MobilenetV2DeeplabRizan.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "MobilenetV2DeeplabRizan.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python2",
      "display_name": "Python 2"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/RizanPSTU/Python-LCC-data/blob/master/MobilenetV2DeeplabRizan.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T-Il7_pZX0Fl",
        "colab_type": "code",
        "outputId": "a87ae628-77b4-49bf-fec2-be9a18185db8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "ls"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[01;34msample_data\u001b[0m/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EpkBMzeRoq7X",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import sys"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GX8tiwYhowxL",
        "colab_type": "code",
        "outputId": "c25e8103-6326-4eaa-aae5-638a5515015b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "print(sys.version)"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2.7.15rc1 (default, Nov 12 2018, 14:31:15) \n",
            "[GCC 7.3.0]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4BkAsgcKX38g",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#!pip install tensorflow-gpu==1.13.1"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sveDc68sX-Rc",
        "colab_type": "code",
        "outputId": "5a996991-10c4-4b27-cf21-af324201745d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "#!pip install tensorflow==1.13.1\n",
        "import tensorflow as tf\n",
        "print(tf.__version__)"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1.13.1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qYG8p8z8YBYs",
        "colab_type": "code",
        "outputId": "3e274ce2-e9a8-4802-bfa6-99a8e46c0b55",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 170
        }
      },
      "source": [
        "%cd\n",
        "!git clone https://github.com/RizanPSTU/models.git"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/root\n",
            "Cloning into 'models'...\n",
            "remote: Enumerating objects: 110, done.\u001b[K\n",
            "remote: Counting objects: 100% (110/110), done.\u001b[K\n",
            "remote: Compressing objects: 100% (100/100), done.\u001b[K\n",
            "remote: Total 25983 (delta 70), reused 21 (delta 10), pack-reused 25873\u001b[K\n",
            "Receiving objects: 100% (25983/25983), 508.57 MiB | 41.19 MiB/s, done.\n",
            "Resolving deltas: 100% (15601/15601), done.\n",
            "Checking out files: 100% (2907/2907), done.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "R9zzHsHKLhFJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#from google.colab import drive\n",
        "#drive.mount('/content/gdrive')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nCDZ908SLl2E",
        "colab_type": "code",
        "outputId": "0ee1a199-d01a-4e06-f518-8b4de4ef2948",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "ls"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[01;34mmodels\u001b[0m/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qWLFwHVWYKO0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#%cd models/research/\n",
        "# From tensorflow/models/research/\n",
        "#!export PYTHONPATH=$PYTHONPATH:`pwd`:`pwd`/slim"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DvwnKqjwYNX0",
        "colab_type": "code",
        "outputId": "07c5ebd9-e258-4b6d-b870-0792fa095757",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "%cd"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/root\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k9vZrMJhYSM8",
        "colab_type": "code",
        "outputId": "3707f500-d81b-4b82-a4dc-e76200ca5a1a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "ls"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[01;34mmodels\u001b[0m/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2GjPZ2raYTP9",
        "colab_type": "code",
        "outputId": "233e2291-a9a1-4504-a2f9-571b4838dfcd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 73239
        }
      },
      "source": [
        "%cd models/research/deeplab/\n",
        "!sh ./local_test_mobilenetv2.sh"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/root/models/research/deeplab\n",
            "\n",
            "WARNING: The TensorFlow contrib module will not be included in TensorFlow 2.0.\n",
            "For more information, please see:\n",
            "  * https://github.com/tensorflow/community/blob/master/rfcs/20180907-contrib-sunset.md\n",
            "  * https://github.com/tensorflow/addons\n",
            "If you depend on functionality not listed there, please file an issue.\n",
            "\n",
            "testBuildDeepLabWithDensePredictionCell (__main__.DeeplabModelTest) ... WARNING:tensorflow:From /usr/lib/python2.7/contextlib.py:84: test_session (from tensorflow.python.framework.test_util) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use `self.session()` or `self.cached_session()` instead.\n",
            "2019-05-18 11:46:02.704200: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2300000000 Hz\n",
            "2019-05-18 11:46:02.704505: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x55b4aed80c00 executing computations on platform Host. Devices:\n",
            "2019-05-18 11:46:02.704537: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): <undefined>, <undefined>\n",
            "2019-05-18 11:46:02.968521: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:998] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
            "2019-05-18 11:46:02.969076: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x55b4aed81c80 executing computations on platform CUDA. Devices:\n",
            "2019-05-18 11:46:02.969115: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): Tesla T4, Compute Capability 7.5\n",
            "2019-05-18 11:46:02.969466: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties: \n",
            "name: Tesla T4 major: 7 minor: 5 memoryClockRate(GHz): 1.59\n",
            "pciBusID: 0000:00:04.0\n",
            "totalMemory: 14.73GiB freeMemory: 14.60GiB\n",
            "2019-05-18 11:46:02.969490: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:04.269908: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:04.269989: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:04.270004: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:04.270270: W tensorflow/core/common_runtime/gpu/gpu_bfc_allocator.cc:42] Overriding allow_growth setting because the TF_FORCE_GPU_ALLOW_GROWTH environment variable is set. Original config value was 0.\n",
            "2019-05-18 11:46:04.270368: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4523 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "WARNING:tensorflow:From /root/models/research/deeplab/core/feature_extractor.py:196: to_float (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n",
            "ok\n",
            "testBuildDeepLabv2 (__main__.DeeplabModelTest) ... 2019-05-18 11:46:06.217222: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:06.217316: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:06.217334: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:06.217347: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:06.217780: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4523 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "2019-05-18 11:46:10.981748: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:10.981832: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:10.981848: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:10.981861: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:10.982132: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4523 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "2019-05-18 11:46:17.440431: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:17.440533: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:17.440548: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:17.440561: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:17.440888: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4523 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "2019-05-18 11:46:19.536195: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:19.536300: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:19.536315: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:19.536328: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:19.536670: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4523 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "ok\n",
            "testForwardpassDeepLabv3plus (__main__.DeeplabModelTest) ... 2019-05-18 11:46:22.120665: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:22.120744: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:22.120760: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:22.120773: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:22.121037: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4523 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "ok\n",
            "testWrongDeepLabVariant (__main__.DeeplabModelTest) ... ok\n",
            "test_session (__main__.DeeplabModelTest)\n",
            "Use cached_session instead. (deprecated) ... skipped 'Not a test.'\n",
            "\n",
            "----------------------------------------------------------------------\n",
            "Ran 5 tests in 22.071s\n",
            "\n",
            "OK (skipped=1)\n",
            "Downloading VOCtrainval_11-May-2012.tar to ./pascal_voc_seg\n",
            "  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current\n",
            "                                 Dload  Upload   Total   Spent    Left  Speed\n",
            "100   388    0   388    0     0     38      0 --:--:--  0:00:10 --:--:--   100\n",
            "100 62.5M    0 62.5M    0     0  5208k      0 --:--:--  0:00:12 --:--:-- 93.0M\n",
            "Uncompressing VOCtrainval_11-May-2012.tar\n",
            "Removing the color map in ground truth annotations...\n",
            "Converting PASCAL VOC 2012 dataset...\n",
            "2019-05-18 11:46:49.111760: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2300000000 Hz\n",
            "2019-05-18 11:46:49.111986: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x55b7700f3600 executing computations on platform Host. Devices:\n",
            "2019-05-18 11:46:49.112018: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): <undefined>, <undefined>\n",
            "2019-05-18 11:46:49.297972: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:998] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
            "2019-05-18 11:46:49.298522: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x55b7700f31e0 executing computations on platform CUDA. Devices:\n",
            "2019-05-18 11:46:49.298553: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): Tesla T4, Compute Capability 7.5\n",
            "2019-05-18 11:46:49.298944: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties: \n",
            "name: Tesla T4 major: 7 minor: 5 memoryClockRate(GHz): 1.59\n",
            "pciBusID: 0000:00:04.0\n",
            "totalMemory: 14.73GiB freeMemory: 14.60GiB\n",
            "2019-05-18 11:46:49.298972: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:49.793185: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:49.793267: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:49.793281: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:49.793552: W tensorflow/core/common_runtime/gpu/gpu_bfc_allocator.cc:42] Overriding allow_growth setting because the TF_FORCE_GPU_ALLOW_GROWTH environment variable is set. Original config value was 0.\n",
            "2019-05-18 11:46:49.793610: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14115 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "2019-05-18 11:46:49.797175: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:49.797223: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:49.797242: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:49.797253: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:49.797490: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14115 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            ">> Converting image 1/316 shard 0WARNING:tensorflow:From ./build_voc2012_data.py:116: __init__ (from tensorflow.python.platform.gfile) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.gfile.GFile.\n",
            ">> Converting image 79/316 shard 0\n",
            ">> Converting image 158/316 shard 1\n",
            ">> Converting image 237/316 shard 2\n",
            ">> Converting image 316/316 shard 3\n",
            "2019-05-18 11:46:50.759652: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:50.759711: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:50.759727: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:50.759739: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:50.760008: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14115 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "2019-05-18 11:46:50.761673: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:50.761710: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:50.761725: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:50.761737: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:50.761964: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14115 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            ">> Converting image 454/1816 shard 0\n",
            ">> Converting image 908/1816 shard 1\n",
            ">> Converting image 1362/1816 shard 2\n",
            ">> Converting image 1816/1816 shard 3\n",
            "2019-05-18 11:46:56.136056: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:56.136110: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:56.136125: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:56.136138: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:56.136412: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14115 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "2019-05-18 11:46:56.138068: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:46:56.138106: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:46:56.138121: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:46:56.138133: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:46:56.138360: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14115 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            ">> Converting image 375/1500 shard 0\n",
            ">> Converting image 750/1500 shard 1\n",
            ">> Converting image 1125/1500 shard 2\n",
            ">> Converting image 1500/1500 shard 3\n",
            "--2019-05-18 11:47:00--  http://download.tensorflow.org/models/deeplabv3_mnv2_pascal_train_aug_2018_01_29.tar.gz\n",
            "Resolving download.tensorflow.org (download.tensorflow.org)... 74.125.141.128, 2607:f8b0:400c:c06::80\n",
            "Connecting to download.tensorflow.org (download.tensorflow.org)|74.125.141.128|:80... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 23882985 (23M) [application/x-tar]\n",
            "Saving to: ‘deeplabv3_mnv2_pascal_train_aug_2018_01_29.tar.gz’\n",
            "\n",
            "deeplabv3_mnv2_pasc 100%[===================>]  22.78M  --.-KB/s    in 0.1s    \n",
            "\n",
