 - semantic_1_loss: 1.1183e-04 - semantic_2_loss: 1.2397e-04\n",
      "Epoch 00075: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 250ms/step - loss: 3.4663e-04 - semantic_0_loss: 1.1083e-04 - semantic_1_loss: 1.1183e-04 - semantic_2_loss: 1.2397e-04 - val_loss: 5.8850e-04 - val_semantic_0_loss: 2.7671e-04 - val_semantic_1_loss: 1.9706e-04 - val_semantic_2_loss: 1.1473e-04 - lr: 4.7534e-05\n",
      "Epoch 76/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.5278e-04 - semantic_0_loss: 1.1348e-04 - semantic_1_loss: 1.1345e-04 - semantic_2_loss: 1.2586e-04\n",
      "Epoch 00076: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.5278e-04 - semantic_0_loss: 1.1348e-04 - semantic_1_loss: 1.1345e-04 - semantic_2_loss: 1.2586e-04 - val_loss: 6.0500e-04 - val_semantic_0_loss: 2.9765e-04 - val_semantic_1_loss: 1.9395e-04 - val_semantic_2_loss: 1.1340e-04 - lr: 4.7059e-05\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 77/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.5506e-04 - semantic_0_loss: 1.1465e-04 - semantic_1_loss: 1.1398e-04 - semantic_2_loss: 1.2643e-04\n",
      "Epoch 00077: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 3.5506e-04 - semantic_0_loss: 1.1465e-04 - semantic_1_loss: 1.1398e-04 - semantic_2_loss: 1.2643e-04 - val_loss: 5.8174e-04 - val_semantic_0_loss: 2.7073e-04 - val_semantic_1_loss: 1.9948e-04 - val_semantic_2_loss: 1.1153e-04 - lr: 4.6588e-05\n",
      "Epoch 78/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.5274e-04 - semantic_0_loss: 1.1248e-04 - semantic_1_loss: 1.1400e-04 - semantic_2_loss: 1.2626e-04\n",
      "Epoch 00078: val_loss did not improve from 0.00057\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.5274e-04 - semantic_0_loss: 1.1248e-04 - semantic_1_loss: 1.1400e-04 - semantic_2_loss: 1.2626e-04 - val_loss: 5.8986e-04 - val_semantic_0_loss: 2.8133e-04 - val_semantic_1_loss: 1.9609e-04 - val_semantic_2_loss: 1.1244e-04 - lr: 4.6122e-05\n",
      "Epoch 79/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.5079e-04 - semantic_0_loss: 1.1280e-04 - semantic_1_loss: 1.1206e-04 - semantic_2_loss: 1.2593e-04\n",
      "Epoch 00079: val_loss improved from 0.00057 to 0.00055, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 57s 277ms/step - loss: 3.5079e-04 - semantic_0_loss: 1.1280e-04 - semantic_1_loss: 1.1206e-04 - semantic_2_loss: 1.2593e-04 - val_loss: 5.4738e-04 - val_semantic_0_loss: 2.4748e-04 - val_semantic_1_loss: 1.9264e-04 - val_semantic_2_loss: 1.0725e-04 - lr: 4.5661e-05\n",
      "Epoch 80/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.4153e-04 - semantic_0_loss: 1.0652e-04 - semantic_1_loss: 1.1101e-04 - semantic_2_loss: 1.2401e-04\n",
      "Epoch 00080: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 3.4153e-04 - semantic_0_loss: 1.0652e-04 - semantic_1_loss: 1.1101e-04 - semantic_2_loss: 1.2401e-04 - val_loss: 5.7827e-04 - val_semantic_0_loss: 2.6881e-04 - val_semantic_1_loss: 1.9806e-04 - val_semantic_2_loss: 1.1140e-04 - lr: 4.5204e-05\n",
      "Epoch 81/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.4883e-04 - semantic_0_loss: 1.0936e-04 - semantic_1_loss: 1.1450e-04 - semantic_2_loss: 1.2497e-04\n",
      "Epoch 00081: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 250ms/step - loss: 3.4883e-04 - semantic_0_loss: 1.0936e-04 - semantic_1_loss: 1.1450e-04 - semantic_2_loss: 1.2497e-04 - val_loss: 6.0659e-04 - val_semantic_0_loss: 3.0291e-04 - val_semantic_1_loss: 1.9400e-04 - val_semantic_2_loss: 1.0969e-04 - lr: 4.4752e-05\n",
      "Epoch 82/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.3470e-04 - semantic_0_loss: 1.0305e-04 - semantic_1_loss: 1.0769e-04 - semantic_2_loss: 1.2395e-04\n",
      "Epoch 00082: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.3470e-04 - semantic_0_loss: 1.0305e-04 - semantic_1_loss: 1.0769e-04 - semantic_2_loss: 1.2395e-04 - val_loss: 5.8426e-04 - val_semantic_0_loss: 2.8759e-04 - val_semantic_1_loss: 1.8656e-04 - val_semantic_2_loss: 1.1010e-04 - lr: 4.4305e-05\n",
      "Epoch 83/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.4559e-04 - semantic_0_loss: 1.0616e-04 - semantic_1_loss: 1.1164e-04 - semantic_2_loss: 1.2780e-04\n",
      "Epoch 00083: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.4559e-04 - semantic_0_loss: 1.0616e-04 - semantic_1_loss: 1.1164e-04 - semantic_2_loss: 1.2780e-04 - val_loss: 5.9236e-04 - val_semantic_0_loss: 2.8704e-04 - val_semantic_1_loss: 1.8861e-04 - val_semantic_2_loss: 1.1670e-04 - lr: 4.3862e-05\n",
      "Epoch 84/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.5786e-04 - semantic_0_loss: 1.0903e-04 - semantic_1_loss: 1.1853e-04 - semantic_2_loss: 1.3030e-04\n",
      "Epoch 00084: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 3.5786e-04 - semantic_0_loss: 1.0903e-04 - semantic_1_loss: 1.1853e-04 - semantic_2_loss: 1.3030e-04 - val_loss: 5.7313e-04 - val_semantic_0_loss: 2.7408e-04 - val_semantic_1_loss: 1.8985e-04 - val_semantic_2_loss: 1.0920e-04 - lr: 4.3423e-05\n",
      "Epoch 85/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.3499e-04 - semantic_0_loss: 1.0322e-04 - semantic_1_loss: 1.0802e-04 - semantic_2_loss: 1.2374e-04\n",
      "Epoch 00085: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 3.3499e-04 - semantic_0_loss: 1.0322e-04 - semantic_1_loss: 1.0802e-04 - semantic_2_loss: 1.2374e-04 - val_loss: 6.1439e-04 - val_semantic_0_loss: 3.0690e-04 - val_semantic_1_loss: 1.9579e-04 - val_semantic_2_loss: 1.1170e-04 - lr: 4.2989e-05\n",
      "Epoch 86/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.3671e-04 - semantic_0_loss: 1.0639e-04 - semantic_1_loss: 1.0606e-04 - semantic_2_loss: 1.2426e-04\n",
      "Epoch 00086: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 3.3671e-04 - semantic_0_loss: 1.0639e-04 - semantic_1_loss: 1.0606e-04 - semantic_2_loss: 1.2426e-04 - val_loss: 6.1604e-04 - val_semantic_0_loss: 3.1282e-04 - val_semantic_1_loss: 1.9180e-04 - val_semantic_2_loss: 1.1142e-04 - lr: 4.2559e-05\n",
      "Epoch 87/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.2473e-04 - semantic_0_loss: 9.9351e-05 - semantic_1_loss: 1.0275e-04 - semantic_2_loss: 1.2263e-04\n",
      "Epoch 00087: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 3.2473e-04 - semantic_0_loss: 9.9351e-05 - semantic_1_loss: 1.0275e-04 - semantic_2_loss: 1.2263e-04 - val_loss: 5.5779e-04 - val_semantic_0_loss: 2.6228e-04 - val_semantic_1_loss: 1.8868e-04 - val_semantic_2_loss: 1.0682e-04 - lr: 4.2133e-05\n",
      "Epoch 88/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.2734e-04 - semantic_0_loss: 9.9244e-05 - semantic_1_loss: 1.0492e-04 - semantic_2_loss: 1.2317e-04\n",
      "Epoch 00088: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 3.2734e-04 - semantic_0_loss: 9.9244e-05 - semantic_1_loss: 1.0492e-04 - semantic_2_loss: 1.2317e-04 - val_loss: 5.9328e-04 - val_semantic_0_loss: 2.8378e-04 - val_semantic_1_loss: 1.9834e-04 - val_semantic_2_loss: 1.1116e-04 - lr: 4.1712e-05\n",
      "Epoch 89/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.2989e-04 - semantic_0_loss: 1.0197e-04 - semantic_1_loss: 1.0525e-04 - semantic_2_loss: 1.2267e-04\n",
      "Epoch 00089: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.2989e-04 - semantic_0_loss: 1.0197e-04 - semantic_1_loss: 1.0525e-04 - semantic_2_loss: 1.2267e-04 - val_loss: 5.7559e-04 - val_semantic_0_loss: 2.7611e-04 - val_semantic_1_loss: 1.9187e-04 - val_semantic_2_loss: 1.0761e-04 - lr: 4.1295e-05\n",
      "Epoch 90/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.3217e-04 - semantic_0_loss: 1.0030e-04 - semantic_1_loss: 1.0792e-04 - semantic_2_loss: 1.2395e-04\n",