            "2019-05-18 11:47:01 (166 MB/s) - ‘deeplabv3_mnv2_pascal_train_aug_2018_01_29.tar.gz’ saved [23882985/23882985]\n",
            "\n",
            "\n",
            "WARNING: The TensorFlow contrib module will not be included in TensorFlow 2.0.\n",
            "For more information, please see:\n",
            "  * https://github.com/tensorflow/community/blob/master/rfcs/20180907-contrib-sunset.md\n",
            "  * https://github.com/tensorflow/addons\n",
            "If you depend on functionality not listed there, please file an issue.\n",
            "\n",
            "INFO:tensorflow:Training on trainval set\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/control_flow_ops.py:423: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n",
            "WARNING:tensorflow:From /root/models/research/deeplab/core/preprocess_utils.py:203: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/training/learning_rate_decay_v2.py:321: div (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Deprecated in favor of operator or tf.math.divide.\n",
            "WARNING:tensorflow:From /root/models/research/deeplab/core/feature_extractor.py:196: to_float (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/keras/layers/core.py:143: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.\n",
            "WARNING:tensorflow:From /root/models/research/deeplab/train.py:418: Print (from tensorflow.python.ops.logging_ops) is deprecated and will be removed after 2018-08-20.\n",
            "Instructions for updating:\n",
            "Use tf.print instead of tf.Print. Note that tf.print returns a no-output operator that directly prints the output. Outside of defuns or eager mode, this operator will not be executed unless it is directly specified in session.run or used as a control dependency for other operators. This is only a concern in graph mode. Below is an example of how to ensure tf.print executes in graph mode:\n",
            "```python\n",
            "    sess = tf.Session()\n",
            "    with sess.as_default():\n",
            "        tensor = tf.range(10)\n",
            "        print_op = tf.print(tensor)\n",
            "        with tf.control_dependencies([print_op]):\n",
            "          out = tf.add(tensor, tensor)\n",
            "        sess.run(out)\n",
            "    ```\n",
            "Additionally, to use tf.print in python 2.7, users must make sure to import\n",
            "the following:\n",
            "\n",
            "  `from __future__ import print_function`\n",
            "\n",
            "INFO:tensorflow:Initializing model from path: /root/models/research/deeplab/datasets/pascal_voc_seg/init_models/deeplabv3_mnv2_pascal_train_aug/model.ckpt-30000\n",
            "INFO:tensorflow:Create CheckpointSaverHook.\n",
            "INFO:tensorflow:Graph was finalized.\n",
            "2019-05-18 11:47:11.163372: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2300000000 Hz\n",
            "2019-05-18 11:47:11.163606: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x56442d090f20 executing computations on platform Host. Devices:\n",
            "2019-05-18 11:47:11.163636: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): <undefined>, <undefined>\n",
            "2019-05-18 11:47:11.345052: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:998] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
            "2019-05-18 11:47:11.345579: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x56442d090c60 executing computations on platform CUDA. Devices:\n",
            "2019-05-18 11:47:11.345609: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): Tesla T4, Compute Capability 7.5\n",
            "2019-05-18 11:47:11.346000: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties: \n",
            "name: Tesla T4 major: 7 minor: 5 memoryClockRate(GHz): 1.59\n",
            "pciBusID: 0000:00:04.0\n",
            "totalMemory: 14.73GiB freeMemory: 14.60GiB\n",
            "2019-05-18 11:47:11.346029: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 11:47:11.835768: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 11:47:11.835848: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 11:47:11.835861: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 11:47:11.836141: W tensorflow/core/common_runtime/gpu/gpu_bfc_allocator.cc:42] Overriding allow_growth setting because the TF_FORCE_GPU_ALLOW_GROWTH environment variable is set. Original config value was 0.\n",
            "2019-05-18 11:47:11.836196: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14115 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "INFO:tensorflow:Running local_init_op.\n",
            "INFO:tensorflow:Done running local_init_op.\n",
            "INFO:tensorflow:Saving checkpoints for 0 into /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt.\n",
            "Total loss is :[1.97182631]\n",
            "Total loss is :[1.6965872]\n",
            "INFO:tensorflow:global_step/sec: 2.45509\n",
            "Total loss is :[1.94275701]\n",
            "INFO:tensorflow:global_step/sec: 3.98228\n",
            "Total loss is :[1.99201298]\n",
            "INFO:tensorflow:global_step/sec: 3.84622\n",
            "Total loss is :[1.74010789]\n",
            "INFO:tensorflow:global_step/sec: 3.73321\n",
            "Total loss is :[1.86184466]\n",
            "INFO:tensorflow:global_step/sec: 3.89887\n",
            "Total loss is :[1.83141768]\n",
            "INFO:tensorflow:global_step/sec: 3.99113\n",
            "Total loss is :[1.57875252]\n",
            "INFO:tensorflow:global_step/sec: 4.09931\n",
            "Total loss is :[1.58666456]\n",
            "INFO:tensorflow:global_step/sec: 3.90661\n",
            "Total loss is :[1.42494941]\n",
            "INFO:tensorflow:global_step/sec: 3.63308\n",
            "Total loss is :[1.63460135]\n",
            "INFO:tensorflow:global_step/sec: 3.86212\n",
            "Total loss is :[1.38292992]\n",
            "INFO:tensorflow:global_step/sec: 3.77372\n",
            "Total loss is :[1.46232939]\n",
            "INFO:tensorflow:global_step/sec: 3.54106\n",
            "Total loss is :[1.4302181]\n",
            "INFO:tensorflow:global_step/sec: 3.82601\n",
            "Total loss is :[1.20496237]\n",
            "INFO:tensorflow:global_step/sec: 3.72448\n",
            "Total loss is :[1.23169732]\n",
            "INFO:tensorflow:global_step/sec: 3.70915\n",
            "Total loss is :[1.29885626]\n",
            "INFO:tensorflow:global_step/sec: 3.94134\n",
            "Total loss is :[1.03172529]\n",
            "INFO:tensorflow:global_step/sec: 3.65774\n",
            "Total loss is :[1.02999878]\n",
            "INFO:tensorflow:global_step/sec: 3.82757\n",
            "Total loss is :[0.940396786]\n",
            "INFO:tensorflow:global_step/sec: 3.86723\n",
            "Total loss is :[1.08341384]\n",
            "INFO:tensorflow:global_step/sec: 3.72972\n",
            "Total loss is :[0.910629511]\n",
            "INFO:tensorflow:global_step/sec: 3.68704\n",
            "Total loss is :[0.794299781]\n",
            "INFO:tensorflow:global_step/sec: 3.68828\n",
            "Total loss is :[0.907747805]\n",
            "INFO:tensorflow:global_step/sec: 3.84411\n",
            "Total loss is :[0.925743401]\n",
            "INFO:tensorflow:global_step/sec: 3.7077\n",
            "Total loss is :[0.644994795]\n",
            "INFO:tensorflow:global_step/sec: 3.71717\n",
            "Total loss is :[0.821690559]\n",
            "INFO:tensorflow:global_step/sec: 3.73702\n",
            "Total loss is :[0.910626411]\n",
            "INFO:tensorflow:global_step/sec: 3.79339\n",
            "Total loss is :[0.66978246]\n",
            "INFO:tensorflow:global_step/sec: 3.72286\n",
            "Total loss is :[0.657754123]\n",
            "INFO:tensorflow:global_step/sec: 3.68421\n",
            "Total loss is :[0.621376038]\n",
            "INFO:tensorflow:global_step/sec: 3.88407\n",
            "Total loss is :[0.898867786]\n",
            "INFO:tensorflow:global_step/sec: 3.76281\n",
            "Total loss is :[0.797445297]\n",
            "INFO:tensorflow:global_step/sec: 3.77309\n",
            "Total loss is :[0.880799115]\n",
            "INFO:tensorflow:global_step/sec: 3.80857\n",
            "Total loss is :[0.487444609]\n",
            "INFO:tensorflow:global_step/sec: 3.67236\n",
            "Total loss is :[0.933359325]\n",
            "INFO:tensorflow:global_step/sec: 3.73187\n",
            "Total loss is :[0.638934374]\n",
            "INFO:tensorflow:global_step/sec: 3.69563\n",
            "Total loss is :[0.600791514]\n",
            "INFO:tensorflow:global_step/sec: 3.70901\n",
            "Total loss is :[0.543476224]\n",
            "INFO:tensorflow:global_step/sec: 3.89671\n",
            "Total loss is :[0.595339179]\n",
            "INFO:tensorflow:global_step/sec: 3.60411\n",
            "Total loss is :[0.455471635]\n",
            "INFO:tensorflow:global_step/sec: 3.42939\n",
            "Total loss is :[0.625200927]\n",
            "INFO:tensorflow:global_step/sec: 3.65017\n",
            "Total loss is :[0.893780887]\n",
            "INFO:tensorflow:global_step/sec: 3.7918\n",
            "Total loss is :[0.479433775]\n",
            "INFO:tensorflow:global_step/sec: 3.7377\n",
            "Total loss is :[0.479895234]\n",
            "INFO:tensorflow:global_step/sec: 3.68196\n",
            "Total loss is :[0.65074712]\n",
            "INFO:tensorflow:global_step/sec: 3.91666\n",
            "Total loss is :[0.423738331]\n",
            "INFO:tensorflow:global_step/sec: 4.03495\n",
            "Total loss is :[0.461569667]\n",
            "INFO:tensorflow:global_step/sec: 3.78922\n",
            "Total loss is :[0.494530588]\n",
            "INFO:tensorflow:global_step/sec: 3.72108\n",
            "Total loss is :[0.563483417]\n",
            "INFO:tensorflow:global_step/sec: 3.80124\n",
            "Total loss is :[0.812053621]\n",
            "INFO:tensorflow:global_step/sec: 3.7443\n",
            "Total loss is :[0.426277488]\n",
            "INFO:tensorflow:global_step/sec: 4.02252\n",
            "Total loss is :[0.509328187]\n",
            "INFO:tensorflow:global_step/sec: 3.71918\n",
            "Total loss is :[0.442103982]\n",
            "INFO:tensorflow:global_step/sec: 3.79461\n",
            "Total loss is :[0.622822344]\n",
            "INFO:tensorflow:global_step/sec: 3.64577\n",
            "Total loss is :[0.441793203]\n",
            "INFO:tensorflow:global_step/sec: 3.77162\n",
            "Total loss is :[0.539706588]\n",
            "INFO:tensorflow:global_step/sec: 3.80506\n",
            "Total loss is :[0.503447056]\n",
            "INFO:tensorflow:global_step/sec: 3.63489\n",
            "Total loss is :[0.672578037]\n",
            "INFO:tensorflow:global_step/sec: 3.82441\n",
            "Total loss is :[0.598698378]\n",
            "INFO:tensorflow:global_step/sec: 3.83716\n",
            "Total loss is :[0.459009498]\n",
            "INFO:tensorflow:global_step/sec: 3.61083\n",
            "Total loss is :[0.360868037]\n",
            "INFO:tensorflow:global_step/sec: 3.77814\n",
            "Total loss is :[0.344972342]\n",
            "INFO:tensorflow:global_step/sec: 3.65\n",
            "Total loss is :[0.436561763]\n",
            "INFO:tensorflow:global_step/sec: 3.62684\n",
            "Total loss is :[0.404322505]\n",
            "INFO:tensorflow:global_step/sec: 3.84286\n",
            "Total loss is :[0.668304741]\n",
            "INFO:tensorflow:global_step/sec: 3.98197\n",
            "Total loss is :[0.361849695]\n",
            "INFO:tensorflow:global_step/sec: 3.8285\n",
            "Total loss is :[0.32983318]\n",
            "INFO:tensorflow:global_step/sec: 3.66139\n",
            "Total loss is :[0.476248443]\n",
            "INFO:tensorflow:global_step/sec: 3.84499\n",
            "Total loss is :[0.512673438]\n",
            "INFO:tensorflow:global_step/sec: 3.96918\n",
            "Total loss is :[0.283066213]\n",
            "INFO:tensorflow:global_step/sec: 3.75653\n",
            "Total loss is :[0.457303494]\n",
            "INFO:tensorflow:global_step/sec: 3.48311\n",
            "Total loss is :[0.291887462]\n",
            "INFO:tensorflow:global_step/sec: 3.63158\n",
            "Total loss is :[0.322274953]\n",
            "INFO:tensorflow:global_step/sec: 3.81535\n",
            "Total loss is :[0.429797113]\n",
            "INFO:tensorflow:global_step/sec: 3.82336\n",
            "Total loss is :[0.445393533]\n",
            "INFO:tensorflow:global_step/sec: 3.81142\n",
            "Total loss is :[0.437563211]\n",
            "INFO:tensorflow:global_step/sec: 3.73878\n",
            "Total loss is :[0.47125119]\n",
            "INFO:tensorflow:global_step/sec: 3.76041\n",
            "Total loss is :[0.288172066]\n",
            "INFO:tensorflow:global_step/sec: 3.95725\n",
            "Total loss is :[0.456343651]\n",
            "INFO:tensorflow:global_step/sec: 3.73559\n",
            "Total loss is :[0.324334681]\n",
            "INFO:tensorflow:global_step/sec: 3.80455\n",
            "Total loss is :[0.431522697]\n",
            "INFO:tensorflow:global_step/sec: 3.67942\n",
            "Total loss is :[0.374766946]\n",
            "INFO:tensorflow:global_step/sec: 3.73961\n",
            "Total loss is :[0.541820228]\n",
            "INFO:tensorflow:global_step/sec: 3.76259\n",
            "Total loss is :[0.768884838]\n",
            "INFO:tensorflow:global_step/sec: 3.80449\n",
            "Total loss is :[0.318235934]\n",
            "INFO:tensorflow:global_step/sec: 3.78162\n",
            "Total loss is :[0.441124141]\n",
            "INFO:tensorflow:global_step/sec: 3.64304\n",
            "Total loss is :[0.272073984]\n",
            "INFO:tensorflow:global_step/sec: 3.91149\n",
            "Total loss is :[0.353292674]\n",
            "INFO:tensorflow:global_step/sec: 4.01397\n",
            "Total loss is :[0.430567682]\n",
            "INFO:tensorflow:global_step/sec: 3.63138\n",
            "Total loss is :[0.224123091]\n",
            "INFO:tensorflow:global_step/sec: 3.96073\n",
            "Total loss is :[0.383111179]\n",
            "INFO:tensorflow:global_step/sec: 3.73728\n",
            "Total loss is :[0.247039035]\n",
            "INFO:tensorflow:global_step/sec: 3.88246\n",
            "Total loss is :[0.452105492]\n",
            "INFO:tensorflow:global_step/sec: 3.77214\n",
            "Total loss is :[0.461135894]\n",
            "INFO:tensorflow:global_step/sec: 3.85154\n",
            "Total loss is :[0.375481963]\n",
            "INFO:tensorflow:global_step/sec: 3.82011\n",
            "Total loss is :[0.26224342]\n",
            "INFO:tensorflow:global_step/sec: 3.93342\n",
            "Total loss is :[0.440345764]\n",
            "INFO:tensorflow:global_step/sec: 3.9704\n",
            "Total loss is :[0.422533035]\n",
            "INFO:tensorflow:global_step/sec: 3.84358\n",
            "Total loss is :[0.396268904]\n",
            "INFO:tensorflow:global_step/sec: 3.80997\n",
            "Total loss is :[0.338829]\n",
            "INFO:tensorflow:global_step/sec: 3.55333\n",
            "Total loss is :[0.418192685]\n",
            "INFO:tensorflow:global_step/sec: 3.81574\n",
            "Total loss is :[0.344474792]\n",
            "INFO:tensorflow:global_step/sec: 3.73812\n",
            "Total loss is :[0.327417463]\n",
            "INFO:tensorflow:global_step/sec: 3.64656\n",
            "Total loss is :[0.423914552]\n",
            "INFO:tensorflow:global_step/sec: 3.94398\n",
            "Total loss is :[0.313591599]\n",
            "INFO:tensorflow:global_step/sec: 3.83313\n",
            "Total loss is :[0.451521248]\n",
            "INFO:tensorflow:global_step/sec: 3.65109\n",
            "Total loss is :[0.364644]\n",
            "INFO:tensorflow:global_step/sec: 3.74728\n",
            "Total loss is :[0.373864263]\n",
            "INFO:tensorflow:global_step/sec: 3.89995\n",
            "Total loss is :[0.308362305]\n",
            "INFO:tensorflow:global_step/sec: 3.62834\n",
            "Total loss is :[0.673058867]\n",
            "INFO:tensorflow:global_step/sec: 3.78204\n",
            "Total loss is :[0.377242506]\n",
            "INFO:tensorflow:global_step/sec: 3.64899\n",
            "Total loss is :[0.546634853]\n",
            "INFO:tensorflow:global_step/sec: 3.92761\n",
            "Total loss is :[0.454080403]\n",
            "INFO:tensorflow:global_step/sec: 3.63539\n",
            "Total loss is :[0.368814468]\n",
            "INFO:tensorflow:global_step/sec: 4.14622\n",
            "Total loss is :[0.400556087]\n",
            "INFO:tensorflow:global_step/sec: 3.75786\n",
            "Total loss is :[0.469407737]\n",
            "INFO:tensorflow:global_step/sec: 3.78734\n",
            "Total loss is :[0.333558589]\n",
            "INFO:tensorflow:global_step/sec: 3.8191\n",
            "Total loss is :[0.351160944]\n",
            "INFO:tensorflow:global_step/sec: 3.72849\n",
            "Total loss is :[0.297389388]\n",
            "INFO:tensorflow:global_step/sec: 3.65581\n",
            "Total loss is :[0.288198382]\n",
            "INFO:tensorflow:global_step/sec: 3.71793\n",
            "Total loss is :[0.398453444]\n",
            "INFO:tensorflow:global_step/sec: 3.87783\n",
            "Total loss is :[0.503316283]\n",
            "INFO:tensorflow:global_step/sec: 3.70506\n",
            "Total loss is :[0.35628587]\n",
            "INFO:tensorflow:global_step/sec: 3.76913\n",
            "Total loss is :[0.36128667]\n",
            "INFO:tensorflow:global_step/sec: 3.75592\n",
            "Total loss is :[0.231737673]\n",
            "INFO:tensorflow:global_step/sec: 3.71389\n",
            "Total loss is :[0.290693402]\n",
            "INFO:tensorflow:global_step/sec: 3.67211\n",
            "Total loss is :[0.434627622]\n",
            "INFO:tensorflow:global_step/sec: 3.62652\n",
            "Total loss is :[0.37538445]\n",
            "INFO:tensorflow:global_step/sec: 3.72869\n",
            "Total loss is :[0.512936652]\n",
            "INFO:tensorflow:global_step/sec: 3.72638\n",
            "Total loss is :[0.456084341]\n",
            "INFO:tensorflow:global_step/sec: 3.60858\n",
            "Total loss is :[0.337006211]\n",
            "INFO:tensorflow:global_step/sec: 3.49515\n",
            "Total loss is :[0.267980397]\n",
            "INFO:tensorflow:global_step/sec: 3.84949\n",
            "Total loss is :[0.269583523]\n",
            "INFO:tensorflow:global_step/sec: 3.81931\n",
            "Total loss is :[0.343002558]\n",
            "INFO:tensorflow:global_step/sec: 3.88912\n",
            "Total loss is :[0.397343189]\n",
            "INFO:tensorflow:global_step/sec: 3.76421\n",
            "Total loss is :[0.455869585]\n",
            "INFO:tensorflow:global_step/sec: 3.7438\n",
            "Total loss is :[0.422541559]\n",
            "INFO:tensorflow:global_step/sec: 3.91157\n",
            "Total loss is :[0.30931741]\n",
            "INFO:tensorflow:global_step/sec: 3.8991\n",
            "Total loss is :[0.535090268]\n",
            "INFO:tensorflow:global_step/sec: 3.76521\n",
            "Total loss is :[0.451974809]\n",
            "INFO:tensorflow:global_step/sec: 3.84898\n",
            "Total loss is :[0.264599025]\n",
            "INFO:tensorflow:global_step/sec: 3.74324\n",
            "Total loss is :[0.341718078]\n",
            "INFO:tensorflow:global_step/sec: 3.92697\n",
            "Total loss is :[0.307323158]\n",
            "INFO:tensorflow:global_step/sec: 3.78055\n",
            "Total loss is :[0.257077932]\n",
            "INFO:tensorflow:global_step/sec: 3.79953\n",
            "Total loss is :[0.307522178]\n",
            "INFO:tensorflow:global_step/sec: 3.74607\n",
            "Total loss is :[0.274429202]\n",
            "INFO:tensorflow:global_step/sec: 3.82498\n",
            "Total loss is :[0.368234158]\n",
            "INFO:tensorflow:global_step/sec: 3.92407\n",
            "Total loss is :[0.276514471]\n",
            "INFO:tensorflow:global_step/sec: 3.77532\n",
            "Total loss is :[0.340914249]\n",
            "INFO:tensorflow:global_step/sec: 3.92119\n",
            "Total loss is :[0.875716209]\n",
            "INFO:tensorflow:global_step/sec: 3.7314\n",
            "Total loss is :[0.259877414]\n",
            "INFO:tensorflow:global_step/sec: 3.81067\n",
            "Total loss is :[0.362995058]\n",
            "INFO:tensorflow:global_step/sec: 3.87743\n",
            "Total loss is :[0.311533868]\n",
            "INFO:tensorflow:global_step/sec: 3.89427\n",
            "Total loss is :[0.251000375]\n",
            "INFO:tensorflow:global_step/sec: 3.82491\n",
            "Total loss is :[0.325764537]\n",
            "INFO:tensorflow:global_step/sec: 3.80936\n",
            "Total loss is :[0.28865394]\n",
            "INFO:tensorflow:global_step/sec: 3.81188\n",
            "Total loss is :[0.361862451]\n",
            "INFO:tensorflow:global_step/sec: 3.69399\n",
            "Total loss is :[0.394281536]\n",
            "INFO:tensorflow:global_step/sec: 3.52473\n",
            "Total loss is :[0.307316661]\n",
            "INFO:tensorflow:global_step/sec: 3.60903\n",
            "Total loss is :[0.255130291]\n",
            "INFO:tensorflow:global_step/sec: 3.6838\n",
            "Total loss is :[0.325065553]\n",
            "INFO:tensorflow:global_step/sec: 3.82173\n",
            "Total loss is :[0.227417707]\n",
            "INFO:tensorflow:global_step/sec: 3.70651\n",
            "Total loss is :[0.276993692]\n",
            "INFO:tensorflow:global_step/sec: 3.73342\n",
            "Total loss is :[0.32400623]\n",
            "INFO:tensorflow:global_step/sec: 3.6508\n",
            "Total loss is :[0.28749153]\n",
            "INFO:tensorflow:global_step/sec: 3.63897\n",
            "Total loss is :[0.335588366]\n",
            "INFO:tensorflow:global_step/sec: 3.86039\n",
            "Total loss is :[0.274865985]\n",
            "INFO:tensorflow:global_step/sec: 3.94141\n",
            "Total loss is :[0.286020547]\n",
            "INFO:tensorflow:global_step/sec: 3.79176\n",
            "Total loss is :[0.200061023]\n",
            "INFO:tensorflow:global_step/sec: 3.66654\n",
            "Total loss is :[0.221048743]\n",
            "INFO:tensorflow:global_step/sec: 3.73928\n",
            "Total loss is :[0.28447789]\n",
            "INFO:tensorflow:global_step/sec: 3.88187\n",
            "Total loss is :[0.244613126]\n",
            "INFO:tensorflow:global_step/sec: 3.58833\n",
            "Total loss is :[0.256079227]\n",
            "INFO:tensorflow:global_step/sec: 4.00536\n",
            "Total loss is :[0.304711044]\n",
            "INFO:tensorflow:global_step/sec: 3.72918\n",
            "Total loss is :[0.464289784]\n",
            "INFO:tensorflow:global_step/sec: 3.70921\n",
            "Total loss is :[0.251888275]\n",
            "INFO:tensorflow:global_step/sec: 3.78406\n",
            "Total loss is :[0.265569448]\n",
            "INFO:tensorflow:global_step/sec: 3.74474\n",
            "Total loss is :[0.397205204]\n",
            "INFO:tensorflow:global_step/sec: 4.13369\n",
            "Total loss is :[0.319041729]\n",
            "INFO:tensorflow:global_step/sec: 3.94051\n",
            "Total loss is :[0.306850433]\n",
            "INFO:tensorflow:global_step/sec: 3.69742\n",
            "Total loss is :[0.832992852]\n",
            "INFO:tensorflow:global_step/sec: 3.95876\n",
            "Total loss is :[0.451162547]\n",
            "INFO:tensorflow:global_step/sec: 3.85803\n",
            "Total loss is :[0.423820198]\n",
            "INFO:tensorflow:global_step/sec: 3.77849\n",
            "Total loss is :[0.365349859]\n",
            "INFO:tensorflow:global_step/sec: 3.76683\n",
            "Total loss is :[0.294503778]\n",
            "INFO:tensorflow:global_step/sec: 3.87104\n",
            "Total loss is :[0.225562468]\n",
            "INFO:tensorflow:global_step/sec: 3.69223\n",
            "Total loss is :[0.271604478]\n",
            "INFO:tensorflow:global_step/sec: 3.84023\n",
            "Total loss is :[0.240021482]\n",
            "INFO:tensorflow:global_step/sec: 3.87769\n",
            "Total loss is :[0.311703086]\n",
            "INFO:tensorflow:global_step/sec: 3.76277\n",
            "Total loss is :[0.238661453]\n",
            "INFO:tensorflow:global_step/sec: 3.61452\n",
            "Total loss is :[0.29165411]\n",
            "INFO:tensorflow:global_step/sec: 3.52267\n",
            "Total loss is :[0.299214065]\n",
            "INFO:tensorflow:global_step/sec: 3.90514\n",
            "Total loss is :[0.327298164]\n",
            "INFO:tensorflow:global_step/sec: 3.89012\n",
            "Total loss is :[0.299903303]\n",
            "INFO:tensorflow:global_step/sec: 3.49511\n",
            "Total loss is :[0.282398045]\n",
            "INFO:tensorflow:global_step/sec: 3.83558\n",
            "Total loss is :[0.256102294]\n",
            "INFO:tensorflow:global_step/sec: 3.78723\n",
            "Total loss is :[0.28312555]\n",
            "INFO:tensorflow:global_step/sec: 3.89766\n",
            "Total loss is :[0.331434399]\n",
            "INFO:tensorflow:global_step/sec: 3.5967\n",
            "Total loss is :[0.308350325]\n",
            "INFO:tensorflow:global_step/sec: 3.73416\n",
            "Total loss is :[0.389599502]\n",
            "INFO:tensorflow:global_step/sec: 3.77027\n",
            "Total loss is :[0.307557702]\n",
            "INFO:tensorflow:global_step/sec: 3.88399\n",
            "Total loss is :[0.26380837]\n",
            "INFO:tensorflow:global_step/sec: 3.65971\n",
            "Total loss is :[0.224986464]\n",
            "INFO:tensorflow:global_step/sec: 3.74458\n",
            "Total loss is :[0.225264221]\n",
            "INFO:tensorflow:global_step/sec: 3.82461\n",
            "Total loss is :[0.473010272]\n",
            "INFO:tensorflow:global_step/sec: 3.85614\n",
            "Total loss is :[0.400509089]\n",
            "INFO:tensorflow:global_step/sec: 3.70326\n",
            "Total loss is :[0.278773099]\n",
            "INFO:tensorflow:global_step/sec: 3.79229\n",
            "Total loss is :[0.316504836]\n",
            "INFO:tensorflow:global_step/sec: 3.62775\n",
            "Total loss is :[0.364434302]\n",
            "INFO:tensorflow:global_step/sec: 3.7073\n",
            "Total loss is :[0.269960701]\n",
            "INFO:tensorflow:global_step/sec: 3.69438\n",
            "Total loss is :[0.267631739]\n",
            "INFO:tensorflow:global_step/sec: 3.75983\n",
            "Total loss is :[0.205174893]\n",
            "INFO:tensorflow:global_step/sec: 3.83021\n",
            "Total loss is :[0.507591546]\n",
            "INFO:tensorflow:global_step/sec: 3.58146\n",
            "Total loss is :[0.244864717]\n",
            "INFO:tensorflow:global_step/sec: 3.70775\n",
            "Total loss is :[0.247357607]\n",
            "INFO:tensorflow:global_step/sec: 3.66222\n",
            "Total loss is :[0.312319368]\n",
            "INFO:tensorflow:global_step/sec: 3.71735\n",
            "Total loss is :[0.360207438]\n",
            "INFO:tensorflow:global_step/sec: 3.79372\n",
            "Total loss is :[0.539743185]\n",
            "INFO:tensorflow:global_step/sec: 3.8731\n",
            "Total loss is :[0.234354392]\n",
            "INFO:tensorflow:global_step/sec: 3.75538\n",
            "Total loss is :[0.32679224]\n",
            "INFO:tensorflow:global_step/sec: 3.82386\n",
            "Total loss is :[0.339679718]\n",
            "INFO:tensorflow:global_step/sec: 3.66008\n",
            "Total loss is :[0.21763663]\n",
            "INFO:tensorflow:global_step/sec: 3.29732\n",
            "Total loss is :[0.33584857]\n",
            "INFO:tensorflow:global_step/sec: 3.83361\n",
            "Total loss is :[0.311049312]\n",
            "INFO:tensorflow:global_step/sec: 3.89348\n",
            "Total loss is :[0.368483156]\n",
            "INFO:tensorflow:global_step/sec: 3.73888\n",
            "Total loss is :[0.371518135]\n",
            "INFO:tensorflow:global_step/sec: 3.82672\n",
            "Total loss is :[0.307050169]\n",
            "INFO:tensorflow:global_step/sec: 3.72528\n",
            "Total loss is :[0.331807882]\n",
            "INFO:tensorflow:global_step/sec: 3.73179\n",
            "Total loss is :[0.348937809]\n",
            "INFO:tensorflow:global_step/sec: 3.6191\n",
            "Total loss is :[0.304764122]\n",
            "INFO:tensorflow:global_step/sec: 3.70454\n",
            "Total loss is :[0.324270487]\n",
            "INFO:tensorflow:global_step/sec: 3.71691\n",
            "Total loss is :[0.282896042]\n",
            "INFO:tensorflow:global_step/sec: 3.93452\n",
            "Total loss is :[0.254865855]\n",
            "INFO:tensorflow:global_step/sec: 4.0492\n",
            "Total loss is :[0.303209364]\n",
            "INFO:tensorflow:global_step/sec: 3.87921\n",
            "Total loss is :[0.307868809]\n",
            "INFO:tensorflow:global_step/sec: 3.78676\n",
            "Total loss is :[0.333850801]\n",
            "INFO:tensorflow:global_step/sec: 3.84591\n",
            "Total loss is :[0.245765254]\n",
            "INFO:tensorflow:global_step/sec: 3.84899\n",
            "Total loss is :[0.397239119]\n",
            "INFO:tensorflow:global_step/sec: 3.97568\n",
            "Total loss is :[0.214731157]\n",
            "INFO:tensorflow:global_step/sec: 3.73582\n",
            "Total loss is :[0.274277806]\n",
            "INFO:tensorflow:global_step/sec: 3.61228\n",
            "Total loss is :[0.290357262]\n",
            "INFO:tensorflow:global_step/sec: 3.85611\n",
            "Total loss is :[0.291198909]\n",
            "INFO:tensorflow:global_step/sec: 3.83975\n",
            "Total loss is :[0.251106888]\n",
            "INFO:tensorflow:global_step/sec: 3.92018\n",
            "Total loss is :[0.277261883]\n",
            "INFO:tensorflow:global_step/sec: 3.64717\n",
            "Total loss is :[0.29055202]\n",
            "INFO:tensorflow:global_step/sec: 3.75756\n",
            "Total loss is :[0.329027712]\n",
            "INFO:tensorflow:global_step/sec: 3.82971\n",
            "Total loss is :[0.292631835]\n",
            "INFO:tensorflow:global_step/sec: 3.96808\n",
            "Total loss is :[0.350934088]\n",
            "INFO:tensorflow:global_step/sec: 3.60057\n",
            "Total loss is :[0.289459705]\n",
            "INFO:tensorflow:global_step/sec: 3.8841\n",
            "Total loss is :[0.290966034]\n",
            "INFO:tensorflow:global_step/sec: 4.02409\n",
            "Total loss is :[0.241524935]\n",
            "INFO:tensorflow:global_step/sec: 3.91707\n",
            "Total loss is :[0.268740147]\n",
            "INFO:tensorflow:global_step/sec: 3.57142\n",