      "Epoch 00090: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.3217e-04 - semantic_0_loss: 1.0030e-04 - semantic_1_loss: 1.0792e-04 - semantic_2_loss: 1.2395e-04 - val_loss: 6.1289e-04 - val_semantic_0_loss: 2.9900e-04 - val_semantic_1_loss: 2.0079e-04 - val_semantic_2_loss: 1.1310e-04 - lr: 4.0882e-05\n",
      "Epoch 91/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.2507e-04 - semantic_0_loss: 9.8847e-05 - semantic_1_loss: 1.0456e-04 - semantic_2_loss: 1.2167e-04\n",
      "Epoch 00091: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 3.2507e-04 - semantic_0_loss: 9.8847e-05 - semantic_1_loss: 1.0456e-04 - semantic_2_loss: 1.2167e-04 - val_loss: 5.9260e-04 - val_semantic_0_loss: 2.7924e-04 - val_semantic_1_loss: 1.9982e-04 - val_semantic_2_loss: 1.1355e-04 - lr: 4.0473e-05\n",
      "Epoch 92/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.2185e-04 - semantic_0_loss: 9.5071e-05 - semantic_1_loss: 1.0448e-04 - semantic_2_loss: 1.2230e-04\n",
      "Epoch 00092: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 3.2185e-04 - semantic_0_loss: 9.5071e-05 - semantic_1_loss: 1.0448e-04 - semantic_2_loss: 1.2230e-04 - val_loss: 5.7383e-04 - val_semantic_0_loss: 2.7693e-04 - val_semantic_1_loss: 1.8506e-04 - val_semantic_2_loss: 1.1184e-04 - lr: 4.0068e-05\n",
      "Epoch 93/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.3061e-04 - semantic_0_loss: 9.9149e-05 - semantic_1_loss: 1.0799e-04 - semantic_2_loss: 1.2348e-04\n",
      "Epoch 00093: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.3061e-04 - semantic_0_loss: 9.9149e-05 - semantic_1_loss: 1.0799e-04 - semantic_2_loss: 1.2348e-04 - val_loss: 5.7907e-04 - val_semantic_0_loss: 2.7878e-04 - val_semantic_1_loss: 1.8903e-04 - val_semantic_2_loss: 1.1125e-04 - lr: 3.9668e-05\n",
      "Epoch 94/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.2195e-04 - semantic_0_loss: 9.7254e-05 - semantic_1_loss: 1.0346e-04 - semantic_2_loss: 1.2124e-04\n",
      "Epoch 00094: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 3.2195e-04 - semantic_0_loss: 9.7254e-05 - semantic_1_loss: 1.0346e-04 - semantic_2_loss: 1.2124e-04 - val_loss: 5.7606e-04 - val_semantic_0_loss: 2.7962e-04 - val_semantic_1_loss: 1.8944e-04 - val_semantic_2_loss: 1.0699e-04 - lr: 3.9271e-05\n",
      "Epoch 95/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.2020e-04 - semantic_0_loss: 9.4353e-05 - semantic_1_loss: 1.0406e-04 - semantic_2_loss: 1.2178e-04\n",
      "Epoch 00095: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.2020e-04 - semantic_0_loss: 9.4353e-05 - semantic_1_loss: 1.0406e-04 - semantic_2_loss: 1.2178e-04 - val_loss: 6.0037e-04 - val_semantic_0_loss: 2.9773e-04 - val_semantic_1_loss: 1.9112e-04 - val_semantic_2_loss: 1.1152e-04 - lr: 3.8878e-05\n",
      "Epoch 96/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.2480e-04 - semantic_0_loss: 9.6645e-05 - semantic_1_loss: 1.0580e-04 - semantic_2_loss: 1.2236e-04\n",
      "Epoch 00096: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.2480e-04 - semantic_0_loss: 9.6645e-05 - semantic_1_loss: 1.0580e-04 - semantic_2_loss: 1.2236e-04 - val_loss: 5.7261e-04 - val_semantic_0_loss: 2.6734e-04 - val_semantic_1_loss: 1.9538e-04 - val_semantic_2_loss: 1.0989e-04 - lr: 3.8490e-05\n",
      "Epoch 97/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.1582e-04 - semantic_0_loss: 9.3814e-05 - semantic_1_loss: 1.0058e-04 - semantic_2_loss: 1.2142e-04\n",
      "Epoch 00097: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.1582e-04 - semantic_0_loss: 9.3814e-05 - semantic_1_loss: 1.0058e-04 - semantic_2_loss: 1.2142e-04 - val_loss: 5.7003e-04 - val_semantic_0_loss: 2.6818e-04 - val_semantic_1_loss: 1.9144e-04 - val_semantic_2_loss: 1.1042e-04 - lr: 3.8105e-05\n",
      "Epoch 98/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.1806e-04 - semantic_0_loss: 9.7016e-05 - semantic_1_loss: 1.0015e-04 - semantic_2_loss: 1.2089e-04\n",
      "Epoch 00098: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 3.1806e-04 - semantic_0_loss: 9.7016e-05 - semantic_1_loss: 1.0015e-04 - semantic_2_loss: 1.2089e-04 - val_loss: 6.0672e-04 - val_semantic_0_loss: 2.9274e-04 - val_semantic_1_loss: 2.0263e-04 - val_semantic_2_loss: 1.1135e-04 - lr: 3.7724e-05\n",
      "Epoch 99/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.1075e-04 - semantic_0_loss: 9.1317e-05 - semantic_1_loss: 9.9257e-05 - semantic_2_loss: 1.2017e-04\n",
      "Epoch 00099: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 3.1075e-04 - semantic_0_loss: 9.1317e-05 - semantic_1_loss: 9.9257e-05 - semantic_2_loss: 1.2017e-04 - val_loss: 5.7157e-04 - val_semantic_0_loss: 2.6313e-04 - val_semantic_1_loss: 1.9209e-04 - val_semantic_2_loss: 1.1635e-04 - lr: 3.7346e-05\n",
      "Epoch 100/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.1780e-04 - semantic_0_loss: 9.3487e-05 - semantic_1_loss: 1.0189e-04 - semantic_2_loss: 1.2243e-04\n",
      "Epoch 00100: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 3.1780e-04 - semantic_0_loss: 9.3487e-05 - semantic_1_loss: 1.0189e-04 - semantic_2_loss: 1.2243e-04 - val_loss: 5.7846e-04 - val_semantic_0_loss: 2.7857e-04 - val_semantic_1_loss: 1.8917e-04 - val_semantic_2_loss: 1.1073e-04 - lr: 3.6973e-05\n",
      "Epoch 101/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.1044e-04 - semantic_0_loss: 8.9726e-05 - semantic_1_loss: 1.0117e-04 - semantic_2_loss: 1.1954e-04\n",
      "Epoch 00101: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 3.1044e-04 - semantic_0_loss: 8.9726e-05 - semantic_1_loss: 1.0117e-04 - semantic_2_loss: 1.1954e-04 - val_loss: 6.1569e-04 - val_semantic_0_loss: 3.0967e-04 - val_semantic_1_loss: 1.9533e-04 - val_semantic_2_loss: 1.1069e-04 - lr: 3.6603e-05\n",
      "Epoch 102/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.1398e-04 - semantic_0_loss: 9.1815e-05 - semantic_1_loss: 1.0089e-04 - semantic_2_loss: 1.2127e-04\n",
      "Epoch 00102: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.1398e-04 - semantic_0_loss: 9.1815e-05 - semantic_1_loss: 1.0089e-04 - semantic_2_loss: 1.2127e-04 - val_loss: 5.4830e-04 - val_semantic_0_loss: 2.5525e-04 - val_semantic_1_loss: 1.8617e-04 - val_semantic_2_loss: 1.0688e-04 - lr: 3.6237e-05\n",
      "Epoch 103/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.1419e-04 - semantic_0_loss: 9.1063e-05 - semantic_1_loss: 1.0177e-04 - semantic_2_loss: 1.2136e-04\n",
      "Epoch 00103: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.1419e-04 - semantic_0_loss: 9.1063e-05 - semantic_1_loss: 1.0177e-04 - semantic_2_loss: 1.2136e-04 - val_loss: 6.2226e-04 - val_semantic_0_loss: 3.1213e-04 - val_semantic_1_loss: 1.9440e-04 - val_semantic_2_loss: 1.1573e-04 - lr: 3.5875e-05\n",
      "Epoch 104/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.1393e-04 - semantic_0_loss: 9.1373e-05 - semantic_1_loss: 1.0297e-04 - semantic_2_loss: 1.1959e-04\n",
      "Epoch 00104: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.1393e-04 - semantic_0_loss: 9.1373e-05 - semantic_1_loss: 1.0297e-04 - semantic_2_loss: 1.1959e-04 - val_loss: 5.5203e-04 - val_semantic_0_loss: 2.5240e-04 - val_semantic_1_loss: 1.8990e-04 - val_semantic_2_loss: 1.0974e-04 - lr: 3.5516e-05\n",
      "Epoch 105/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0820e-04 - semantic_0_loss: 8.8800e-05 - semantic_1_loss: 9.9330e-05 - semantic_2_loss: 1.2007e-04\n",
      "Epoch 00105: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 3.0820e-04 - semantic_0_loss: 8.8800e-05 - semantic_1_loss: 9.9330e-05 - semantic_2_loss: 1.2007e-04 - val_loss: 5.5476e-04 - val_semantic_0_loss: 2.6275e-04 - val_semantic_1_loss: 1.8603e-04 - val_semantic_2_loss: 1.0599e-04 - lr: 3.5161e-05\n",