            "Total loss is :[0.266757101]\n",
            "INFO:tensorflow:global_step/sec: 3.75857\n",
            "Total loss is :[0.222024977]\n",
            "INFO:tensorflow:global_step/sec: 3.58117\n",
            "Total loss is :[0.223771825]\n",
            "INFO:tensorflow:global_step/sec: 3.69859\n",
            "Total loss is :[0.272180021]\n",
            "INFO:tensorflow:global_step/sec: 3.62256\n",
            "Total loss is :[0.326810092]\n",
            "INFO:tensorflow:global_step/sec: 3.95432\n",
            "Total loss is :[0.318193525]\n",
            "INFO:tensorflow:global_step/sec: 3.79793\n",
            "Total loss is :[0.46017912]\n",
            "INFO:tensorflow:global_step/sec: 3.59328\n",
            "Total loss is :[0.317282379]\n",
            "INFO:tensorflow:global_step/sec: 3.92161\n",
            "Total loss is :[0.269512653]\n",
            "INFO:tensorflow:global_step/sec: 3.77054\n",
            "Total loss is :[0.873187721]\n",
            "INFO:tensorflow:global_step/sec: 3.74803\n",
            "Total loss is :[0.225629747]\n",
            "INFO:tensorflow:global_step/sec: 3.5754\n",
            "Total loss is :[0.250204772]\n",
            "INFO:tensorflow:global_step/sec: 3.99823\n",
            "Total loss is :[0.337206304]\n",
            "INFO:tensorflow:global_step/sec: 3.82094\n",
            "Total loss is :[0.449121356]\n",
            "INFO:tensorflow:global_step/sec: 3.8461\n",
            "Total loss is :[0.410321683]\n",
            "INFO:tensorflow:global_step/sec: 3.74243\n",
            "Total loss is :[0.209926784]\n",
            "INFO:tensorflow:global_step/sec: 3.8405\n",
            "Total loss is :[0.205177546]\n",
            "INFO:tensorflow:global_step/sec: 3.94717\n",
            "Total loss is :[0.240317762]\n",
            "INFO:tensorflow:global_step/sec: 4.0206\n",
            "Total loss is :[0.262508154]\n",
            "INFO:tensorflow:global_step/sec: 3.85\n",
            "Total loss is :[0.347034335]\n",
            "INFO:tensorflow:global_step/sec: 3.86345\n",
            "Total loss is :[0.296335876]\n",
            "INFO:tensorflow:global_step/sec: 3.53579\n",
            "Total loss is :[0.392883807]\n",
            "INFO:tensorflow:global_step/sec: 3.80674\n",
            "Total loss is :[0.244457722]\n",
            "INFO:tensorflow:global_step/sec: 3.59679\n",
            "Total loss is :[0.29454422]\n",
            "INFO:tensorflow:global_step/sec: 3.49056\n",
            "Total loss is :[0.302960873]\n",
            "INFO:tensorflow:global_step/sec: 3.50141\n",
            "Total loss is :[0.331972718]\n",
            "INFO:tensorflow:global_step/sec: 3.57851\n",
            "Total loss is :[0.306207299]\n",
            "INFO:tensorflow:global_step/sec: 3.81948\n",
            "Total loss is :[0.224736288]\n",
            "INFO:tensorflow:global_step/sec: 3.76497\n",
            "Total loss is :[0.209492862]\n",
            "INFO:tensorflow:global_step/sec: 3.60369\n",
            "Total loss is :[0.287779689]\n",
            "INFO:tensorflow:global_step/sec: 3.58791\n",
            "Total loss is :[0.269972712]\n",
            "INFO:tensorflow:global_step/sec: 3.75856\n",
            "Total loss is :[0.22036317]\n",
            "INFO:tensorflow:global_step/sec: 3.55504\n",
            "Total loss is :[0.402255833]\n",
            "INFO:tensorflow:global_step/sec: 3.71814\n",
            "Total loss is :[0.362910539]\n",
            "INFO:tensorflow:global_step/sec: 3.73432\n",
            "Total loss is :[0.31221211]\n",
            "INFO:tensorflow:global_step/sec: 3.84177\n",
            "Total loss is :[0.272267044]\n",
            "INFO:tensorflow:global_step/sec: 3.92467\n",
            "Total loss is :[0.421027213]\n",
            "INFO:tensorflow:global_step/sec: 3.6845\n",
            "Total loss is :[0.24998346]\n",
            "INFO:tensorflow:global_step/sec: 3.71926\n",
            "Total loss is :[0.251667172]\n",
            "INFO:tensorflow:global_step/sec: 3.77781\n",
            "Total loss is :[0.262133479]\n",
            "INFO:tensorflow:global_step/sec: 3.86183\n",
            "Total loss is :[0.283750951]\n",
            "INFO:tensorflow:global_step/sec: 3.74255\n",
            "Total loss is :[0.269519985]\n",
            "INFO:tensorflow:global_step/sec: 3.69165\n",
            "Total loss is :[0.236109659]\n",
            "INFO:tensorflow:global_step/sec: 3.68758\n",
            "Total loss is :[0.228036761]\n",
            "INFO:tensorflow:global_step/sec: 3.80919\n",
            "Total loss is :[0.207501024]\n",
            "INFO:tensorflow:global_step/sec: 3.80142\n",
            "Total loss is :[0.383904219]\n",
            "INFO:tensorflow:global_step/sec: 3.82748\n",
            "Total loss is :[0.314211398]\n",
            "INFO:tensorflow:global_step/sec: 3.73322\n",
            "Total loss is :[0.251594275]\n",
            "INFO:tensorflow:global_step/sec: 3.47954\n",
            "Total loss is :[0.293248177]\n",
            "INFO:tensorflow:global_step/sec: 3.70366\n",
            "Total loss is :[0.332097948]\n",
            "INFO:tensorflow:global_step/sec: 3.67794\n",
            "Total loss is :[0.247543916]\n",
            "INFO:tensorflow:global_step/sec: 3.7801\n",
            "Total loss is :[0.306791961]\n",
            "INFO:tensorflow:global_step/sec: 3.71157\n",
            "Total loss is :[0.383974522]\n",
            "INFO:tensorflow:global_step/sec: 3.83132\n",
            "Total loss is :[0.195359588]\n",
            "INFO:tensorflow:global_step/sec: 3.80645\n",
            "Total loss is :[0.208187014]\n",
            "INFO:tensorflow:global_step/sec: 3.74016\n",
            "Total loss is :[0.274773717]\n",
            "INFO:tensorflow:global_step/sec: 3.66157\n",
            "Total loss is :[0.325956821]\n",
            "INFO:tensorflow:global_step/sec: 3.91883\n",
            "Total loss is :[0.227313966]\n",
            "INFO:tensorflow:global_step/sec: 3.69872\n",
            "Total loss is :[0.2156232]\n",
            "INFO:tensorflow:global_step/sec: 3.61774\n",
            "Total loss is :[0.282607079]\n",
            "INFO:tensorflow:global_step/sec: 3.5932\n",
            "Total loss is :[0.241588771]\n",
            "INFO:tensorflow:global_step/sec: 3.60821\n",
            "Total loss is :[0.192615554]\n",
            "INFO:tensorflow:global_step/sec: 3.78205\n",
            "Total loss is :[0.235470414]\n",
            "INFO:tensorflow:global_step/sec: 3.85079\n",
            "Total loss is :[0.262783557]\n",
            "INFO:tensorflow:global_step/sec: 3.81597\n",
            "Total loss is :[0.266338199]\n",
            "INFO:tensorflow:global_step/sec: 3.67972\n",
            "Total loss is :[0.252396911]\n",
            "INFO:tensorflow:global_step/sec: 3.66136\n",
            "Total loss is :[0.273067832]\n",
            "INFO:tensorflow:global_step/sec: 3.8055\n",
            "Total loss is :[0.222864]\n",
            "INFO:tensorflow:global_step/sec: 3.68278\n",
            "Total loss is :[0.284360856]\n",
            "INFO:tensorflow:global_step/sec: 3.7523\n",
            "Total loss is :[0.545472145]\n",
            "INFO:tensorflow:global_step/sec: 3.7237\n",
            "Total loss is :[0.243591189]\n",
            "INFO:tensorflow:global_step/sec: 3.78839\n",
            "Total loss is :[0.240023553]\n",
            "INFO:tensorflow:global_step/sec: 3.82134\n",
            "Total loss is :[0.291723579]\n",
            "INFO:tensorflow:global_step/sec: 3.77007\n",
            "Total loss is :[0.237334013]\n",
            "INFO:tensorflow:global_step/sec: 3.73792\n",
            "Total loss is :[0.356902629]\n",
            "INFO:tensorflow:global_step/sec: 3.91451\n",
            "Total loss is :[0.329251111]\n",
            "INFO:tensorflow:global_step/sec: 3.83556\n",
            "Total loss is :[0.305225343]\n",
            "INFO:tensorflow:global_step/sec: 3.86403\n",
            "Total loss is :[0.448041439]\n",
            "INFO:tensorflow:global_step/sec: 3.72068\n",
            "Total loss is :[0.35213]\n",
            "INFO:tensorflow:global_step/sec: 3.69175\n",
            "Total loss is :[0.223009571]\n",
            "INFO:tensorflow:global_step/sec: 3.84463\n",
            "Total loss is :[0.30075866]\n",
            "INFO:tensorflow:global_step/sec: 3.98756\n",
            "Total loss is :[0.240053892]\n",
            "INFO:tensorflow:global_step/sec: 3.86385\n",
            "Total loss is :[0.21874471]\n",
            "INFO:tensorflow:global_step/sec: 3.7713\n",
            "Total loss is :[0.284411252]\n",
            "INFO:tensorflow:global_step/sec: 3.80653\n",
            "Total loss is :[0.317497373]\n",
            "INFO:tensorflow:global_step/sec: 3.8569\n",
            "Total loss is :[0.258364886]\n",
            "INFO:tensorflow:global_step/sec: 3.88812\n",
            "Total loss is :[0.422503293]\n",
            "INFO:tensorflow:global_step/sec: 3.81334\n",
            "Total loss is :[0.431566656]\n",
            "INFO:tensorflow:global_step/sec: 3.90385\n",
            "Total loss is :[0.265091032]\n",
            "INFO:tensorflow:global_step/sec: 3.91987\n",
            "Total loss is :[0.220458359]\n",
            "INFO:tensorflow:global_step/sec: 3.74726\n",
            "Total loss is :[0.235482186]\n",
            "INFO:tensorflow:global_step/sec: 3.63222\n",
            "Total loss is :[0.213972509]\n",
            "INFO:tensorflow:global_step/sec: 3.73207\n",
            "Total loss is :[0.236686215]\n",
            "INFO:tensorflow:global_step/sec: 3.60151\n",
            "Total loss is :[0.243186504]\n",
            "INFO:tensorflow:global_step/sec: 3.81801\n",
            "Total loss is :[0.45160985]\n",
            "INFO:tensorflow:global_step/sec: 3.95812\n",
            "Total loss is :[0.205198646]\n",
            "INFO:tensorflow:global_step/sec: 3.8519\n",
            "Total loss is :[0.20876123]\n",
            "INFO:tensorflow:global_step/sec: 3.87987\n",
            "Total loss is :[0.208370581]\n",
            "INFO:tensorflow:global_step/sec: 3.82376\n",
            "Total loss is :[0.255686283]\n",
            "INFO:tensorflow:global_step/sec: 3.66901\n",
            "Total loss is :[0.179536432]\n",
            "INFO:tensorflow:global_step/sec: 3.77108\n",
            "Total loss is :[0.310663313]\n",
            "INFO:tensorflow:global_step/sec: 3.81138\n",
            "Total loss is :[0.234986976]\n",
            "INFO:tensorflow:global_step/sec: 3.66599\n",
            "Total loss is :[0.248409867]\n",
            "INFO:tensorflow:global_step/sec: 3.75107\n",
            "Total loss is :[0.291893363]\n",
            "INFO:tensorflow:global_step/sec: 3.80286\n",
            "Total loss is :[0.349560618]\n",
            "INFO:tensorflow:global_step/sec: 3.99824\n",
            "Total loss is :[0.240002826]\n",
            "INFO:tensorflow:global_step/sec: 3.65149\n",
            "Total loss is :[0.320989]\n",
            "INFO:tensorflow:global_step/sec: 3.82188\n",
            "Total loss is :[0.292569]\n",
            "INFO:tensorflow:global_step/sec: 3.62667\n",
            "Total loss is :[0.248811528]\n",
            "INFO:tensorflow:global_step/sec: 3.97324\n",
            "Total loss is :[0.249198586]\n",
            "INFO:tensorflow:global_step/sec: 3.91725\n",
            "Total loss is :[0.216739804]\n",
            "INFO:tensorflow:global_step/sec: 3.79477\n",
            "Total loss is :[1.14900148]\n",
            "INFO:tensorflow:global_step/sec: 3.93786\n",
            "Total loss is :[0.241000623]\n",
            "INFO:tensorflow:global_step/sec: 3.71611\n",
            "Total loss is :[0.233922333]\n",
            "INFO:tensorflow:global_step/sec: 3.68084\n",
            "Total loss is :[0.229016811]\n",
            "INFO:tensorflow:global_step/sec: 3.8366\n",
            "Total loss is :[0.228871524]\n",
            "INFO:tensorflow:global_step/sec: 3.80765\n",
            "Total loss is :[0.242263392]\n",
            "INFO:tensorflow:global_step/sec: 3.97945\n",
            "Total loss is :[0.36760518]\n",
            "INFO:tensorflow:global_step/sec: 3.80256\n",
            "Total loss is :[0.201327443]\n",
            "INFO:tensorflow:global_step/sec: 3.81232\n",
            "Total loss is :[0.24237223]\n",
            "INFO:tensorflow:global_step/sec: 3.75742\n",
            "Total loss is :[0.203245685]\n",
            "INFO:tensorflow:global_step/sec: 3.64615\n",
            "Total loss is :[0.219281852]\n",
            "INFO:tensorflow:global_step/sec: 3.94\n",
            "Total loss is :[0.211566985]\n",
            "INFO:tensorflow:global_step/sec: 3.60924\n",
            "Total loss is :[0.26178056]\n",
            "INFO:tensorflow:global_step/sec: 3.74853\n",
            "Total loss is :[0.1890122]\n",
            "INFO:tensorflow:global_step/sec: 3.69258\n",
            "Total loss is :[0.331302404]\n",
            "INFO:tensorflow:global_step/sec: 3.71035\n",
            "Total loss is :[0.379769325]\n",
            "INFO:tensorflow:global_step/sec: 3.68564\n",
            "Total loss is :[0.184906989]\n",
            "INFO:tensorflow:global_step/sec: 3.80873\n",
            "Total loss is :[0.228091925]\n",
            "INFO:tensorflow:global_step/sec: 3.7898\n",
            "Total loss is :[0.260926217]\n",
            "INFO:tensorflow:global_step/sec: 3.83873\n",
            "Total loss is :[0.259531707]\n",
            "INFO:tensorflow:global_step/sec: 3.9257\n",
            "Total loss is :[0.288283616]\n",
            "INFO:tensorflow:global_step/sec: 3.79868\n",
            "Total loss is :[0.296962082]\n",
            "INFO:tensorflow:global_step/sec: 3.67963\n",
            "Total loss is :[0.324320823]\n",
            "INFO:tensorflow:global_step/sec: 3.62368\n",
            "Total loss is :[0.295468569]\n",
            "INFO:tensorflow:global_step/sec: 3.89034\n",
            "Total loss is :[0.260748267]\n",
            "INFO:tensorflow:global_step/sec: 3.56048\n",
            "Total loss is :[0.200174958]\n",
            "INFO:tensorflow:global_step/sec: 3.71859\n",
            "Total loss is :[0.303023577]\n",
            "INFO:tensorflow:global_step/sec: 3.69706\n",
            "Total loss is :[0.204497933]\n",
            "INFO:tensorflow:global_step/sec: 3.8427\n",
            "Total loss is :[0.217957914]\n",
            "INFO:tensorflow:global_step/sec: 3.61085\n",
            "Total loss is :[0.197224706]\n",
            "INFO:tensorflow:global_step/sec: 3.67565\n",
            "Total loss is :[0.388614357]\n",
            "INFO:tensorflow:global_step/sec: 3.67953\n",
            "Total loss is :[0.25355953]\n",
            "INFO:tensorflow:global_step/sec: 3.73736\n",
            "Total loss is :[0.216833472]\n",
            "INFO:tensorflow:global_step/sec: 3.91532\n",
            "Total loss is :[0.236538634]\n",
            "INFO:tensorflow:global_step/sec: 3.7719\n",
            "Total loss is :[0.281086832]\n",
            "INFO:tensorflow:global_step/sec: 3.58822\n",
            "Total loss is :[0.463810116]\n",
            "INFO:tensorflow:global_step/sec: 3.7798\n",
            "Total loss is :[0.358375818]\n",
            "INFO:tensorflow:global_step/sec: 3.94763\n",
            "Total loss is :[0.231389508]\n",
            "INFO:tensorflow:global_step/sec: 3.82963\n",
            "Total loss is :[0.208065495]\n",
            "INFO:tensorflow:global_step/sec: 3.71241\n",
            "Total loss is :[0.251094103]\n",
            "INFO:tensorflow:global_step/sec: 3.71112\n",
            "Total loss is :[0.260713071]\n",
            "INFO:tensorflow:global_step/sec: 3.5246\n",
            "Total loss is :[0.195624039]\n",
            "INFO:tensorflow:global_step/sec: 3.48444\n",
            "Total loss is :[0.249037713]\n",
            "INFO:tensorflow:global_step/sec: 3.68036\n",
            "Total loss is :[0.189153284]\n",
            "INFO:tensorflow:global_step/sec: 3.59013\n",
            "Total loss is :[0.391105831]\n",
            "INFO:tensorflow:global_step/sec: 3.83422\n",
            "Total loss is :[0.210893065]\n",
            "INFO:tensorflow:global_step/sec: 3.89115\n",
            "Total loss is :[0.247869134]\n",
            "INFO:tensorflow:global_step/sec: 3.67495\n",
            "Total loss is :[0.216998905]\n",
            "INFO:tensorflow:global_step/sec: 3.6963\n",
            "Total loss is :[0.254195869]\n",
            "INFO:tensorflow:global_step/sec: 3.93508\n",
            "Total loss is :[0.219285145]\n",
            "INFO:tensorflow:global_step/sec: 3.77768\n",
            "Total loss is :[0.291158766]\n",
            "INFO:tensorflow:global_step/sec: 3.99506\n",
            "Total loss is :[0.253851742]\n",
            "INFO:tensorflow:global_step/sec: 3.58833\n",
            "Total loss is :[0.224239975]\n",
            "INFO:tensorflow:global_step/sec: 3.8396\n",
            "Total loss is :[0.216768727]\n",
            "INFO:tensorflow:global_step/sec: 3.73646\n",
            "Total loss is :[0.443103492]\n",
            "INFO:tensorflow:global_step/sec: 3.83712\n",
            "Total loss is :[0.334275365]\n",
            "INFO:tensorflow:global_step/sec: 3.94387\n",
            "Total loss is :[0.231175691]\n",
            "INFO:tensorflow:global_step/sec: 3.79178\n",
            "Total loss is :[0.240650192]\n",
            "INFO:tensorflow:global_step/sec: 3.61801\n",
            "Total loss is :[0.359250933]\n",
            "INFO:tensorflow:global_step/sec: 3.89324\n",
            "Total loss is :[0.237874612]\n",
            "INFO:tensorflow:global_step/sec: 3.62255\n",
            "Total loss is :[0.207668513]\n",
            "INFO:tensorflow:global_step/sec: 3.795\n",
            "Total loss is :[0.434180379]\n",
            "INFO:tensorflow:global_step/sec: 3.58196\n",
            "Total loss is :[0.250956565]\n",
            "INFO:tensorflow:global_step/sec: 3.76858\n",
            "Total loss is :[0.24276498]\n",
            "INFO:tensorflow:global_step/sec: 3.72372\n",
            "Total loss is :[0.224383608]\n",
            "INFO:tensorflow:global_step/sec: 3.88644\n",
            "Total loss is :[0.33861661]\n",
            "INFO:tensorflow:global_step/sec: 3.7908\n",
            "Total loss is :[0.266594231]\n",
            "INFO:tensorflow:global_step/sec: 3.70932\n",
            "Total loss is :[0.237463593]\n",
            "INFO:tensorflow:global_step/sec: 3.672\n",
            "Total loss is :[0.413841248]\n",
            "INFO:tensorflow:global_step/sec: 3.99392\n",
            "Total loss is :[0.313956231]\n",
            "INFO:tensorflow:global_step/sec: 3.70128\n",
            "Total loss is :[0.260611475]\n",
            "INFO:tensorflow:global_step/sec: 3.65726\n",
            "Total loss is :[0.238990918]\n",
            "INFO:tensorflow:global_step/sec: 3.68254\n",
            "Total loss is :[0.26610437]\n",
            "INFO:tensorflow:global_step/sec: 3.49124\n",
            "Total loss is :[0.224047035]\n",
            "INFO:tensorflow:global_step/sec: 3.8601\n",
            "Total loss is :[0.260935932]\n",
            "INFO:tensorflow:global_step/sec: 3.88029\n",
            "Total loss is :[0.387250841]\n",
            "INFO:tensorflow:global_step/sec: 3.97088\n",
            "Total loss is :[0.266283512]\n",
            "INFO:tensorflow:global_step/sec: 3.67243\n",
            "Total loss is :[0.236444652]\n",
            "INFO:tensorflow:global_step/sec: 3.707\n",
            "Total loss is :[0.23085703]\n",
            "INFO:tensorflow:global_step/sec: 3.8889\n",
            "Total loss is :[0.459299862]\n",
            "INFO:tensorflow:global_step/sec: 3.8744\n",
            "Total loss is :[0.272475839]\n",
            "INFO:tensorflow:global_step/sec: 3.68664\n",
            "Total loss is :[0.49491632]\n",
            "INFO:tensorflow:global_step/sec: 3.7384\n",
            "Total loss is :[0.266659945]\n",
            "INFO:tensorflow:global_step/sec: 3.72638\n",
            "Total loss is :[0.202252775]\n",
            "INFO:tensorflow:global_step/sec: 3.74114\n",
            "INFO:tensorflow:Saving checkpoints for 4478 into /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt.\n",
            "Total loss is :[0.228982866]\n",
            "INFO:tensorflow:global_step/sec: 2.95997\n",
            "Total loss is :[0.191559792]\n",
            "INFO:tensorflow:global_step/sec: 3.73513\n",
            "Total loss is :[0.262892812]\n",
            "INFO:tensorflow:global_step/sec: 3.83473\n",
            "Total loss is :[0.191752762]\n",
            "INFO:tensorflow:global_step/sec: 3.94179\n",
            "Total loss is :[0.229125977]\n",
            "INFO:tensorflow:global_step/sec: 3.78597\n",
            "Total loss is :[0.199561074]\n",
            "INFO:tensorflow:global_step/sec: 3.80324\n",
            "Total loss is :[0.226033]\n",
            "INFO:tensorflow:global_step/sec: 3.97278\n",
            "Total loss is :[0.219007194]\n",
            "INFO:tensorflow:global_step/sec: 3.87546\n",
            "Total loss is :[0.223533034]\n",
            "INFO:tensorflow:global_step/sec: 3.9669\n",
            "Total loss is :[0.20202069]\n",
            "INFO:tensorflow:global_step/sec: 3.61336\n",
            "Total loss is :[0.302160025]\n",
            "INFO:tensorflow:global_step/sec: 3.64589\n",
            "Total loss is :[0.214637905]\n",
            "INFO:tensorflow:global_step/sec: 3.73717\n",
            "Total loss is :[0.227779821]\n",
            "INFO:tensorflow:global_step/sec: 3.78535\n",
            "Total loss is :[0.199993491]\n",
            "INFO:tensorflow:global_step/sec: 3.69233\n",
            "Total loss is :[0.233642876]\n",
            "INFO:tensorflow:global_step/sec: 3.77445\n",
            "Total loss is :[0.296111256]\n",
            "INFO:tensorflow:global_step/sec: 3.79921\n",
            "Total loss is :[0.201166153]\n",
            "INFO:tensorflow:global_step/sec: 3.7134\n",
            "Total loss is :[0.249405921]\n",
            "INFO:tensorflow:global_step/sec: 3.55232\n",
            "Total loss is :[0.212648675]\n",
            "INFO:tensorflow:global_step/sec: 3.65674\n",
            "Total loss is :[0.274458408]\n",
            "INFO:tensorflow:global_step/sec: 4.04011\n",
            "Total loss is :[0.214154363]\n",
            "INFO:tensorflow:global_step/sec: 3.70067\n",
            "Total loss is :[0.2800937]\n",
            "INFO:tensorflow:global_step/sec: 3.7306\n",
            "Total loss is :[0.221605629]\n",
            "INFO:tensorflow:global_step/sec: 3.87243\n",
            "Total loss is :[0.174158692]\n",
            "INFO:tensorflow:global_step/sec: 3.74457\n",
            "Total loss is :[0.494017243]\n",
            "INFO:tensorflow:global_step/sec: 3.82831\n",
            "Total loss is :[0.3266626]\n",
            "INFO:tensorflow:global_step/sec: 3.81365\n",
            "Total loss is :[0.326964468]\n",
            "INFO:tensorflow:global_step/sec: 3.9148\n",
            "Total loss is :[0.297746]\n",
            "INFO:tensorflow:global_step/sec: 3.80401\n",
            "Total loss is :[0.255459964]\n",
            "INFO:tensorflow:global_step/sec: 3.73924\n",
            "Total loss is :[0.227972478]\n",
            "INFO:tensorflow:global_step/sec: 3.80559\n",
            "Total loss is :[0.261693537]\n",
            "INFO:tensorflow:global_step/sec: 3.63117\n",
            "Total loss is :[0.235052675]\n",
            "INFO:tensorflow:global_step/sec: 3.7911\n",
            "Total loss is :[0.275703192]\n",
            "INFO:tensorflow:global_step/sec: 3.64746\n",
            "Total loss is :[0.191460371]\n",
            "INFO:tensorflow:global_step/sec: 3.70775\n",
            "Total loss is :[0.29029125]\n",
            "INFO:tensorflow:global_step/sec: 3.946\n",
            "Total loss is :[0.18154259]\n",
            "INFO:tensorflow:global_step/sec: 3.81939\n",
            "Total loss is :[0.321613878]\n",
            "INFO:tensorflow:global_step/sec: 3.89852\n",
            "Total loss is :[0.21423994]\n",
            "INFO:tensorflow:global_step/sec: 3.71349\n",
            "Total loss is :[0.269175529]\n",
            "INFO:tensorflow:global_step/sec: 3.66907\n",
            "Total loss is :[0.295588851]\n",
            "INFO:tensorflow:global_step/sec: 3.88328\n",
            "Total loss is :[0.236744165]\n",
            "INFO:tensorflow:global_step/sec: 3.69812\n",
            "Total loss is :[0.190592051]\n",
            "INFO:tensorflow:global_step/sec: 3.85137\n",
            "Total loss is :[0.303284973]\n",
            "INFO:tensorflow:global_step/sec: 3.81179\n",
            "Total loss is :[0.234571174]\n",
            "INFO:tensorflow:global_step/sec: 3.70381\n",
            "Total loss is :[0.256820828]\n",
            "INFO:tensorflow:global_step/sec: 3.73791\n",
            "Total loss is :[0.190686509]\n",
            "INFO:tensorflow:global_step/sec: 3.74641\n",
            "Total loss is :[0.215127647]\n",
            "INFO:tensorflow:global_step/sec: 3.37022\n",
            "Total loss is :[0.248713657]\n",
            "INFO:tensorflow:global_step/sec: 3.66363\n",
            "Total loss is :[0.296035796]\n",
            "INFO:tensorflow:global_step/sec: 3.86199\n",
            "Total loss is :[0.240421131]\n",
            "INFO:tensorflow:global_step/sec: 3.83699\n",
            "Total loss is :[0.243784338]\n",
            "INFO:tensorflow:global_step/sec: 3.67433\n",
            "Total loss is :[0.236933157]\n",
            "INFO:tensorflow:global_step/sec: 3.81893\n",
            "Total loss is :[0.199777]\n",
            "INFO:tensorflow:global_step/sec: 3.57478\n",
            "Total loss is :[0.224995494]\n",
            "INFO:tensorflow:global_step/sec: 3.62161\n",
            "Total loss is :[0.230920047]\n",
            "INFO:tensorflow:global_step/sec: 3.60292\n",
            "Total loss is :[0.245273724]\n",
            "INFO:tensorflow:global_step/sec: 3.65597\n",
            "Total loss is :[0.319368511]\n",
            "INFO:tensorflow:global_step/sec: 3.8468\n",
            "Total loss is :[0.199033663]\n",
            "INFO:tensorflow:global_step/sec: 3.8838\n",
            "Total loss is :[0.325917482]\n",
            "INFO:tensorflow:global_step/sec: 3.62868\n",
            "Total loss is :[0.295968324]\n",
            "INFO:tensorflow:global_step/sec: 3.7947\n",
            "Total loss is :[0.20858264]\n",
            "INFO:tensorflow:global_step/sec: 3.58057\n",
            "Total loss is :[0.23879467]\n",
            "INFO:tensorflow:global_step/sec: 3.67817\n",
            "Total loss is :[0.29280746]\n",
            "INFO:tensorflow:global_step/sec: 3.76068\n",
            "Total loss is :[0.224286869]\n",
            "INFO:tensorflow:global_step/sec: 3.57879\n",
            "Total loss is :[0.821423769]\n",
            "INFO:tensorflow:global_step/sec: 3.71966\n",
            "Total loss is :[0.267953247]\n",
            "INFO:tensorflow:global_step/sec: 3.67283\n",
            "Total loss is :[0.225041568]\n",
            "INFO:tensorflow:global_step/sec: 3.67244\n",
            "Total loss is :[0.181069642]\n",
            "INFO:tensorflow:global_step/sec: 4.06363\n",
            "Total loss is :[0.248838708]\n",
            "INFO:tensorflow:global_step/sec: 3.78442\n",
            "Total loss is :[0.214229897]\n",
            "INFO:tensorflow:global_step/sec: 3.7929\n",
            "Total loss is :[0.250948042]\n",
            "INFO:tensorflow:global_step/sec: 4.03939\n",
            "Total loss is :[0.180076927]\n",
            "INFO:tensorflow:global_step/sec: 3.9199\n",
            "Total loss is :[0.275901079]\n",
            "INFO:tensorflow:global_step/sec: 3.91765\n",
            "Total loss is :[0.31812]\n",
            "INFO:tensorflow:global_step/sec: 3.75974\n",
            "Total loss is :[0.236590281]\n",
            "INFO:tensorflow:global_step/sec: 3.8225\n",
            "Total loss is :[0.305401146]\n",
            "INFO:tensorflow:global_step/sec: 3.92789\n",
            "Total loss is :[0.518835843]\n",
            "INFO:tensorflow:global_step/sec: 3.65069\n",
            "Total loss is :[0.199948534]\n",
            "INFO:tensorflow:global_step/sec: 3.65463\n",
            "Total loss is :[0.204870075]\n",
            "INFO:tensorflow:global_step/sec: 3.68272\n",
            "Total loss is :[0.236149088]\n",
            "INFO:tensorflow:global_step/sec: 3.58878\n",
            "Total loss is :[0.269073606]\n",
            "INFO:tensorflow:global_step/sec: 3.92072\n",
            "Total loss is :[0.231342942]\n",
            "INFO:tensorflow:global_step/sec: 3.99251\n",
            "Total loss is :[0.246332705]\n",
            "INFO:tensorflow:global_step/sec: 3.89142\n",
            "Total loss is :[0.262742877]\n",
            "INFO:tensorflow:global_step/sec: 3.68907\n",
            "Total loss is :[0.270510107]\n",
            "INFO:tensorflow:global_step/sec: 3.67826\n",
            "Total loss is :[0.166756749]\n",
            "INFO:tensorflow:global_step/sec: 3.78151\n",
            "Total loss is :[0.221733958]\n",
            "INFO:tensorflow:global_step/sec: 3.69485\n",
            "Total loss is :[0.229960918]\n",
            "INFO:tensorflow:global_step/sec: 3.74696\n",
            "Total loss is :[0.228047907]\n",
            "INFO:tensorflow:global_step/sec: 3.70489\n",
            "Total loss is :[0.242186591]\n",
            "INFO:tensorflow:global_step/sec: 3.76985\n",
            "Total loss is :[0.187517971]\n",
            "INFO:tensorflow:global_step/sec: 3.87791\n",
            "Total loss is :[0.220814899]\n",
            "INFO:tensorflow:global_step/sec: 3.81511\n",
            "Total loss is :[0.206790254]\n",
            "INFO:tensorflow:global_step/sec: 3.56562\n",
            "Total loss is :[0.165027037]\n",
            "INFO:tensorflow:global_step/sec: 4.04137\n",
            "Total loss is :[0.273003936]\n",
            "INFO:tensorflow:global_step/sec: 3.91283\n",
            "Total loss is :[0.218096986]\n",
            "INFO:tensorflow:global_step/sec: 3.84124\n",
            "Total loss is :[0.266693741]\n",
            "INFO:tensorflow:global_step/sec: 3.98952\n",
            "Total loss is :[0.308726609]\n",
            "INFO:tensorflow:global_step/sec: 3.70519\n",
            "Total loss is :[0.232568234]\n",
            "INFO:tensorflow:global_step/sec: 3.92902\n",
            "Total loss is :[0.25555712]\n",
            "INFO:tensorflow:global_step/sec: 3.88108\n",
            "Total loss is :[0.314858049]\n",
            "INFO:tensorflow:global_step/sec: 3.62929\n",
            "Total loss is :[0.189569265]\n",
            "INFO:tensorflow:global_step/sec: 3.63957\n",
            "Total loss is :[0.262070537]\n",
            "INFO:tensorflow:global_step/sec: 3.83191\n",
            "Total loss is :[0.245748073]\n",