      "Epoch 106/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0545e-04 - semantic_0_loss: 8.8889e-05 - semantic_1_loss: 9.8474e-05 - semantic_2_loss: 1.1809e-04\n",
      "Epoch 00106: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 3.0545e-04 - semantic_0_loss: 8.8889e-05 - semantic_1_loss: 9.8474e-05 - semantic_2_loss: 1.1809e-04 - val_loss: 5.7569e-04 - val_semantic_0_loss: 2.7972e-04 - val_semantic_1_loss: 1.8561e-04 - val_semantic_2_loss: 1.1036e-04 - lr: 3.4809e-05\n",
      "Epoch 107/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0478e-04 - semantic_0_loss: 8.7976e-05 - semantic_1_loss: 9.6737e-05 - semantic_2_loss: 1.2006e-04\n",
      "Epoch 00107: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 3.0478e-04 - semantic_0_loss: 8.7976e-05 - semantic_1_loss: 9.6737e-05 - semantic_2_loss: 1.2006e-04 - val_loss: 5.8731e-04 - val_semantic_0_loss: 2.8702e-04 - val_semantic_1_loss: 1.9034e-04 - val_semantic_2_loss: 1.0994e-04 - lr: 3.4461e-05\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 108/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0740e-04 - semantic_0_loss: 8.8243e-05 - semantic_1_loss: 9.9026e-05 - semantic_2_loss: 1.2013e-04\n",
      "Epoch 00108: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.0740e-04 - semantic_0_loss: 8.8243e-05 - semantic_1_loss: 9.9026e-05 - semantic_2_loss: 1.2013e-04 - val_loss: 5.7545e-04 - val_semantic_0_loss: 2.7971e-04 - val_semantic_1_loss: 1.8786e-04 - val_semantic_2_loss: 1.0788e-04 - lr: 3.4117e-05\n",
      "Epoch 109/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0771e-04 - semantic_0_loss: 8.6259e-05 - semantic_1_loss: 1.0089e-04 - semantic_2_loss: 1.2057e-04\n",
      "Epoch 00109: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 244ms/step - loss: 3.0771e-04 - semantic_0_loss: 8.6259e-05 - semantic_1_loss: 1.0089e-04 - semantic_2_loss: 1.2057e-04 - val_loss: 6.2359e-04 - val_semantic_0_loss: 3.2995e-04 - val_semantic_1_loss: 1.8380e-04 - val_semantic_2_loss: 1.0985e-04 - lr: 3.3775e-05\n",
      "Epoch 110/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0377e-04 - semantic_0_loss: 8.6408e-05 - semantic_1_loss: 9.7915e-05 - semantic_2_loss: 1.1945e-04\n",
      "Epoch 00110: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.0377e-04 - semantic_0_loss: 8.6408e-05 - semantic_1_loss: 9.7915e-05 - semantic_2_loss: 1.1945e-04 - val_loss: 5.7505e-04 - val_semantic_0_loss: 2.8336e-04 - val_semantic_1_loss: 1.8471e-04 - val_semantic_2_loss: 1.0698e-04 - lr: 3.3438e-05\n",
      "Epoch 111/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0099e-04 - semantic_0_loss: 8.4139e-05 - semantic_1_loss: 9.7839e-05 - semantic_2_loss: 1.1901e-04\n",
      "Epoch 00111: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 3.0099e-04 - semantic_0_loss: 8.4139e-05 - semantic_1_loss: 9.7839e-05 - semantic_2_loss: 1.1901e-04 - val_loss: 5.7805e-04 - val_semantic_0_loss: 2.8000e-04 - val_semantic_1_loss: 1.8610e-04 - val_semantic_2_loss: 1.1195e-04 - lr: 3.3103e-05\n",
      "Epoch 112/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0327e-04 - semantic_0_loss: 8.5537e-05 - semantic_1_loss: 9.8305e-05 - semantic_2_loss: 1.1942e-04\n",
      "Epoch 00112: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.0327e-04 - semantic_0_loss: 8.5537e-05 - semantic_1_loss: 9.8305e-05 - semantic_2_loss: 1.1942e-04 - val_loss: 5.4859e-04 - val_semantic_0_loss: 2.5811e-04 - val_semantic_1_loss: 1.8498e-04 - val_semantic_2_loss: 1.0550e-04 - lr: 3.2772e-05\n",
      "Epoch 113/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0399e-04 - semantic_0_loss: 8.6232e-05 - semantic_1_loss: 9.7666e-05 - semantic_2_loss: 1.2010e-04\n",
      "Epoch 00113: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 250ms/step - loss: 3.0399e-04 - semantic_0_loss: 8.6232e-05 - semantic_1_loss: 9.7666e-05 - semantic_2_loss: 1.2010e-04 - val_loss: 5.6341e-04 - val_semantic_0_loss: 2.6897e-04 - val_semantic_1_loss: 1.8800e-04 - val_semantic_2_loss: 1.0644e-04 - lr: 3.2445e-05\n",
      "Epoch 114/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0058e-04 - semantic_0_loss: 8.4147e-05 - semantic_1_loss: 9.7727e-05 - semantic_2_loss: 1.1871e-04\n",
      "Epoch 00114: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 3.0058e-04 - semantic_0_loss: 8.4147e-05 - semantic_1_loss: 9.7727e-05 - semantic_2_loss: 1.1871e-04 - val_loss: 5.5905e-04 - val_semantic_0_loss: 2.6110e-04 - val_semantic_1_loss: 1.8609e-04 - val_semantic_2_loss: 1.1187e-04 - lr: 3.2120e-05\n",
      "Epoch 115/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9885e-04 - semantic_0_loss: 8.3952e-05 - semantic_1_loss: 9.6462e-05 - semantic_2_loss: 1.1844e-04\n",
      "Epoch 00115: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 2.9885e-04 - semantic_0_loss: 8.3952e-05 - semantic_1_loss: 9.6462e-05 - semantic_2_loss: 1.1844e-04 - val_loss: 5.5221e-04 - val_semantic_0_loss: 2.5860e-04 - val_semantic_1_loss: 1.8416e-04 - val_semantic_2_loss: 1.0944e-04 - lr: 3.1799e-05\n",
      "Epoch 116/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 3.0640e-04 - semantic_0_loss: 8.8017e-05 - semantic_1_loss: 9.8400e-05 - semantic_2_loss: 1.1998e-04\n",
      "Epoch 00116: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 244ms/step - loss: 3.0640e-04 - semantic_0_loss: 8.8017e-05 - semantic_1_loss: 9.8400e-05 - semantic_2_loss: 1.1998e-04 - val_loss: 5.7456e-04 - val_semantic_0_loss: 2.6861e-04 - val_semantic_1_loss: 1.9163e-04 - val_semantic_2_loss: 1.1432e-04 - lr: 3.1481e-05\n",
      "Epoch 117/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9991e-04 - semantic_0_loss: 8.1684e-05 - semantic_1_loss: 9.8167e-05 - semantic_2_loss: 1.2006e-04\n",
      "Epoch 00117: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 2.9991e-04 - semantic_0_loss: 8.1684e-05 - semantic_1_loss: 9.8167e-05 - semantic_2_loss: 1.2006e-04 - val_loss: 5.6165e-04 - val_semantic_0_loss: 2.6557e-04 - val_semantic_1_loss: 1.8545e-04 - val_semantic_2_loss: 1.1063e-04 - lr: 3.1166e-05\n",
      "Epoch 118/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9549e-04 - semantic_0_loss: 8.1336e-05 - semantic_1_loss: 9.5122e-05 - semantic_2_loss: 1.1904e-04\n",
      "Epoch 00118: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 2.9549e-04 - semantic_0_loss: 8.1336e-05 - semantic_1_loss: 9.5122e-05 - semantic_2_loss: 1.1904e-04 - val_loss: 5.9813e-04 - val_semantic_0_loss: 2.9439e-04 - val_semantic_1_loss: 1.8993e-04 - val_semantic_2_loss: 1.1381e-04 - lr: 3.0854e-05\n",
      "Epoch 119/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9825e-04 - semantic_0_loss: 8.3216e-05 - semantic_1_loss: 9.6173e-05 - semantic_2_loss: 1.1886e-04\n",
      "Epoch 00119: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 2.9825e-04 - semantic_0_loss: 8.3216e-05 - semantic_1_loss: 9.6173e-05 - semantic_2_loss: 1.1886e-04 - val_loss: 6.1287e-04 - val_semantic_0_loss: 3.0839e-04 - val_semantic_1_loss: 1.9104e-04 - val_semantic_2_loss: 1.1344e-04 - lr: 3.0546e-05\n",
      "Epoch 120/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9504e-04 - semantic_0_loss: 8.1940e-05 - semantic_1_loss: 9.4804e-05 - semantic_2_loss: 1.1829e-04\n",
      "Epoch 00120: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.9504e-04 - semantic_0_loss: 8.1940e-05 - semantic_1_loss: 9.4804e-05 - semantic_2_loss: 1.1829e-04 - val_loss: 5.4909e-04 - val_semantic_0_loss: 2.5322e-04 - val_semantic_1_loss: 1.8564e-04 - val_semantic_2_loss: 1.1022e-04 - lr: 3.0240e-05\n",