            "INFO:tensorflow:global_step/sec: 3.84111\n",
            "Total loss is :[0.232990876]\n",
            "INFO:tensorflow:global_step/sec: 3.8114\n",
            "Total loss is :[0.302829921]\n",
            "INFO:tensorflow:global_step/sec: 3.64282\n",
            "Total loss is :[0.218701422]\n",
            "INFO:tensorflow:global_step/sec: 3.66936\n",
            "Total loss is :[0.223903432]\n",
            "INFO:tensorflow:global_step/sec: 3.8413\n",
            "Total loss is :[0.807944298]\n",
            "INFO:tensorflow:global_step/sec: 3.83693\n",
            "Total loss is :[0.175624058]\n",
            "INFO:tensorflow:global_step/sec: 3.63173\n",
            "Total loss is :[0.2940211]\n",
            "INFO:tensorflow:global_step/sec: 3.73257\n",
            "Total loss is :[0.228727251]\n",
            "INFO:tensorflow:global_step/sec: 3.81805\n",
            "Total loss is :[0.24292478]\n",
            "INFO:tensorflow:global_step/sec: 3.90107\n",
            "Total loss is :[0.220998749]\n",
            "INFO:tensorflow:global_step/sec: 3.73915\n",
            "Total loss is :[0.250170976]\n",
            "INFO:tensorflow:global_step/sec: 3.83823\n",
            "Total loss is :[0.272921503]\n",
            "INFO:tensorflow:global_step/sec: 3.64129\n",
            "Total loss is :[0.207428664]\n",
            "INFO:tensorflow:global_step/sec: 4.01687\n",
            "Total loss is :[0.239047021]\n",
            "INFO:tensorflow:global_step/sec: 3.82282\n",
            "Total loss is :[0.399996877]\n",
            "INFO:tensorflow:global_step/sec: 3.70429\n",
            "Total loss is :[0.249389142]\n",
            "INFO:tensorflow:global_step/sec: 3.8574\n",
            "Total loss is :[0.320445359]\n",
            "INFO:tensorflow:global_step/sec: 4.07121\n",
            "Total loss is :[0.242252246]\n",
            "INFO:tensorflow:global_step/sec: 3.70029\n",
            "Total loss is :[0.224191353]\n",
            "INFO:tensorflow:global_step/sec: 3.61707\n",
            "Total loss is :[0.263446]\n",
            "INFO:tensorflow:global_step/sec: 3.79618\n",
            "Total loss is :[0.214991808]\n",
            "INFO:tensorflow:global_step/sec: 3.77122\n",
            "Total loss is :[0.233403504]\n",
            "INFO:tensorflow:global_step/sec: 3.79415\n",
            "Total loss is :[0.205477074]\n",
            "INFO:tensorflow:global_step/sec: 3.66833\n",
            "Total loss is :[0.170762211]\n",
            "INFO:tensorflow:global_step/sec: 3.83264\n",
            "Total loss is :[0.251590729]\n",
            "INFO:tensorflow:global_step/sec: 3.80705\n",
            "Total loss is :[0.270376384]\n",
            "INFO:tensorflow:global_step/sec: 3.8126\n",
            "Total loss is :[0.261257917]\n",
            "INFO:tensorflow:global_step/sec: 3.68946\n",
            "Total loss is :[0.203956187]\n",
            "INFO:tensorflow:global_step/sec: 3.73178\n",
            "Total loss is :[0.286698341]\n",
            "INFO:tensorflow:global_step/sec: 3.7785\n",
            "Total loss is :[0.187598124]\n",
            "INFO:tensorflow:global_step/sec: 3.77997\n",
            "Total loss is :[0.215122163]\n",
            "INFO:tensorflow:global_step/sec: 3.59499\n",
            "Total loss is :[0.230343103]\n",
            "INFO:tensorflow:global_step/sec: 3.66798\n",
            "Total loss is :[0.246201277]\n",
            "INFO:tensorflow:global_step/sec: 3.53565\n",
            "Total loss is :[0.221095338]\n",
            "INFO:tensorflow:global_step/sec: 3.53455\n",
            "Total loss is :[0.205745026]\n",
            "INFO:tensorflow:global_step/sec: 3.80355\n",
            "Total loss is :[0.237862989]\n",
            "INFO:tensorflow:global_step/sec: 3.74079\n",
            "Total loss is :[0.206730366]\n",
            "INFO:tensorflow:global_step/sec: 4.05226\n",
            "Total loss is :[0.224796861]\n",
            "INFO:tensorflow:global_step/sec: 3.88662\n",
            "Total loss is :[0.288425177]\n",
            "INFO:tensorflow:global_step/sec: 3.94343\n",
            "Total loss is :[0.209208146]\n",
            "INFO:tensorflow:global_step/sec: 3.67159\n",
            "Total loss is :[0.190545589]\n",
            "INFO:tensorflow:global_step/sec: 3.96398\n",
            "Total loss is :[0.277528107]\n",
            "INFO:tensorflow:global_step/sec: 3.72514\n",
            "Total loss is :[0.207926974]\n",
            "INFO:tensorflow:global_step/sec: 3.88206\n",
            "Total loss is :[0.271178544]\n",
            "INFO:tensorflow:global_step/sec: 3.64247\n",
            "Total loss is :[0.243436381]\n",
            "INFO:tensorflow:global_step/sec: 4.02461\n",
            "Total loss is :[0.226183131]\n",
            "INFO:tensorflow:global_step/sec: 3.88623\n",
            "Total loss is :[0.222424135]\n",
            "INFO:tensorflow:global_step/sec: 3.72303\n",
            "Total loss is :[0.178368539]\n",
            "INFO:tensorflow:global_step/sec: 3.79083\n",
            "Total loss is :[0.24510543]\n",
            "INFO:tensorflow:global_step/sec: 3.65556\n",
            "Total loss is :[0.295012802]\n",
            "INFO:tensorflow:global_step/sec: 3.98529\n",
            "Total loss is :[0.298944235]\n",
            "INFO:tensorflow:global_step/sec: 3.58161\n",
            "Total loss is :[0.216577008]\n",
            "INFO:tensorflow:global_step/sec: 3.78979\n",
            "Total loss is :[0.194694743]\n",
            "INFO:tensorflow:global_step/sec: 3.76418\n",
            "Total loss is :[0.189872175]\n",
            "INFO:tensorflow:global_step/sec: 3.84919\n",
            "Total loss is :[0.196982235]\n",
            "INFO:tensorflow:global_step/sec: 3.64185\n",
            "Total loss is :[0.273969293]\n",
            "INFO:tensorflow:global_step/sec: 3.7271\n",
            "Total loss is :[0.249822289]\n",
            "INFO:tensorflow:global_step/sec: 3.76427\n",
            "Total loss is :[0.231973857]\n",
            "INFO:tensorflow:global_step/sec: 3.85256\n",
            "Total loss is :[0.18049638]\n",
            "INFO:tensorflow:global_step/sec: 3.98414\n",
            "Total loss is :[0.227184296]\n",
            "INFO:tensorflow:global_step/sec: 3.90108\n",
            "Total loss is :[0.189612895]\n",
            "INFO:tensorflow:global_step/sec: 3.78883\n",
            "Total loss is :[0.365441382]\n",
            "INFO:tensorflow:global_step/sec: 3.90422\n",
            "Total loss is :[0.210215524]\n",
            "INFO:tensorflow:global_step/sec: 3.90223\n",
            "Total loss is :[0.232516646]\n",
            "INFO:tensorflow:global_step/sec: 3.56217\n",
            "Total loss is :[0.482031494]\n",
            "INFO:tensorflow:global_step/sec: 3.63882\n",
            "Total loss is :[0.329874158]\n",
            "INFO:tensorflow:global_step/sec: 3.66314\n",
            "Total loss is :[0.305098206]\n",
            "INFO:tensorflow:global_step/sec: 3.84073\n",
            "Total loss is :[0.297790498]\n",
            "INFO:tensorflow:global_step/sec: 3.78591\n",
            "Total loss is :[0.274347275]\n",
            "INFO:tensorflow:global_step/sec: 3.71893\n",
            "Total loss is :[0.284607321]\n",
            "INFO:tensorflow:global_step/sec: 3.78114\n",
            "Total loss is :[0.203363553]\n",
            "INFO:tensorflow:global_step/sec: 3.70504\n",
            "Total loss is :[0.206131473]\n",
            "INFO:tensorflow:global_step/sec: 3.8208\n",
            "Total loss is :[0.290631115]\n",
            "INFO:tensorflow:global_step/sec: 3.81264\n",
            "Total loss is :[0.281164587]\n",
            "INFO:tensorflow:global_step/sec: 3.80312\n",
            "Total loss is :[0.222872525]\n",
            "INFO:tensorflow:global_step/sec: 3.56387\n",
            "Total loss is :[0.330158]\n",
            "INFO:tensorflow:global_step/sec: 3.70366\n",
            "Total loss is :[0.289390534]\n",
            "INFO:tensorflow:global_step/sec: 3.51141\n",
            "Total loss is :[0.257415891]\n",
            "INFO:tensorflow:global_step/sec: 3.66288\n",
            "Total loss is :[0.240243822]\n",
            "INFO:tensorflow:global_step/sec: 3.87304\n",
            "Total loss is :[0.25654164]\n",
            "INFO:tensorflow:global_step/sec: 3.60953\n",
            "Total loss is :[0.197041944]\n",
            "INFO:tensorflow:global_step/sec: 3.69122\n",
            "Total loss is :[0.200262696]\n",
            "INFO:tensorflow:global_step/sec: 3.88829\n",
            "Total loss is :[0.250148952]\n",
            "INFO:tensorflow:global_step/sec: 3.72751\n",
            "Total loss is :[0.314551204]\n",
            "INFO:tensorflow:global_step/sec: 3.80423\n",
            "Total loss is :[0.207888931]\n",
            "INFO:tensorflow:global_step/sec: 3.57467\n",
            "Total loss is :[0.260112733]\n",
            "INFO:tensorflow:global_step/sec: 3.83819\n",
            "Total loss is :[0.23107864]\n",
            "INFO:tensorflow:global_step/sec: 3.67976\n",
            "Total loss is :[0.209592924]\n",
            "INFO:tensorflow:global_step/sec: 3.66703\n",
            "Total loss is :[0.234234616]\n",
            "INFO:tensorflow:global_step/sec: 3.79221\n",
            "Total loss is :[0.256034672]\n",
            "INFO:tensorflow:global_step/sec: 3.90981\n",
            "Total loss is :[0.235540658]\n",
            "INFO:tensorflow:global_step/sec: 3.92877\n",
            "Total loss is :[0.203591824]\n",
            "INFO:tensorflow:global_step/sec: 3.64852\n",
            "Total loss is :[0.230580628]\n",
            "INFO:tensorflow:global_step/sec: 3.59801\n",
            "Total loss is :[0.272409528]\n",
            "INFO:tensorflow:global_step/sec: 3.59638\n",
            "Total loss is :[0.186226875]\n",
            "INFO:tensorflow:global_step/sec: 3.7971\n",
            "Total loss is :[0.283182859]\n",
            "INFO:tensorflow:global_step/sec: 3.48794\n",
            "Total loss is :[0.191502482]\n",
            "INFO:tensorflow:global_step/sec: 3.53134\n",
            "Total loss is :[0.241033554]\n",
            "INFO:tensorflow:global_step/sec: 3.97046\n",
            "Total loss is :[0.226280525]\n",
            "INFO:tensorflow:global_step/sec: 3.71199\n",
            "Total loss is :[0.293449521]\n",
            "INFO:tensorflow:global_step/sec: 4.00489\n",
            "Total loss is :[0.282945037]\n",
            "INFO:tensorflow:global_step/sec: 3.74132\n",
            "Total loss is :[0.224626064]\n",
            "INFO:tensorflow:global_step/sec: 3.81937\n",
            "Total loss is :[0.188345194]\n",
            "INFO:tensorflow:global_step/sec: 3.89966\n",
            "Total loss is :[0.225138441]\n",
            "INFO:tensorflow:global_step/sec: 3.86172\n",
            "Total loss is :[0.191971943]\n",
            "INFO:tensorflow:global_step/sec: 3.82881\n",
            "Total loss is :[0.249507576]\n",
            "INFO:tensorflow:global_step/sec: 3.79467\n",
            "Total loss is :[0.287437141]\n",
            "INFO:tensorflow:global_step/sec: 3.60728\n",
            "Total loss is :[0.234532326]\n",
            "INFO:tensorflow:global_step/sec: 3.79105\n",
            "Total loss is :[0.283198863]\n",
            "INFO:tensorflow:global_step/sec: 3.70637\n",
            "Total loss is :[0.300738037]\n",
            "INFO:tensorflow:global_step/sec: 3.76639\n",
            "Total loss is :[0.200456053]\n",
            "INFO:tensorflow:global_step/sec: 3.71811\n",
            "Total loss is :[0.205138892]\n",
            "INFO:tensorflow:global_step/sec: 3.84356\n",
            "Total loss is :[0.272731364]\n",
            "INFO:tensorflow:global_step/sec: 3.61227\n",
            "Total loss is :[0.190879792]\n",
            "INFO:tensorflow:global_step/sec: 3.91197\n",
            "Total loss is :[0.185031474]\n",
            "INFO:tensorflow:global_step/sec: 3.85296\n",
            "Total loss is :[0.287040651]\n",
            "INFO:tensorflow:global_step/sec: 3.73994\n",
            "Total loss is :[0.255126148]\n",
            "INFO:tensorflow:global_step/sec: 3.70137\n",
            "Total loss is :[0.231798187]\n",
            "INFO:tensorflow:global_step/sec: 3.81939\n",
            "Total loss is :[0.230125189]\n",
            "INFO:tensorflow:global_step/sec: 3.63487\n",
            "Total loss is :[0.253275156]\n",
            "INFO:tensorflow:global_step/sec: 3.64512\n",
            "Total loss is :[0.208938524]\n",
            "INFO:tensorflow:global_step/sec: 3.86737\n",
            "Total loss is :[0.212685108]\n",
            "INFO:tensorflow:global_step/sec: 3.68584\n",
            "Total loss is :[0.195861101]\n",
            "INFO:tensorflow:global_step/sec: 3.82579\n",
            "Total loss is :[0.306141376]\n",
            "INFO:tensorflow:global_step/sec: 3.56836\n",
            "Total loss is :[0.197261602]\n",
            "INFO:tensorflow:global_step/sec: 3.54176\n",
            "Total loss is :[0.178729773]\n",
            "INFO:tensorflow:global_step/sec: 3.64725\n",
            "Total loss is :[0.215200365]\n",
            "INFO:tensorflow:global_step/sec: 3.85337\n",
            "Total loss is :[0.26549691]\n",
            "INFO:tensorflow:global_step/sec: 3.97325\n",
            "Total loss is :[0.212812975]\n",
            "INFO:tensorflow:global_step/sec: 3.79722\n",
            "Total loss is :[0.236602724]\n",
            "INFO:tensorflow:global_step/sec: 3.58801\n",
            "Total loss is :[0.294263393]\n",
            "INFO:tensorflow:global_step/sec: 3.93105\n",
            "Total loss is :[0.250605702]\n",
            "INFO:tensorflow:global_step/sec: 3.82363\n",
            "Total loss is :[0.292900175]\n",
            "INFO:tensorflow:global_step/sec: 3.68594\n",
            "Total loss is :[0.213082552]\n",
            "INFO:tensorflow:global_step/sec: 3.753\n",
            "Total loss is :[0.21876651]\n",
            "INFO:tensorflow:global_step/sec: 3.7665\n",
            "Total loss is :[0.193526864]\n",
            "INFO:tensorflow:global_step/sec: 3.99106\n",
            "Total loss is :[0.261018157]\n",
            "INFO:tensorflow:global_step/sec: 3.72753\n",
            "Total loss is :[0.199656963]\n",
            "INFO:tensorflow:global_step/sec: 3.76426\n",
            "Total loss is :[0.219540253]\n",
            "INFO:tensorflow:global_step/sec: 3.68374\n",
            "Total loss is :[0.21500279]\n",
            "INFO:tensorflow:global_step/sec: 3.87147\n",
            "Total loss is :[0.258109808]\n",
            "INFO:tensorflow:global_step/sec: 4.00086\n",
            "Total loss is :[0.24095571]\n",
            "INFO:tensorflow:global_step/sec: 3.90095\n",
            "Total loss is :[0.244869411]\n",
            "INFO:tensorflow:global_step/sec: 3.75685\n",
            "Total loss is :[0.270792931]\n",
            "INFO:tensorflow:global_step/sec: 3.56731\n",
            "Total loss is :[0.248009861]\n",
            "INFO:tensorflow:global_step/sec: 3.8842\n",
            "Total loss is :[0.240845412]\n",
            "INFO:tensorflow:global_step/sec: 3.605\n",
            "Total loss is :[0.279377848]\n",
            "INFO:tensorflow:global_step/sec: 3.85745\n",
            "Total loss is :[0.245981574]\n",
            "INFO:tensorflow:global_step/sec: 3.71096\n",
            "Total loss is :[0.187525764]\n",
            "INFO:tensorflow:global_step/sec: 3.78237\n",