      "Epoch 121/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9045e-04 - semantic_0_loss: 7.9012e-05 - semantic_1_loss: 9.3136e-05 - semantic_2_loss: 1.1830e-04\n",
      "Epoch 00121: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 2.9045e-04 - semantic_0_loss: 7.9012e-05 - semantic_1_loss: 9.3136e-05 - semantic_2_loss: 1.1830e-04 - val_loss: 5.9184e-04 - val_semantic_0_loss: 2.7982e-04 - val_semantic_1_loss: 1.9283e-04 - val_semantic_2_loss: 1.1920e-04 - lr: 2.9938e-05\n",
      "Epoch 122/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9719e-04 - semantic_0_loss: 8.2385e-05 - semantic_1_loss: 9.6346e-05 - semantic_2_loss: 1.1846e-04\n",
      "Epoch 00122: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 2.9719e-04 - semantic_0_loss: 8.2385e-05 - semantic_1_loss: 9.6346e-05 - semantic_2_loss: 1.1846e-04 - val_loss: 5.6768e-04 - val_semantic_0_loss: 2.6919e-04 - val_semantic_1_loss: 1.8815e-04 - val_semantic_2_loss: 1.1034e-04 - lr: 2.9639e-05\n",
      "Epoch 123/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9360e-04 - semantic_0_loss: 8.1110e-05 - semantic_1_loss: 9.4291e-05 - semantic_2_loss: 1.1820e-04\n",
      "Epoch 00123: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 2.9360e-04 - semantic_0_loss: 8.1110e-05 - semantic_1_loss: 9.4291e-05 - semantic_2_loss: 1.1820e-04 - val_loss: 5.8234e-04 - val_semantic_0_loss: 2.8780e-04 - val_semantic_1_loss: 1.8337e-04 - val_semantic_2_loss: 1.1117e-04 - lr: 2.9342e-05\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 124/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9956e-04 - semantic_0_loss: 8.4465e-05 - semantic_1_loss: 9.7153e-05 - semantic_2_loss: 1.1794e-04\n",
      "Epoch 00124: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 2.9956e-04 - semantic_0_loss: 8.4465e-05 - semantic_1_loss: 9.7153e-05 - semantic_2_loss: 1.1794e-04 - val_loss: 5.6342e-04 - val_semantic_0_loss: 2.6856e-04 - val_semantic_1_loss: 1.8252e-04 - val_semantic_2_loss: 1.1235e-04 - lr: 2.9049e-05\n",
      "Epoch 125/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9149e-04 - semantic_0_loss: 7.9878e-05 - semantic_1_loss: 9.3255e-05 - semantic_2_loss: 1.1836e-04\n",
      "Epoch 00125: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 2.9149e-04 - semantic_0_loss: 7.9878e-05 - semantic_1_loss: 9.3255e-05 - semantic_2_loss: 1.1836e-04 - val_loss: 5.6010e-04 - val_semantic_0_loss: 2.6988e-04 - val_semantic_1_loss: 1.8052e-04 - val_semantic_2_loss: 1.0970e-04 - lr: 2.8758e-05\n",
      "Epoch 126/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.9124e-04 - semantic_0_loss: 7.8421e-05 - semantic_1_loss: 9.3998e-05 - semantic_2_loss: 1.1882e-04\n",
      "Epoch 00126: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 2.9124e-04 - semantic_0_loss: 7.8421e-05 - semantic_1_loss: 9.3998e-05 - semantic_2_loss: 1.1882e-04 - val_loss: 5.7444e-04 - val_semantic_0_loss: 2.7512e-04 - val_semantic_1_loss: 1.8839e-04 - val_semantic_2_loss: 1.1092e-04 - lr: 2.8471e-05\n",
      "Epoch 127/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8957e-04 - semantic_0_loss: 7.9173e-05 - semantic_1_loss: 9.2792e-05 - semantic_2_loss: 1.1761e-04\n",
      "Epoch 00127: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 2.8957e-04 - semantic_0_loss: 7.9173e-05 - semantic_1_loss: 9.2792e-05 - semantic_2_loss: 1.1761e-04 - val_loss: 5.6841e-04 - val_semantic_0_loss: 2.7169e-04 - val_semantic_1_loss: 1.8667e-04 - val_semantic_2_loss: 1.1005e-04 - lr: 2.8186e-05\n",
      "Epoch 128/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8716e-04 - semantic_0_loss: 7.7533e-05 - semantic_1_loss: 9.2321e-05 - semantic_2_loss: 1.1731e-04\n",
      "Epoch 00128: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.8716e-04 - semantic_0_loss: 7.7533e-05 - semantic_1_loss: 9.2321e-05 - semantic_2_loss: 1.1731e-04 - val_loss: 5.8334e-04 - val_semantic_0_loss: 2.8617e-04 - val_semantic_1_loss: 1.8760e-04 - val_semantic_2_loss: 1.0956e-04 - lr: 2.7904e-05\n",
      "Epoch 129/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8747e-04 - semantic_0_loss: 7.7483e-05 - semantic_1_loss: 9.2922e-05 - semantic_2_loss: 1.1707e-04\n",
      "Epoch 00129: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 2.8747e-04 - semantic_0_loss: 7.7483e-05 - semantic_1_loss: 9.2922e-05 - semantic_2_loss: 1.1707e-04 - val_loss: 5.6361e-04 - val_semantic_0_loss: 2.6720e-04 - val_semantic_1_loss: 1.8607e-04 - val_semantic_2_loss: 1.1034e-04 - lr: 2.7625e-05\n",
      "Epoch 130/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8592e-04 - semantic_0_loss: 7.6877e-05 - semantic_1_loss: 9.2114e-05 - semantic_2_loss: 1.1693e-04\n",
      "Epoch 00130: val_loss did not improve from 0.00055\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.8592e-04 - semantic_0_loss: 7.6877e-05 - semantic_1_loss: 9.2114e-05 - semantic_2_loss: 1.1693e-04 - val_loss: 5.6151e-04 - val_semantic_0_loss: 2.6521e-04 - val_semantic_1_loss: 1.8491e-04 - val_semantic_2_loss: 1.1140e-04 - lr: 2.7349e-05\n",
      "Epoch 131/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8377e-04 - semantic_0_loss: 7.6032e-05 - semantic_1_loss: 9.1156e-05 - semantic_2_loss: 1.1658e-04\n",
      "Epoch 00131: val_loss improved from 0.00055 to 0.00054, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 275ms/step - loss: 2.8377e-04 - semantic_0_loss: 7.6032e-05 - semantic_1_loss: 9.1156e-05 - semantic_2_loss: 1.1658e-04 - val_loss: 5.3586e-04 - val_semantic_0_loss: 2.4469e-04 - val_semantic_1_loss: 1.7962e-04 - val_semantic_2_loss: 1.1155e-04 - lr: 2.7075e-05\n",
      "Epoch 132/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8995e-04 - semantic_0_loss: 7.8833e-05 - semantic_1_loss: 9.3013e-05 - semantic_2_loss: 1.1811e-04\n",
      "Epoch 00132: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 2.8995e-04 - semantic_0_loss: 7.8833e-05 - semantic_1_loss: 9.3013e-05 - semantic_2_loss: 1.1811e-04 - val_loss: 5.6551e-04 - val_semantic_0_loss: 2.6868e-04 - val_semantic_1_loss: 1.8197e-04 - val_semantic_2_loss: 1.1485e-04 - lr: 2.6805e-05\n",
      "Epoch 133/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8620e-04 - semantic_0_loss: 7.6494e-05 - semantic_1_loss: 9.2076e-05 - semantic_2_loss: 1.1763e-04\n",
      "Epoch 00133: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 2.8620e-04 - semantic_0_loss: 7.6494e-05 - semantic_1_loss: 9.2076e-05 - semantic_2_loss: 1.1763e-04 - val_loss: 5.8052e-04 - val_semantic_0_loss: 2.8589e-04 - val_semantic_1_loss: 1.8456e-04 - val_semantic_2_loss: 1.1006e-04 - lr: 2.6537e-05\n",
      "Epoch 134/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8118e-04 - semantic_0_loss: 7.5009e-05 - semantic_1_loss: 8.9978e-05 - semantic_2_loss: 1.1619e-04\n",
      "Epoch 00134: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 2.8118e-04 - semantic_0_loss: 7.5009e-05 - semantic_1_loss: 8.9978e-05 - semantic_2_loss: 1.1619e-04 - val_loss: 5.6869e-04 - val_semantic_0_loss: 2.7285e-04 - val_semantic_1_loss: 1.8624e-04 - val_semantic_2_loss: 1.0960e-04 - lr: 2.6271e-05\n",
      "Epoch 135/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8203e-04 - semantic_0_loss: 7.2736e-05 - semantic_1_loss: 9.2603e-05 - semantic_2_loss: 1.1669e-04\n",
      "Epoch 00135: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 250ms/step - loss: 2.8203e-04 - semantic_0_loss: 7.2736e-05 - semantic_1_loss: 9.2603e-05 - semantic_2_loss: 1.1669e-04 - val_loss: 5.4936e-04 - val_semantic_0_loss: 2.5936e-04 - val_semantic_1_loss: 1.8370e-04 - val_semantic_2_loss: 1.0630e-04 - lr: 2.6009e-05\n",
      "Epoch 136/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8367e-04 - semantic_0_loss: 7.5473e-05 - semantic_1_loss: 9.2006e-05 - semantic_2_loss: 1.1619e-04\n",
      "Epoch 00136: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 2.8367e-04 - semantic_0_loss: 7.5473e-05 - semantic_1_loss: 9.2006e-05 - semantic_2_loss: 1.1619e-04 - val_loss: 5.4930e-04 - val_semantic_0_loss: 2.6006e-04 - val_semantic_1_loss: 1.8351e-04 - val_semantic_2_loss: 1.0574e-04 - lr: 2.5748e-05\n",
      "Epoch 137/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8434e-04 - semantic_0_loss: 7.6012e-05 - semantic_1_loss: 9.1998e-05 - semantic_2_loss: 1.1633e-04\n",
      "Epoch 00137: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.8434e-04 - semantic_0_loss: 7.6012e-05 - semantic_1_loss: 9.1998e-05 - semantic_2_loss: 1.1633e-04 - val_loss: 5.6424e-04 - val_semantic_0_loss: 2.6910e-04 - val_semantic_1_loss: 1.8421e-04 - val_semantic_2_loss: 1.1094e-04 - lr: 2.5491e-05\n",
      "Epoch 138/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8533e-04 - semantic_0_loss: 7.6656e-05 - semantic_1_loss: 9.2099e-05 - semantic_2_loss: 1.1658e-04\n",
      "Epoch 00138: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.8533e-04 - semantic_0_loss: 7.6656e-05 - semantic_1_loss: 9.2099e-05 - semantic_2_loss: 1.1658e-04 - val_loss: 5.6704e-04 - val_semantic_0_loss: 2.7250e-04 - val_semantic_1_loss: 1.8477e-04 - val_semantic_2_loss: 1.0977e-04 - lr: 2.5236e-05\n",
      "Epoch 139/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8209e-04 - semantic_0_loss: 7.4896e-05 - semantic_1_loss: 9.0698e-05 - semantic_2_loss: 1.1650e-04\n",
      "Epoch 00139: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 2.8209e-04 - semantic_0_loss: 7.4896e-05 - semantic_1_loss: 9.0698e-05 - semantic_2_loss: 1.1650e-04 - val_loss: 5.4942e-04 - val_semantic_0_loss: 2.6645e-04 - val_semantic_1_loss: 1.7525e-04 - val_semantic_2_loss: 1.0772e-04 - lr: 2.4984e-05\n",
      "Epoch 140/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7507e-04 - semantic_0_loss: 7.2471e-05 - semantic_1_loss: 8.7049e-05 - semantic_2_loss: 1.1555e-04\n",
      "Epoch 00140: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 2.7507e-04 - semantic_0_loss: 7.2471e-05 - semantic_1_loss: 8.7049e-05 - semantic_2_loss: 1.1555e-04 - val_loss: 5.6792e-04 - val_semantic_0_loss: 2.7262e-04 - val_semantic_1_loss: 1.8628e-04 - val_semantic_2_loss: 1.0902e-04 - lr: 2.4734e-05\n",
      "Epoch 141/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8189e-04 - semantic_0_loss: 7.6092e-05 - semantic_1_loss: 8.9769e-05 - semantic_2_loss: 1.1603e-04\n",
      "Epoch 00141: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.8189e-04 - semantic_0_loss: 7.6092e-05 - semantic_1_loss: 8.9769e-05 - semantic_2_loss: 1.1603e-04 - val_loss: 5.4628e-04 - val_semantic_0_loss: 2.5976e-04 - val_semantic_1_loss: 1.7983e-04 - val_semantic_2_loss: 1.0669e-04 - lr: 2.4487e-05\n",
      "Epoch 142/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.8344e-04 - semantic_0_loss: 7.5414e-05 - semantic_1_loss: 9.1382e-05 - semantic_2_loss: 1.1664e-04\n",
      "Epoch 00142: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 2.8344e-04 - semantic_0_loss: 7.5414e-05 - semantic_1_loss: 9.1382e-05 - semantic_2_loss: 1.1664e-04 - val_loss: 5.6865e-04 - val_semantic_0_loss: 2.7186e-04 - val_semantic_1_loss: 1.8434e-04 - val_semantic_2_loss: 1.1245e-04 - lr: 2.4242e-05\n",
      "Epoch 143/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7535e-04 - semantic_0_loss: 7.1120e-05 - semantic_1_loss: 8.7889e-05 - semantic_2_loss: 1.1634e-04\n",
      "Epoch 00143: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 251ms/step - loss: 2.7535e-04 - semantic_0_loss: 7.1120e-05 - semantic_1_loss: 8.7889e-05 - semantic_2_loss: 1.1634e-04 - val_loss: 5.7045e-04 - val_semantic_0_loss: 2.8052e-04 - val_semantic_1_loss: 1.8166e-04 - val_semantic_2_loss: 1.0826e-04 - lr: 2.3999e-05\n",
      "Epoch 144/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7341e-04 - semantic_0_loss: 7.0122e-05 - semantic_1_loss: 8.8620e-05 - semantic_2_loss: 1.1467e-04\n",
      "Epoch 00144: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 2.7341e-04 - semantic_0_loss: 7.0122e-05 - semantic_1_loss: 8.8620e-05 - semantic_2_loss: 1.1467e-04 - val_loss: 5.9128e-04 - val_semantic_0_loss: 2.8765e-04 - val_semantic_1_loss: 1.8492e-04 - val_semantic_2_loss: 1.1871e-04 - lr: 2.3759e-05\n",
      "Epoch 145/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7595e-04 - semantic_0_loss: 7.1241e-05 - semantic_1_loss: 8.9142e-05 - semantic_2_loss: 1.1556e-04\n",
      "Epoch 00145: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 2.7595e-04 - semantic_0_loss: 7.1241e-05 - semantic_1_loss: 8.9142e-05 - semantic_2_loss: 1.1556e-04 - val_loss: 5.6812e-04 - val_semantic_0_loss: 2.7804e-04 - val_semantic_1_loss: 1.7893e-04 - val_semantic_2_loss: 1.1115e-04 - lr: 2.3522e-05\n",
      "Epoch 146/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7625e-04 - semantic_0_loss: 7.3317e-05 - semantic_1_loss: 8.8055e-05 - semantic_2_loss: 1.1487e-04\n",
      "Epoch 00146: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 2.7625e-04 - semantic_0_loss: 7.3317e-05 - semantic_1_loss: 8.8055e-05 - semantic_2_loss: 1.1487e-04 - val_loss: 5.6166e-04 - val_semantic_0_loss: 2.7368e-04 - val_semantic_1_loss: 1.7922e-04 - val_semantic_2_loss: 1.0876e-04 - lr: 2.3286e-05\n",
      "Epoch 147/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7574e-04 - semantic_0_loss: 7.1320e-05 - semantic_1_loss: 8.9329e-05 - semantic_2_loss: 1.1510e-04\n",
      "Epoch 00147: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.7574e-04 - semantic_0_loss: 7.1320e-05 - semantic_1_loss: 8.9329e-05 - semantic_2_loss: 1.1510e-04 - val_loss: 5.5162e-04 - val_semantic_0_loss: 2.6056e-04 - val_semantic_1_loss: 1.8253e-04 - val_semantic_2_loss: 1.0853e-04 - lr: 2.3054e-05\n",
      "Epoch 148/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7892e-04 - semantic_0_loss: 7.3785e-05 - semantic_1_loss: 8.9905e-05 - semantic_2_loss: 1.1522e-04\n",
      "Epoch 00148: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 2.7892e-04 - semantic_0_loss: 7.3785e-05 - semantic_1_loss: 8.9905e-05 - semantic_2_loss: 1.1522e-04 - val_loss: 5.7751e-04 - val_semantic_0_loss: 2.9017e-04 - val_semantic_1_loss: 1.7965e-04 - val_semantic_2_loss: 1.0769e-04 - lr: 2.2823e-05\n",
      "Epoch 149/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7881e-04 - semantic_0_loss: 7.1848e-05 - semantic_1_loss: 9.0868e-05 - semantic_2_loss: 1.1610e-04\n",
      "Epoch 00149: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.7881e-04 - semantic_0_loss: 7.1848e-05 - semantic_1_loss: 9.0868e-05 - semantic_2_loss: 1.1610e-04 - val_loss: 5.5307e-04 - val_semantic_0_loss: 2.5988e-04 - val_semantic_1_loss: 1.8387e-04 - val_semantic_2_loss: 1.0932e-04 - lr: 2.2595e-05\n",
      "Epoch 150/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7625e-04 - semantic_0_loss: 7.1623e-05 - semantic_1_loss: 8.9264e-05 - semantic_2_loss: 1.1537e-04\n",
      "Epoch 00150: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.7625e-04 - semantic_0_loss: 7.1623e-05 - semantic_1_loss: 8.9264e-05 - semantic_2_loss: 1.1537e-04 - val_loss: 5.9133e-04 - val_semantic_0_loss: 2.9428e-04 - val_semantic_1_loss: 1.8647e-04 - val_semantic_2_loss: 1.1058e-04 - lr: 2.2369e-05\n",
      "Epoch 151/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7685e-04 - semantic_0_loss: 7.1765e-05 - semantic_1_loss: 8.9498e-05 - semantic_2_loss: 1.1559e-04\n",