            "Total loss is :[0.275335]\n",
            "INFO:tensorflow:global_step/sec: 3.8543\n",
            "Total loss is :[0.241086572]\n",
            "INFO:tensorflow:global_step/sec: 3.77311\n",
            "Total loss is :[0.240127712]\n",
            "INFO:tensorflow:global_step/sec: 3.74888\n",
            "Total loss is :[0.194084287]\n",
            "INFO:tensorflow:global_step/sec: 3.80653\n",
            "Total loss is :[0.193203121]\n",
            "INFO:tensorflow:global_step/sec: 3.66449\n",
            "Total loss is :[0.267905295]\n",
            "INFO:tensorflow:global_step/sec: 3.70553\n",
            "Total loss is :[0.2608459]\n",
            "INFO:tensorflow:global_step/sec: 3.39782\n",
            "Total loss is :[0.279508054]\n",
            "INFO:tensorflow:global_step/sec: 3.59547\n",
            "Total loss is :[0.172920763]\n",
            "INFO:tensorflow:global_step/sec: 3.75848\n",
            "Total loss is :[0.247147635]\n",
            "INFO:tensorflow:global_step/sec: 3.68699\n",
            "Total loss is :[0.234143198]\n",
            "INFO:tensorflow:global_step/sec: 3.67481\n",
            "Total loss is :[0.190153271]\n",
            "INFO:tensorflow:global_step/sec: 3.72929\n",
            "Total loss is :[0.157850206]\n",
            "INFO:tensorflow:global_step/sec: 3.68331\n",
            "Total loss is :[0.375023723]\n",
            "INFO:tensorflow:global_step/sec: 3.93652\n",
            "Total loss is :[0.210217893]\n",
            "INFO:tensorflow:global_step/sec: 3.77715\n",
            "Total loss is :[0.238060459]\n",
            "INFO:tensorflow:global_step/sec: 3.82605\n",
            "Total loss is :[0.202755347]\n",
            "INFO:tensorflow:global_step/sec: 3.73062\n",
            "Total loss is :[0.258885354]\n",
            "INFO:tensorflow:global_step/sec: 4.01323\n",
            "Total loss is :[0.23982127]\n",
            "INFO:tensorflow:global_step/sec: 3.78106\n",
            "Total loss is :[0.278347015]\n",
            "INFO:tensorflow:global_step/sec: 3.64057\n",
            "Total loss is :[0.242836356]\n",
            "INFO:tensorflow:global_step/sec: 3.56531\n",
            "Total loss is :[0.320413]\n",
            "INFO:tensorflow:global_step/sec: 3.62627\n",
            "Total loss is :[0.219722971]\n",
            "INFO:tensorflow:global_step/sec: 3.83143\n",
            "Total loss is :[0.238925904]\n",
            "INFO:tensorflow:global_step/sec: 3.92261\n",
            "Total loss is :[0.193053067]\n",
            "INFO:tensorflow:global_step/sec: 3.61765\n",
            "Total loss is :[0.196411461]\n",
            "INFO:tensorflow:global_step/sec: 3.82561\n",
            "Total loss is :[0.201330066]\n",
            "INFO:tensorflow:global_step/sec: 3.85818\n",
            "Total loss is :[0.278368503]\n",
            "INFO:tensorflow:global_step/sec: 3.71838\n",
            "Total loss is :[0.255072236]\n",
            "INFO:tensorflow:global_step/sec: 3.78187\n",
            "Total loss is :[0.226935863]\n",
            "INFO:tensorflow:global_step/sec: 3.79271\n",
            "Total loss is :[0.184692487]\n",
            "INFO:tensorflow:global_step/sec: 3.57896\n",
            "Total loss is :[0.208561629]\n",
            "INFO:tensorflow:global_step/sec: 3.93669\n",
            "Total loss is :[0.256130427]\n",
            "INFO:tensorflow:global_step/sec: 3.91576\n",
            "Total loss is :[0.226391375]\n",
            "INFO:tensorflow:global_step/sec: 3.95826\n",
            "Total loss is :[0.260924041]\n",
            "INFO:tensorflow:global_step/sec: 3.68784\n",
            "Total loss is :[0.193216398]\n",
            "INFO:tensorflow:global_step/sec: 3.46107\n",
            "Total loss is :[0.218429938]\n",
            "INFO:tensorflow:global_step/sec: 3.82744\n",
            "Total loss is :[0.297650814]\n",
            "INFO:tensorflow:global_step/sec: 3.63592\n",
            "Total loss is :[0.173272327]\n",
            "INFO:tensorflow:global_step/sec: 3.76258\n",
            "Total loss is :[0.320785731]\n",
            "INFO:tensorflow:global_step/sec: 3.69633\n",
            "Total loss is :[0.208029598]\n",
            "INFO:tensorflow:global_step/sec: 3.88692\n",
            "Total loss is :[0.186320871]\n",
            "INFO:tensorflow:global_step/sec: 3.71046\n",
            "Total loss is :[0.216802195]\n",
            "INFO:tensorflow:global_step/sec: 3.49339\n",
            "Total loss is :[0.332257211]\n",
            "INFO:tensorflow:global_step/sec: 3.46801\n",
            "Total loss is :[0.224880397]\n",
            "INFO:tensorflow:global_step/sec: 3.78491\n",
            "Total loss is :[0.235288501]\n",
            "INFO:tensorflow:global_step/sec: 3.81294\n",
            "Total loss is :[0.239831686]\n",
            "INFO:tensorflow:global_step/sec: 3.67445\n",
            "Total loss is :[0.242833793]\n",
            "INFO:tensorflow:global_step/sec: 3.75837\n",
            "Total loss is :[0.247785896]\n",
            "INFO:tensorflow:global_step/sec: 3.76914\n",
            "Total loss is :[0.200291976]\n",
            "INFO:tensorflow:global_step/sec: 4.05314\n",
            "Total loss is :[0.248749793]\n",
            "INFO:tensorflow:global_step/sec: 3.84393\n",
            "Total loss is :[0.190515548]\n",
            "INFO:tensorflow:global_step/sec: 3.76017\n",
            "Total loss is :[0.246142536]\n",
            "INFO:tensorflow:global_step/sec: 3.69751\n",
            "Total loss is :[0.201457143]\n",
            "INFO:tensorflow:global_step/sec: 3.6958\n",
            "Total loss is :[0.290549576]\n",
            "INFO:tensorflow:global_step/sec: 3.78017\n",
            "Total loss is :[0.24530521]\n",
            "INFO:tensorflow:global_step/sec: 3.91076\n",
            "Total loss is :[0.256325185]\n",
            "INFO:tensorflow:global_step/sec: 3.85847\n",
            "Total loss is :[0.292313248]\n",
            "INFO:tensorflow:global_step/sec: 3.70568\n",
            "Total loss is :[0.224373713]\n",
            "INFO:tensorflow:global_step/sec: 3.8188\n",
            "Total loss is :[0.265598357]\n",
            "INFO:tensorflow:global_step/sec: 3.80343\n",
            "Total loss is :[0.222743481]\n",
            "INFO:tensorflow:global_step/sec: 3.80192\n",
            "Total loss is :[0.290614039]\n",
            "INFO:tensorflow:global_step/sec: 3.79099\n",
            "Total loss is :[0.213512629]\n",
            "INFO:tensorflow:global_step/sec: 3.83846\n",
            "Total loss is :[0.233774304]\n",
            "INFO:tensorflow:global_step/sec: 3.81085\n",
            "Total loss is :[0.223184884]\n",
            "INFO:tensorflow:global_step/sec: 3.69993\n",
            "Total loss is :[0.175423652]\n",
            "INFO:tensorflow:global_step/sec: 3.60241\n",
            "Total loss is :[0.248054177]\n",
            "INFO:tensorflow:global_step/sec: 3.5625\n",
            "Total loss is :[0.24532041]\n",
            "INFO:tensorflow:global_step/sec: 4.00625\n",
            "Total loss is :[0.222442597]\n",
            "INFO:tensorflow:global_step/sec: 3.66506\n",
            "Total loss is :[0.187422588]\n",
            "INFO:tensorflow:global_step/sec: 3.76045\n",
            "Total loss is :[0.242555678]\n",
            "INFO:tensorflow:global_step/sec: 3.7748\n",
            "Total loss is :[0.250718027]\n",
            "INFO:tensorflow:global_step/sec: 3.7934\n",
            "Total loss is :[0.242212]\n",
            "INFO:tensorflow:global_step/sec: 3.8102\n",
            "Total loss is :[0.174013555]\n",
            "INFO:tensorflow:global_step/sec: 3.87493\n",
            "Total loss is :[0.186394721]\n",
            "INFO:tensorflow:global_step/sec: 3.79425\n",
            "Total loss is :[0.277037472]\n",
            "INFO:tensorflow:global_step/sec: 3.89443\n",
            "Total loss is :[0.201130599]\n",
            "INFO:tensorflow:global_step/sec: 3.84511\n",
            "Total loss is :[0.243058085]\n",
            "INFO:tensorflow:global_step/sec: 4.01493\n",
            "Total loss is :[0.196824864]\n",
            "INFO:tensorflow:global_step/sec: 3.93516\n",
            "Total loss is :[0.209705234]\n",
            "INFO:tensorflow:global_step/sec: 3.55076\n",
            "Total loss is :[0.165912539]\n",
            "INFO:tensorflow:global_step/sec: 3.8102\n",
            "Total loss is :[0.218307361]\n",
            "INFO:tensorflow:global_step/sec: 3.7553\n",
            "Total loss is :[0.18406111]\n",
            "INFO:tensorflow:global_step/sec: 3.98429\n",
            "Total loss is :[0.396238625]\n",
            "INFO:tensorflow:global_step/sec: 3.85024\n",
            "Total loss is :[0.223668128]\n",
            "INFO:tensorflow:global_step/sec: 3.72246\n",
            "Total loss is :[0.209053352]\n",
            "INFO:tensorflow:global_step/sec: 3.79128\n",
            "Total loss is :[0.206946865]\n",
            "INFO:tensorflow:global_step/sec: 3.71908\n",
            "Total loss is :[0.197239399]\n",
            "INFO:tensorflow:global_step/sec: 3.78412\n",
            "Total loss is :[0.182073861]\n",
            "INFO:tensorflow:global_step/sec: 3.86241\n",
            "Total loss is :[0.182476759]\n",
            "INFO:tensorflow:global_step/sec: 3.87234\n",
            "Total loss is :[0.197506562]\n",
            "INFO:tensorflow:global_step/sec: 3.66787\n",
            "Total loss is :[0.239920437]\n",
            "INFO:tensorflow:global_step/sec: 3.82941\n",
            "Total loss is :[0.212414667]\n",
            "INFO:tensorflow:global_step/sec: 3.69681\n",
            "Total loss is :[0.285460681]\n",
            "INFO:tensorflow:global_step/sec: 3.6838\n",
            "Total loss is :[0.174623579]\n",
            "INFO:tensorflow:global_step/sec: 4.03107\n",
            "Total loss is :[0.209160924]\n",
            "INFO:tensorflow:global_step/sec: 3.68343\n",
            "Total loss is :[0.207718253]\n",
            "INFO:tensorflow:global_step/sec: 3.44162\n",
            "Total loss is :[0.240962744]\n",
            "INFO:tensorflow:global_step/sec: 3.42197\n",
            "Total loss is :[0.214279562]\n",
            "INFO:tensorflow:global_step/sec: 3.55776\n",
            "Total loss is :[0.220714271]\n",
            "INFO:tensorflow:global_step/sec: 3.91899\n",
            "Total loss is :[0.391114295]\n",
            "INFO:tensorflow:global_step/sec: 3.69129\n",
            "Total loss is :[0.262651592]\n",
            "INFO:tensorflow:global_step/sec: 3.87951\n",
            "Total loss is :[0.26563549]\n",
            "INFO:tensorflow:global_step/sec: 3.59107\n",
            "Total loss is :[0.262253]\n",
            "INFO:tensorflow:global_step/sec: 3.74102\n",
            "Total loss is :[0.256513268]\n",
            "INFO:tensorflow:global_step/sec: 3.69447\n",
            "Total loss is :[0.248944446]\n",
            "INFO:tensorflow:global_step/sec: 3.81285\n",
            "Total loss is :[0.18668595]\n",
            "INFO:tensorflow:global_step/sec: 3.70174\n",
            "Total loss is :[0.184740692]\n",
            "INFO:tensorflow:global_step/sec: 3.75738\n",
            "Total loss is :[0.188578099]\n",
            "INFO:tensorflow:global_step/sec: 3.96366\n",
            "Total loss is :[0.412233174]\n",
            "INFO:tensorflow:global_step/sec: 3.77745\n",
            "Total loss is :[0.199058473]\n",
            "INFO:tensorflow:global_step/sec: 3.59649\n",
            "Total loss is :[0.187979668]\n",
            "INFO:tensorflow:global_step/sec: 3.67007\n",
            "Total loss is :[0.191260338]\n",
            "INFO:tensorflow:global_step/sec: 3.87218\n",
            "Total loss is :[0.183864057]\n",
            "INFO:tensorflow:global_step/sec: 4.04425\n",
            "Total loss is :[0.258253574]\n",
            "INFO:tensorflow:global_step/sec: 3.89498\n",
            "Total loss is :[0.209856585]\n",
            "INFO:tensorflow:global_step/sec: 3.88037\n",
            "Total loss is :[0.209707826]\n",
            "INFO:tensorflow:global_step/sec: 3.73423\n",
            "Total loss is :[0.370987594]\n",
            "INFO:tensorflow:global_step/sec: 3.87721\n",
            "Total loss is :[0.192281052]\n",
            "INFO:tensorflow:global_step/sec: 3.82605\n",
            "Total loss is :[0.198550045]\n",
            "INFO:tensorflow:global_step/sec: 3.91472\n",
            "Total loss is :[0.173058301]\n",
            "INFO:tensorflow:global_step/sec: 3.79072\n",
            "Total loss is :[0.236032277]\n",
            "INFO:tensorflow:global_step/sec: 3.81944\n",
            "Total loss is :[0.227266908]\n",
            "INFO:tensorflow:global_step/sec: 3.83936\n",
            "Total loss is :[0.301246524]\n",
            "INFO:tensorflow:global_step/sec: 3.85803\n",
            "Total loss is :[0.205216497]\n",
            "INFO:tensorflow:global_step/sec: 3.71565\n",
            "Total loss is :[0.213105932]\n",
            "INFO:tensorflow:global_step/sec: 3.75458\n",
            "Total loss is :[0.182639629]\n",
            "INFO:tensorflow:global_step/sec: 3.74951\n",
            "Total loss is :[0.175298989]\n",
            "INFO:tensorflow:global_step/sec: 3.60777\n",
            "Total loss is :[0.18711479]\n",
            "INFO:tensorflow:global_step/sec: 3.76609\n",
            "Total loss is :[0.201100141]\n",
            "INFO:tensorflow:global_step/sec: 3.78954\n",
            "Total loss is :[0.232059836]\n",
            "INFO:tensorflow:global_step/sec: 3.69106\n",
            "Total loss is :[0.260724425]\n",
            "INFO:tensorflow:global_step/sec: 3.75525\n",
            "Total loss is :[0.165951401]\n",
            "INFO:tensorflow:global_step/sec: 3.74404\n",
            "Total loss is :[0.230869532]\n",
            "INFO:tensorflow:global_step/sec: 3.90584\n",
            "Total loss is :[0.181339309]\n",
            "INFO:tensorflow:global_step/sec: 3.84951\n",
            "Total loss is :[0.270445287]\n",
            "INFO:tensorflow:global_step/sec: 3.96134\n",
            "Total loss is :[0.210446626]\n",
            "INFO:tensorflow:global_step/sec: 3.81211\n",
            "Total loss is :[0.287271529]\n",
            "INFO:tensorflow:global_step/sec: 3.7866\n",
            "Total loss is :[0.284804463]\n",
            "INFO:tensorflow:global_step/sec: 3.68557\n",
            "Total loss is :[0.291610181]\n",
            "INFO:tensorflow:global_step/sec: 3.78205\n",
            "Total loss is :[0.183515668]\n",
            "INFO:tensorflow:global_step/sec: 3.91609\n",
            "Total loss is :[0.223046497]\n",
            "INFO:tensorflow:global_step/sec: 3.92064\n",
            "Total loss is :[0.248297259]\n",
            "INFO:tensorflow:global_step/sec: 3.72512\n",
            "Total loss is :[0.210553542]\n",
            "INFO:tensorflow:global_step/sec: 3.65789\n",
            "Total loss is :[0.166018441]\n",
            "INFO:tensorflow:global_step/sec: 3.7404\n",
            "Total loss is :[0.211672977]\n",
            "INFO:tensorflow:global_step/sec: 3.58\n",
            "Total loss is :[0.448628217]\n",
            "INFO:tensorflow:global_step/sec: 3.70513\n",
            "Total loss is :[0.398420542]\n",
            "INFO:tensorflow:global_step/sec: 3.96695\n",
            "Total loss is :[0.287460178]\n",
            "INFO:tensorflow:global_step/sec: 3.63997\n",
            "Total loss is :[0.192329973]\n",