      "Epoch 00151: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 2.7685e-04 - semantic_0_loss: 7.1765e-05 - semantic_1_loss: 8.9498e-05 - semantic_2_loss: 1.1559e-04 - val_loss: 5.4960e-04 - val_semantic_0_loss: 2.5381e-04 - val_semantic_1_loss: 1.8624e-04 - val_semantic_2_loss: 1.0955e-04 - lr: 2.2145e-05\n",
      "Epoch 152/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7188e-04 - semantic_0_loss: 6.9625e-05 - semantic_1_loss: 8.7050e-05 - semantic_2_loss: 1.1521e-04\n",
      "Epoch 00152: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.7188e-04 - semantic_0_loss: 6.9625e-05 - semantic_1_loss: 8.7050e-05 - semantic_2_loss: 1.1521e-04 - val_loss: 6.0551e-04 - val_semantic_0_loss: 3.0677e-04 - val_semantic_1_loss: 1.8486e-04 - val_semantic_2_loss: 1.1388e-04 - lr: 2.1924e-05\n",
      "Epoch 153/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7311e-04 - semantic_0_loss: 6.9046e-05 - semantic_1_loss: 8.7997e-05 - semantic_2_loss: 1.1607e-04\n",
      "Epoch 00153: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 251ms/step - loss: 2.7311e-04 - semantic_0_loss: 6.9046e-05 - semantic_1_loss: 8.7997e-05 - semantic_2_loss: 1.1607e-04 - val_loss: 5.5150e-04 - val_semantic_0_loss: 2.6492e-04 - val_semantic_1_loss: 1.7856e-04 - val_semantic_2_loss: 1.0802e-04 - lr: 2.1704e-05\n",
      "Epoch 154/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7307e-04 - semantic_0_loss: 6.9951e-05 - semantic_1_loss: 8.7865e-05 - semantic_2_loss: 1.1525e-04\n",
      "Epoch 00154: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 2.7307e-04 - semantic_0_loss: 6.9951e-05 - semantic_1_loss: 8.7865e-05 - semantic_2_loss: 1.1525e-04 - val_loss: 5.4272e-04 - val_semantic_0_loss: 2.4809e-04 - val_semantic_1_loss: 1.8450e-04 - val_semantic_2_loss: 1.1013e-04 - lr: 2.1487e-05\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 155/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7532e-04 - semantic_0_loss: 7.0807e-05 - semantic_1_loss: 8.9146e-05 - semantic_2_loss: 1.1537e-04\n",
      "Epoch 00155: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 2.7532e-04 - semantic_0_loss: 7.0807e-05 - semantic_1_loss: 8.9146e-05 - semantic_2_loss: 1.1537e-04 - val_loss: 5.5315e-04 - val_semantic_0_loss: 2.6548e-04 - val_semantic_1_loss: 1.7918e-04 - val_semantic_2_loss: 1.0849e-04 - lr: 2.1273e-05\n",
      "Epoch 156/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7054e-04 - semantic_0_loss: 6.8841e-05 - semantic_1_loss: 8.7347e-05 - semantic_2_loss: 1.1436e-04\n",
      "Epoch 00156: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 244ms/step - loss: 2.7054e-04 - semantic_0_loss: 6.8841e-05 - semantic_1_loss: 8.7347e-05 - semantic_2_loss: 1.1436e-04 - val_loss: 5.6491e-04 - val_semantic_0_loss: 2.6852e-04 - val_semantic_1_loss: 1.8438e-04 - val_semantic_2_loss: 1.1200e-04 - lr: 2.1060e-05\n",
      "Epoch 157/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7071e-04 - semantic_0_loss: 6.9178e-05 - semantic_1_loss: 8.7278e-05 - semantic_2_loss: 1.1425e-04\n",
      "Epoch 00157: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.7071e-04 - semantic_0_loss: 6.9178e-05 - semantic_1_loss: 8.7278e-05 - semantic_2_loss: 1.1425e-04 - val_loss: 5.3995e-04 - val_semantic_0_loss: 2.5736e-04 - val_semantic_1_loss: 1.7645e-04 - val_semantic_2_loss: 1.0615e-04 - lr: 2.0849e-05\n",
      "Epoch 158/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7355e-04 - semantic_0_loss: 7.0613e-05 - semantic_1_loss: 8.8130e-05 - semantic_2_loss: 1.1480e-04\n",
      "Epoch 00158: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 2.7355e-04 - semantic_0_loss: 7.0613e-05 - semantic_1_loss: 8.8130e-05 - semantic_2_loss: 1.1480e-04 - val_loss: 5.6166e-04 - val_semantic_0_loss: 2.6227e-04 - val_semantic_1_loss: 1.8394e-04 - val_semantic_2_loss: 1.1545e-04 - lr: 2.0641e-05\n",
      "Epoch 159/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7199e-04 - semantic_0_loss: 6.9699e-05 - semantic_1_loss: 8.7267e-05 - semantic_2_loss: 1.1502e-04\n",
      "Epoch 00159: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 2.7199e-04 - semantic_0_loss: 6.9699e-05 - semantic_1_loss: 8.7267e-05 - semantic_2_loss: 1.1502e-04 - val_loss: 5.8309e-04 - val_semantic_0_loss: 2.8967e-04 - val_semantic_1_loss: 1.8237e-04 - val_semantic_2_loss: 1.1105e-04 - lr: 2.0434e-05\n",
      "Epoch 160/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7204e-04 - semantic_0_loss: 6.8477e-05 - semantic_1_loss: 8.8305e-05 - semantic_2_loss: 1.1526e-04\n",
      "Epoch 00160: val_loss did not improve from 0.00054\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 2.7204e-04 - semantic_0_loss: 6.8477e-05 - semantic_1_loss: 8.8305e-05 - semantic_2_loss: 1.1526e-04 - val_loss: 5.5123e-04 - val_semantic_0_loss: 2.5439e-04 - val_semantic_1_loss: 1.8277e-04 - val_semantic_2_loss: 1.1407e-04 - lr: 2.0230e-05\n",
      "Epoch 161/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6777e-04 - semantic_0_loss: 6.6103e-05 - semantic_1_loss: 8.6274e-05 - semantic_2_loss: 1.1540e-04\n",
      "Epoch 00161: val_loss improved from 0.00054 to 0.00053, saving model to nuclear2_deep_watershed.h5\n",
      "204/204 [==============================] - 56s 275ms/step - loss: 2.6777e-04 - semantic_0_loss: 6.6103e-05 - semantic_1_loss: 8.6274e-05 - semantic_2_loss: 1.1540e-04 - val_loss: 5.2772e-04 - val_semantic_0_loss: 2.3882e-04 - val_semantic_1_loss: 1.8098e-04 - val_semantic_2_loss: 1.0792e-04 - lr: 2.0028e-05\n",
      "Epoch 162/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6651e-04 - semantic_0_loss: 6.7324e-05 - semantic_1_loss: 8.5550e-05 - semantic_2_loss: 1.1364e-04\n",
      "Epoch 00162: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 2.6651e-04 - semantic_0_loss: 6.7324e-05 - semantic_1_loss: 8.5550e-05 - semantic_2_loss: 1.1364e-04 - val_loss: 5.5949e-04 - val_semantic_0_loss: 2.6708e-04 - val_semantic_1_loss: 1.8179e-04 - val_semantic_2_loss: 1.1061e-04 - lr: 1.9827e-05\n",
      "Epoch 163/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6483e-04 - semantic_0_loss: 6.6828e-05 - semantic_1_loss: 8.4829e-05 - semantic_2_loss: 1.1318e-04\n",
      "Epoch 00163: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 2.6483e-04 - semantic_0_loss: 6.6828e-05 - semantic_1_loss: 8.4829e-05 - semantic_2_loss: 1.1318e-04 - val_loss: 5.6749e-04 - val_semantic_0_loss: 2.7730e-04 - val_semantic_1_loss: 1.8181e-04 - val_semantic_2_loss: 1.0838e-04 - lr: 1.9629e-05\n",
      "Epoch 164/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7010e-04 - semantic_0_loss: 6.7126e-05 - semantic_1_loss: 8.7957e-05 - semantic_2_loss: 1.1502e-04\n",
      "Epoch 00164: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 50s 244ms/step - loss: 2.7010e-04 - semantic_0_loss: 6.7126e-05 - semantic_1_loss: 8.7957e-05 - semantic_2_loss: 1.1502e-04 - val_loss: 5.6835e-04 - val_semantic_0_loss: 2.7196e-04 - val_semantic_1_loss: 1.8741e-04 - val_semantic_2_loss: 1.0898e-04 - lr: 1.9433e-05\n",
      "Epoch 165/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6843e-04 - semantic_0_loss: 6.8349e-05 - semantic_1_loss: 8.5858e-05 - semantic_2_loss: 1.1422e-04\n",
      "Epoch 00165: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 2.6843e-04 - semantic_0_loss: 6.8349e-05 - semantic_1_loss: 8.5858e-05 - semantic_2_loss: 1.1422e-04 - val_loss: 5.6581e-04 - val_semantic_0_loss: 2.6422e-04 - val_semantic_1_loss: 1.8639e-04 - val_semantic_2_loss: 1.1519e-04 - lr: 1.9239e-05\n",
      "Epoch 166/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7040e-04 - semantic_0_loss: 6.9873e-05 - semantic_1_loss: 8.7109e-05 - semantic_2_loss: 1.1342e-04\n",
      "Epoch 00166: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 50s 246ms/step - loss: 2.7040e-04 - semantic_0_loss: 6.9873e-05 - semantic_1_loss: 8.7109e-05 - semantic_2_loss: 1.1342e-04 - val_loss: 5.5067e-04 - val_semantic_0_loss: 2.5883e-04 - val_semantic_1_loss: 1.8043e-04 - val_semantic_2_loss: 1.1141e-04 - lr: 1.9046e-05\n",
      "Epoch 167/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6421e-04 - semantic_0_loss: 6.6604e-05 - semantic_1_loss: 8.3361e-05 - semantic_2_loss: 1.1425e-04\n",
      "Epoch 00167: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 2.6421e-04 - semantic_0_loss: 6.6604e-05 - semantic_1_loss: 8.3361e-05 - semantic_2_loss: 1.1425e-04 - val_loss: 5.7298e-04 - val_semantic_0_loss: 2.7520e-04 - val_semantic_1_loss: 1.8486e-04 - val_semantic_2_loss: 1.1293e-04 - lr: 1.8856e-05\n",
      "Epoch 168/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6709e-04 - semantic_0_loss: 6.7087e-05 - semantic_1_loss: 8.5638e-05 - semantic_2_loss: 1.1436e-04\n",
      "Epoch 00168: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 2.6709e-04 - semantic_0_loss: 6.7087e-05 - semantic_1_loss: 8.5638e-05 - semantic_2_loss: 1.1436e-04 - val_loss: 5.5348e-04 - val_semantic_0_loss: 2.5589e-04 - val_semantic_1_loss: 1.8511e-04 - val_semantic_2_loss: 1.1249e-04 - lr: 1.8667e-05\n",
      "Epoch 169/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6102e-04 - semantic_0_loss: 6.3183e-05 - semantic_1_loss: 8.4482e-05 - semantic_2_loss: 1.1336e-04\n",
      "Epoch 00169: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 2.6102e-04 - semantic_0_loss: 6.3183e-05 - semantic_1_loss: 8.4482e-05 - semantic_2_loss: 1.1336e-04 - val_loss: 5.3433e-04 - val_semantic_0_loss: 2.4469e-04 - val_semantic_1_loss: 1.7926e-04 - val_semantic_2_loss: 1.1038e-04 - lr: 1.8480e-05\n",
      "Epoch 170/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6960e-04 - semantic_0_loss: 6.8704e-05 - semantic_1_loss: 8.6602e-05 - semantic_2_loss: 1.1429e-04\n",
      "Epoch 00170: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 2.6960e-04 - semantic_0_loss: 6.8704e-05 - semantic_1_loss: 8.6602e-05 - semantic_2_loss: 1.1429e-04 - val_loss: 6.0696e-04 - val_semantic_0_loss: 3.1566e-04 - val_semantic_1_loss: 1.7645e-04 - val_semantic_2_loss: 1.1486e-04 - lr: 1.8296e-05\n",
      "Epoch 171/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6212e-04 - semantic_0_loss: 6.6178e-05 - semantic_1_loss: 8.2929e-05 - semantic_2_loss: 1.1301e-04\n",
      "Epoch 00171: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 51s 249ms/step - loss: 2.6212e-04 - semantic_0_loss: 6.6178e-05 - semantic_1_loss: 8.2929e-05 - semantic_2_loss: 1.1301e-04 - val_loss: 5.3978e-04 - val_semantic_0_loss: 2.5940e-04 - val_semantic_1_loss: 1.7343e-04 - val_semantic_2_loss: 1.0695e-04 - lr: 1.8113e-05\n",
      "Epoch 172/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6604e-04 - semantic_0_loss: 6.5003e-05 - semantic_1_loss: 8.6234e-05 - semantic_2_loss: 1.1480e-04\n",
      "Epoch 00172: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 50s 247ms/step - loss: 2.6604e-04 - semantic_0_loss: 6.5003e-05 - semantic_1_loss: 8.6234e-05 - semantic_2_loss: 1.1480e-04 - val_loss: 5.6612e-04 - val_semantic_0_loss: 2.6695e-04 - val_semantic_1_loss: 1.8446e-04 - val_semantic_2_loss: 1.1472e-04 - lr: 1.7932e-05\n",
      "Epoch 173/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6473e-04 - semantic_0_loss: 6.5707e-05 - semantic_1_loss: 8.5242e-05 - semantic_2_loss: 1.1378e-04\n",
      "Epoch 00173: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 51s 247ms/step - loss: 2.6473e-04 - semantic_0_loss: 6.5707e-05 - semantic_1_loss: 8.5242e-05 - semantic_2_loss: 1.1378e-04 - val_loss: 5.6436e-04 - val_semantic_0_loss: 2.7417e-04 - val_semantic_1_loss: 1.8137e-04 - val_semantic_2_loss: 1.0882e-04 - lr: 1.7752e-05\n",
      "Epoch 174/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.7256e-04 - semantic_0_loss: 6.8775e-05 - semantic_1_loss: 8.7914e-05 - semantic_2_loss: 1.1587e-04\n",
      "Epoch 00174: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 2.7256e-04 - semantic_0_loss: 6.8775e-05 - semantic_1_loss: 8.7914e-05 - semantic_2_loss: 1.1587e-04 - val_loss: 5.3564e-04 - val_semantic_0_loss: 2.4650e-04 - val_semantic_1_loss: 1.7651e-04 - val_semantic_2_loss: 1.1263e-04 - lr: 1.7575e-05\n",
      "Epoch 175/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6508e-04 - semantic_0_loss: 6.6125e-05 - semantic_1_loss: 8.5040e-05 - semantic_2_loss: 1.1391e-04\n",
      "Epoch 00175: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 51s 248ms/step - loss: 2.6508e-04 - semantic_0_loss: 6.6125e-05 - semantic_1_loss: 8.5040e-05 - semantic_2_loss: 1.1391e-04 - val_loss: 5.7106e-04 - val_semantic_0_loss: 2.8157e-04 - val_semantic_1_loss: 1.7752e-04 - val_semantic_2_loss: 1.1196e-04 - lr: 1.7399e-05\n",
      "Epoch 176/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6655e-04 - semantic_0_loss: 6.7174e-05 - semantic_1_loss: 8.5560e-05 - semantic_2_loss: 1.1382e-04\n",
      "Epoch 00176: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 51s 250ms/step - loss: 2.6655e-04 - semantic_0_loss: 6.7174e-05 - semantic_1_loss: 8.5560e-05 - semantic_2_loss: 1.1382e-04 - val_loss: 5.9238e-04 - val_semantic_0_loss: 2.9106e-04 - val_semantic_1_loss: 1.8568e-04 - val_semantic_2_loss: 1.1563e-04 - lr: 1.7225e-05\n",
      "Epoch 177/200\n",
      "204/204 [==============================] - ETA: 0s - loss: 2.6495e-04 - semantic_0_loss: 6.6742e-05 - semantic_1_loss: 8.4731e-05 - semantic_2_loss: 1.1348e-04\n",
      "Epoch 00177: val_loss did not improve from 0.00053\n",
      "204/204 [==============================] - 50s 245ms/step - loss: 2.6495e-04 - semantic_0_loss: 6.6742e-05 - semantic_1_loss: 8.4731e-05 - semantic_2_loss: 1.1348e-04 - val_loss: 5.3779e-04 - val_semantic_0_loss: 2.4493e-04 - val_semantic_1_loss: 1.8254e-04 - val_semantic_2_loss: 1.1032e-04 - lr: 1.7053e-05\n",
      "Epoch 178/200\n",
      "111/204 [===============>..............] - ETA: 20s - loss: 2.5886e-04 - semantic_0_loss: 6.3820e-05 - semantic_1_loss: 8.3140e-05 - semantic_2_loss: 1.1190e-04"
     ]
    }
   ],
   "source": [
    "# fit the model\n",
    "\n",
    "\n",
    "print('Training on', count_gpus(), 'GPUs.')\n",
    "\n",
    "train_callbacks = get_callbacks(\n",
    "    model_path,\n",
    "    lr_sched=lr_sched,\n",
    "    monitor='val_loss',\n",
    "    verbose=1)\n",
    "\n",
    "loss_history = model.fit(\n",
    "    train_data,\n",
    "    steps_per_epoch=train_data.y.shape[0] // batch_size,\n",
    "    epochs=200,\n",
    "    validation_data=val_data,\n",
    "    validation_steps=val_data.y.shape[0] // batch_size,\n",
    "    callbacks=train_callbacks)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71f3f44b",
   "metadata": {},
   "source": [
    "<B>Make predictions on Nuclear test dataset.</B> "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "dbf36df1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# create prediction model\n",
    "\n",
    "prediction_model = PanopticNet(\n",
    "    backbone='resnet50',\n",
    "    norm_method='whole_image',\n",
    "    num_semantic_classes=[1, 1], # inner distance, outer distance\n",
    "    input_shape=X_test.shape[1:]\n",
    ")\n",
    "\n",
    "prediction_model.load_weights(model_path, by_name=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "41bdf527",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-10-30 15:09:35.516526: I tensorflow/stream_executor/cuda/cuda_dnn.cc:366] Loaded cuDNN version 8202\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Watershed segmentation of shape (432, 256, 256, 1) in 23.360511504113674 seconds.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/app/software/DeepCell/0.11.1-foss-2021b-CUDA-11.4.1/lib/python3.9/site-packages/deepcell_toolbox/deep_watershed.py:179: FutureWarning: `selem` is a deprecated argument name for `h_maxima`. It will be removed in version 1.0.Please use `footprint` instead.\n",
      "  markers = h_maxima(image=maxima,\n"
     ]
    }
   ],
   "source": [
    "# make predictions on testing data\n",
    "\n",
    "start = default_timer()\n",
    "test_images = prediction_model.predict(X_test)\n",
    "watershed_time = default_timer() - start\n",
    "\n",
    "print('Watershed segmentation of shape', test_images[0].shape,\n",
    "      'in', watershed_time, 'seconds.')\n",
    "\n",
    "\n",
    "masks = deep_watershed(\n",
    "    test_images,\n",
    "    min_distance=10,\n",
    "    detection_threshold=0.1,\n",
    "    distance_threshold=0.01,\n",
    "    exclude_border=False,\n",
    "    small_objects_threshold=0)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "6d476b36",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(432, 256, 256, 1)\n",
      "(432, 256, 256, 1)\n"
     ]
    }
   ],
   "source": [
    "io.imshow(X_test[1,:,:,0])\n",
    "plt.show()\n",
    "io.imshow(y_test[1,:,:,0])\n",
    "plt.show()\n",
    "print(X_test.shape)\n",
    "print(y_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "d70549c9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.982675648953805\n"
     ]
    }
   ],
   "source": [
    "APs = [syotil.csi(y_test[i,:,:,0], masks[i,:,:,0]) for i in range(y_test.shape[0])]\n",
    "print(np.nanmean(APs))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42277d43",
   "metadata": {},
   "source": [
    "**Make prediction on K's data.**<br>\n",
    "Using NuclearSegmentation allows setting image_mpp, which has a substantial influence on performance."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "37a766dc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('semantic_0', (None, 256, 256, 1)), ('semantic_1', (None, 256, 256, 1))]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from deepcell.applications import NuclearSegmentation\n",
    "\n",
    "app = NuclearSegmentation(prediction_model)\n",
    "[(layer.name, layer.output_shape) for layer in filter(lambda x: x.name.startswith('semantic_'), app.model.layers)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "350ad39f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1040, 233)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1040, 233)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# a rectangular patch that requires tiling\n",
    "im0 = io.imread('images/test/M872956_JML_Position8_CD3_test_img.png')\n",
    "print(im0.shape)\n",
    "im=im0\n",
    "io.imshow(im)\n",
    "plt.show()\n",
    "im = np.expand_dims(im, axis=-1)\n",
    "im = np.expand_dims(im, axis=0)\n",
    "\n",
    "mask_true=io.imread(\"images/test/M872956_JML_Position8_CD3_test_masks.png\")\n",
    "print(mask_true.shape)\n",
    "io.imshow(mask_true)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "e5750258",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(256, 256, 1)"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "app.model_image_shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "9057ac38",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1040, 233, 1)\n",
      "(1, 1824, 256, 1)\n",
      "{'padding': True, 'x_pad': (392, 392), 'y_pad': (11, 12)}\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.3165562913907285"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x=im\n",
    "y, tile_info = app._tile_input(im)\n",
    "print(x.shape)\n",
    "print(y.shape)\n",
    "print(tile_info)\n",
    "pred = app.predict(y, image_mpp=2)\n",
    "prd = app._untile_output(pred, tile_info)\n",
    "io.imshow(prd[0,:,:,0])\n",
    "plt.show()\n",
    "syotil.csi(mask_true, prd[0,:,:,0])\n",
    "# image_mpp=1: .2\n",
    "# image_mpp=2: .3\n",
    "# image_mpp=3 .1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "475b5ee7",
   "metadata": {},
   "source": [
    "**Additional testing with a square patch.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "964e6197",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(256, 256, 3)\n",
      "(256, 256)\n",
      "(1, 256, 256, 1)\n",
      "(1, 256, 256, 1)\n"
     ]
    }
   ],
   "source": [
    "# read CD3 data\n",
    "\n",
    "img=io.imread(\"images/square_patches/M872956_JML_Position8_CD3_img_patch256x256.png\")\n",
    "print(img.shape)\n",
    "X_cd3=img[:,:,2]\n",
    "#io.imshow(im)\n",
    "#plt.show()\n",
    "X_cd3=np.expand_dims(X_cd3, -1) \n",
    "X_cd3=np.expand_dims(X_cd3, 0) \n",
    "\n",
    "y_cd3=io.imread(\"images/square_patches/M872956_JML_Position8_CD3_masks_patch256x256.png\")\n",
    "print(y_cd3.shape)\n",
    "#io.imshow(mask_true)\n",
    "#plt.show()\n",
    "y_cd3=np.expand_dims(y_cd3, -1) \n",
    "y_cd3=np.expand_dims(y_cd3, 0) \n",
    "\n",
    "batch_size_cd3 = 1\n",
    "\n",
    "print(X_cd3.shape)\n",
    "print(y_cd3.shape)\n",
    "\n",
    "train_cd3 = datagen.flow(\n",
    "    {'X': X_cd3, 'y': y_cd3},\n",
    "    seed=seed,\n",
    "    transforms=transforms,\n",
    "    transforms_kwargs=transforms_kwargs,\n",
    "    min_objects=min_objects,\n",
    "    batch_size=batch_size)\n",
    "\n",
    "val_cd3 = datagen_val.flow(\n",
    "    {'X': X_cd3, 'y': y_cd3},\n",
    "    seed=seed,\n",
    "    transforms=transforms,\n",
    "    transforms_kwargs=transforms_kwargs,\n",
    "    min_objects=min_objects,\n",
    "    batch_size=batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "69e9643b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 256, 256, 1)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.2994011976047904"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y, tile_info = app._tile_input(X_cd3)\n",
    "print(y.shape)\n",
    "#print(tile_info)\n",
    "pred = app.predict(y, image_mpp=1)  #, image_mpp=1 0.299, with default image_mpp 0.65, it is 0.09\n",
    "prd = app._untile_output(pred, tile_info)\n",
    "io.imshow(prd[0,:,:,0])\n",
    "plt.show()\n",
    "syotil.csi(y_cd3[0,:,:,0], prd[0,:,:,0]) # "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b1de9891",
   "metadata": {},
   "source": [
    "**Predictions can also be made without Application.**<br>\n",
    "Don't set min_distance to 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "d868286f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.08620689655172414"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# make predictions on CD3 data\n",
    "start = default_timer()\n",
    "test_images = prediction_model.predict(X_cd3)\n",
    "watershed_time = default_timer() - start\n",
    "\n",
    "pred_cd3 = deep_watershed(\n",
    "    test_images,\n",
    "    min_distance=10, # this should not be set to 0, https://github.com/vanvalenlab/deepcell-toolbox/blob/master/deepcell_toolbox/deep_watershed.py\n",
    "    detection_threshold=0.001,\n",
    "    distance_threshold=0.0,\n",
    "    exclude_border=False,\n",
    "    small_objects_threshold=0)\n",
    "\n",
    "syotil.csi(y_cd3[0,:,:,0], pred_cd3[0,:,:,0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "07cd889a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#print(y_cd3.shape, pred_cd3.shape)\n",
    "# io.imshow(y_cd3[0,:,:,0])\n",
    "# plt.show()\n",
    "io.imshow(pred_cd3[0,:,:,0])\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}