            "INFO:tensorflow:global_step/sec: 3.79047\n",
            "Total loss is :[0.178622127]\n",
            "INFO:tensorflow:global_step/sec: 3.81751\n",
            "Total loss is :[0.225194]\n",
            "INFO:tensorflow:global_step/sec: 3.81829\n",
            "Total loss is :[0.678238034]\n",
            "INFO:tensorflow:global_step/sec: 3.70395\n",
            "Total loss is :[0.252218068]\n",
            "INFO:tensorflow:global_step/sec: 3.70501\n",
            "Total loss is :[0.226472333]\n",
            "INFO:tensorflow:global_step/sec: 3.87877\n",
            "Total loss is :[0.242566869]\n",
            "INFO:tensorflow:global_step/sec: 3.74717\n",
            "Total loss is :[0.236726791]\n",
            "INFO:tensorflow:global_step/sec: 3.52458\n",
            "Total loss is :[0.257092267]\n",
            "INFO:tensorflow:global_step/sec: 3.65049\n",
            "Total loss is :[0.214182481]\n",
            "INFO:tensorflow:global_step/sec: 3.61255\n",
            "Total loss is :[0.18699199]\n",
            "INFO:tensorflow:global_step/sec: 3.67845\n",
            "Total loss is :[0.183304712]\n",
            "INFO:tensorflow:global_step/sec: 3.66481\n",
            "Total loss is :[0.16342397]\n",
            "INFO:tensorflow:global_step/sec: 3.81857\n",
            "Total loss is :[0.22355029]\n",
            "INFO:tensorflow:global_step/sec: 3.74524\n",
            "Total loss is :[0.192502335]\n",
            "INFO:tensorflow:global_step/sec: 3.61444\n",
            "Total loss is :[0.25115481]\n",
            "INFO:tensorflow:global_step/sec: 3.85344\n",
            "Total loss is :[0.280438066]\n",
            "INFO:tensorflow:global_step/sec: 3.84947\n",
            "Total loss is :[0.283749342]\n",
            "INFO:tensorflow:global_step/sec: 3.75671\n",
            "Total loss is :[0.263635099]\n",
            "INFO:tensorflow:global_step/sec: 4.03452\n",
            "Total loss is :[0.253892571]\n",
            "INFO:tensorflow:global_step/sec: 4.0971\n",
            "Total loss is :[0.190993309]\n",
            "INFO:tensorflow:global_step/sec: 4.08654\n",
            "Total loss is :[0.22866042]\n",
            "INFO:tensorflow:global_step/sec: 3.82975\n",
            "Total loss is :[0.207730234]\n",
            "INFO:tensorflow:global_step/sec: 3.74885\n",
            "Total loss is :[0.178079963]\n",
            "INFO:tensorflow:global_step/sec: 3.97073\n",
            "Total loss is :[0.22481662]\n",
            "INFO:tensorflow:global_step/sec: 3.75029\n",
            "Total loss is :[0.272957981]\n",
            "INFO:tensorflow:global_step/sec: 3.63703\n",
            "Total loss is :[0.308187544]\n",
            "INFO:tensorflow:global_step/sec: 3.95663\n",
            "Total loss is :[0.169837505]\n",
            "INFO:tensorflow:global_step/sec: 3.69537\n",
            "Total loss is :[0.270616144]\n",
            "INFO:tensorflow:global_step/sec: 3.74127\n",
            "Total loss is :[0.311584383]\n",
            "INFO:tensorflow:global_step/sec: 3.58745\n",
            "Total loss is :[0.224974692]\n",
            "INFO:tensorflow:global_step/sec: 3.88702\n",
            "Total loss is :[0.20278503]\n",
            "INFO:tensorflow:global_step/sec: 3.89083\n",
            "Total loss is :[0.203308463]\n",
            "INFO:tensorflow:global_step/sec: 3.87104\n",
            "Total loss is :[0.247353822]\n",
            "INFO:tensorflow:global_step/sec: 3.74098\n",
            "Total loss is :[0.210084841]\n",
            "INFO:tensorflow:global_step/sec: 3.87875\n",
            "Total loss is :[0.205284238]\n",
            "INFO:tensorflow:global_step/sec: 3.73418\n",
            "Total loss is :[0.203322411]\n",
            "INFO:tensorflow:global_step/sec: 3.81123\n",
            "Total loss is :[0.181515217]\n",
            "INFO:tensorflow:global_step/sec: 3.95521\n",
            "Total loss is :[0.255712658]\n",
            "INFO:tensorflow:global_step/sec: 3.61202\n",
            "Total loss is :[0.205224276]\n",
            "INFO:tensorflow:global_step/sec: 3.51128\n",
            "Total loss is :[0.336187512]\n",
            "INFO:tensorflow:global_step/sec: 3.47318\n",
            "Total loss is :[0.273731]\n",
            "INFO:tensorflow:global_step/sec: 3.81935\n",
            "Total loss is :[0.224632218]\n",
            "INFO:tensorflow:global_step/sec: 3.98582\n",
            "Total loss is :[0.202279091]\n",
            "INFO:tensorflow:global_step/sec: 3.83749\n",
            "Total loss is :[0.181346208]\n",
            "INFO:tensorflow:global_step/sec: 3.64488\n",
            "Total loss is :[0.168798834]\n",
            "INFO:tensorflow:global_step/sec: 3.68741\n",
            "Total loss is :[0.189414829]\n",
            "INFO:tensorflow:global_step/sec: 3.8188\n",
            "Total loss is :[0.246664777]\n",
            "INFO:tensorflow:global_step/sec: 3.72068\n",
            "Total loss is :[0.170714319]\n",
            "INFO:tensorflow:global_step/sec: 3.72781\n",
            "Total loss is :[0.236103714]\n",
            "INFO:tensorflow:global_step/sec: 3.74801\n",
            "INFO:tensorflow:Saving checkpoints for 8994 into /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt.\n",
            "Total loss is :[0.181032702]\n",
            "INFO:tensorflow:global_step/sec: 2.99669\n",
            "Total loss is :[0.274406642]\n",
            "INFO:tensorflow:global_step/sec: 3.84931\n",
            "Total loss is :[0.190603182]\n",
            "INFO:tensorflow:global_step/sec: 3.68269\n",
            "Total loss is :[0.188642472]\n",
            "INFO:tensorflow:global_step/sec: 3.68257\n",
            "Total loss is :[0.196901143]\n",
            "INFO:tensorflow:global_step/sec: 3.91098\n",
            "Total loss is :[0.192104489]\n",
            "INFO:tensorflow:global_step/sec: 3.77582\n",
            "Total loss is :[0.233118817]\n",
            "INFO:tensorflow:global_step/sec: 3.80197\n",
            "Total loss is :[0.193789229]\n",
            "INFO:tensorflow:global_step/sec: 3.85514\n",
            "Total loss is :[0.195344657]\n",
            "INFO:tensorflow:global_step/sec: 3.71211\n",
            "Total loss is :[0.166647479]\n",
            "INFO:tensorflow:global_step/sec: 3.66396\n",
            "Total loss is :[0.268518865]\n",
            "INFO:tensorflow:global_step/sec: 3.6059\n",
            "Total loss is :[0.229737043]\n",
            "INFO:tensorflow:global_step/sec: 3.75911\n",
            "Total loss is :[0.204542831]\n",
            "INFO:tensorflow:global_step/sec: 3.75816\n",
            "Total loss is :[0.245332301]\n",
            "INFO:tensorflow:global_step/sec: 3.61371\n",
            "Total loss is :[0.172274008]\n",
            "INFO:tensorflow:global_step/sec: 3.90832\n",
            "Total loss is :[0.175694078]\n",
            "INFO:tensorflow:global_step/sec: 3.81283\n",
            "Total loss is :[0.251796842]\n",
            "INFO:tensorflow:global_step/sec: 3.76585\n",
            "Total loss is :[0.212316126]\n",
            "INFO:tensorflow:global_step/sec: 3.8041\n",
            "Total loss is :[0.290455192]\n",
            "INFO:tensorflow:global_step/sec: 3.57103\n",
            "Total loss is :[0.18383491]\n",
            "INFO:tensorflow:global_step/sec: 3.76331\n",
            "Total loss is :[0.184118345]\n",
            "INFO:tensorflow:global_step/sec: 3.41444\n",
            "Total loss is :[0.221571758]\n",
            "INFO:tensorflow:global_step/sec: 4.03923\n",
            "Total loss is :[0.183619022]\n",
            "INFO:tensorflow:global_step/sec: 3.80321\n",
            "Total loss is :[0.216290429]\n",
            "INFO:tensorflow:global_step/sec: 3.78091\n",
            "Total loss is :[0.348376453]\n",
            "INFO:tensorflow:global_step/sec: 3.80193\n",
            "Total loss is :[0.226164117]\n",
            "INFO:tensorflow:global_step/sec: 3.78276\n",
            "Total loss is :[0.167598724]\n",
            "INFO:tensorflow:global_step/sec: 3.84769\n",
            "Total loss is :[0.236228153]\n",
            "INFO:tensorflow:global_step/sec: 3.89648\n",
            "Total loss is :[0.189585865]\n",
            "INFO:tensorflow:global_step/sec: 3.89709\n",
            "Total loss is :[0.217398122]\n",
            "INFO:tensorflow:global_step/sec: 3.70738\n",
            "Total loss is :[0.205643952]\n",
            "INFO:tensorflow:global_step/sec: 3.71239\n",
            "Total loss is :[0.25270021]\n",
            "INFO:tensorflow:global_step/sec: 3.84141\n",
            "Total loss is :[0.233162448]\n",
            "INFO:tensorflow:global_step/sec: 3.77361\n",
            "Total loss is :[0.22071743]\n",
            "INFO:tensorflow:global_step/sec: 3.72232\n",
            "Total loss is :[0.185951456]\n",
            "INFO:tensorflow:global_step/sec: 3.91256\n",
            "Total loss is :[0.24971956]\n",
            "INFO:tensorflow:global_step/sec: 3.77108\n",
            "Total loss is :[0.282110602]\n",
            "INFO:tensorflow:global_step/sec: 3.95075\n",
            "Total loss is :[0.213024199]\n",
            "INFO:tensorflow:global_step/sec: 3.72096\n",
            "Total loss is :[0.233417898]\n",
            "INFO:tensorflow:global_step/sec: 3.91746\n",
            "Total loss is :[0.179527655]\n",
            "INFO:tensorflow:global_step/sec: 3.68705\n",
            "Total loss is :[0.241019964]\n",
            "INFO:tensorflow:global_step/sec: 3.70485\n",
            "Total loss is :[0.248261109]\n",
            "INFO:tensorflow:global_step/sec: 3.87555\n",
            "Total loss is :[0.181706071]\n",
            "INFO:tensorflow:global_step/sec: 3.79018\n",
            "Total loss is :[0.243421733]\n",
            "INFO:tensorflow:global_step/sec: 3.7535\n",
            "Total loss is :[0.19722648]\n",
            "INFO:tensorflow:global_step/sec: 3.76745\n",
            "Total loss is :[0.185268104]\n",
            "INFO:tensorflow:global_step/sec: 3.84463\n",
            "Total loss is :[0.269488513]\n",
            "INFO:tensorflow:global_step/sec: 3.66885\n",
            "Total loss is :[0.305777192]\n",
            "INFO:tensorflow:global_step/sec: 3.78341\n",
            "Total loss is :[0.233381659]\n",
            "INFO:tensorflow:global_step/sec: 3.55918\n",
            "Total loss is :[0.274871528]\n",
            "INFO:tensorflow:global_step/sec: 3.46248\n",
            "Total loss is :[0.171384901]\n",
            "INFO:tensorflow:global_step/sec: 3.94667\n",
            "Total loss is :[0.171827018]\n",
            "INFO:tensorflow:global_step/sec: 3.88483\n",
            "Total loss is :[0.189054146]\n",
            "INFO:tensorflow:global_step/sec: 3.85174\n",
            "Total loss is :[0.291781962]\n",
            "INFO:tensorflow:global_step/sec: 3.96638\n",
            "Total loss is :[0.252566695]\n",
            "INFO:tensorflow:global_step/sec: 3.88065\n",
            "Total loss is :[0.25725016]\n",
            "INFO:tensorflow:global_step/sec: 3.76913\n",
            "Total loss is :[0.223431438]\n",
            "INFO:tensorflow:global_step/sec: 3.62581\n",
            "Total loss is :[0.240514413]\n",
            "INFO:tensorflow:global_step/sec: 3.61478\n",
            "Total loss is :[0.199159369]\n",
            "INFO:tensorflow:global_step/sec: 3.88958\n",
            "Total loss is :[0.2088269]\n",
            "INFO:tensorflow:global_step/sec: 3.8626\n",
            "Total loss is :[0.281904072]\n",
            "INFO:tensorflow:global_step/sec: 3.73699\n",
            "Total loss is :[0.272835821]\n",
            "INFO:tensorflow:global_step/sec: 3.73791\n",
            "Total loss is :[0.185171619]\n",
            "INFO:tensorflow:global_step/sec: 3.68589\n",
            "Total loss is :[0.444435567]\n",
            "INFO:tensorflow:global_step/sec: 3.66396\n",
            "Total loss is :[0.26923731]\n",
            "INFO:tensorflow:global_step/sec: 4.06781\n",
            "Total loss is :[0.187874258]\n",
            "INFO:tensorflow:global_step/sec: 3.75888\n",
            "Total loss is :[0.210600793]\n",
            "INFO:tensorflow:global_step/sec: 3.8159\n",
            "Total loss is :[0.217680305]\n",
            "INFO:tensorflow:global_step/sec: 3.91742\n",
            "Total loss is :[0.161799401]\n",
            "INFO:tensorflow:global_step/sec: 3.7847\n",
            "Total loss is :[0.26535973]\n",
            "INFO:tensorflow:global_step/sec: 3.85304\n",
            "Total loss is :[0.253733873]\n",
            "INFO:tensorflow:global_step/sec: 3.66446\n",
            "Total loss is :[0.181580663]\n",
            "INFO:tensorflow:global_step/sec: 3.77245\n",
            "Total loss is :[0.195927352]\n",
            "INFO:tensorflow:global_step/sec: 3.67933\n",
            "Total loss is :[0.178715512]\n",
            "INFO:tensorflow:global_step/sec: 3.7715\n",
            "Total loss is :[0.343711913]\n",
            "INFO:tensorflow:global_step/sec: 3.82691\n",
            "Total loss is :[0.190447316]\n",
            "INFO:tensorflow:global_step/sec: 3.44138\n",
            "Total loss is :[0.236131936]\n",
            "INFO:tensorflow:global_step/sec: 3.57802\n",
            "Total loss is :[0.197315603]\n",
            "INFO:tensorflow:global_step/sec: 3.76086\n",
            "Total loss is :[0.221470654]\n",
            "INFO:tensorflow:global_step/sec: 3.65241\n",
            "Total loss is :[0.2260589]\n",
            "INFO:tensorflow:global_step/sec: 3.54502\n",
            "Total loss is :[0.183737636]\n",
            "INFO:tensorflow:global_step/sec: 3.47809\n",
            "Total loss is :[0.23869437]\n",
            "INFO:tensorflow:global_step/sec: 3.50949\n",
            "Total loss is :[0.21427241]\n",
            "INFO:tensorflow:global_step/sec: 3.94244\n",
            "Total loss is :[0.194988802]\n",
            "INFO:tensorflow:global_step/sec: 3.83168\n",
            "Total loss is :[0.219650775]\n",
            "INFO:tensorflow:global_step/sec: 3.80477\n",
            "Total loss is :[0.441796273]\n",
            "INFO:tensorflow:global_step/sec: 3.79487\n",
            "Total loss is :[0.293641031]\n",
            "INFO:tensorflow:global_step/sec: 3.76766\n",
            "Total loss is :[0.17636165]\n",
            "INFO:tensorflow:global_step/sec: 3.87966\n",
            "Total loss is :[0.185147792]\n",
            "INFO:tensorflow:global_step/sec: 3.69514\n",
            "Total loss is :[0.225873679]\n",
            "INFO:tensorflow:global_step/sec: 3.69639\n",
            "Total loss is :[0.264679432]\n",
            "INFO:tensorflow:global_step/sec: 3.77911\n",
            "Total loss is :[0.211613894]\n",
            "INFO:tensorflow:global_step/sec: 3.81134\n",
            "Total loss is :[0.212478787]\n",
            "INFO:tensorflow:global_step/sec: 3.84917\n",
            "Total loss is :[0.230678]\n",
            "INFO:tensorflow:global_step/sec: 3.72091\n",
            "Total loss is :[0.266257]\n",
            "INFO:tensorflow:global_step/sec: 3.80195\n",
            "Total loss is :[0.18342568]\n",
            "INFO:tensorflow:global_step/sec: 3.9944\n",
            "Total loss is :[0.182030201]\n",
            "INFO:tensorflow:global_step/sec: 3.77326\n",
            "Total loss is :[0.20148845]\n",
            "INFO:tensorflow:global_step/sec: 3.82734\n",
            "Total loss is :[0.188026309]\n",
            "INFO:tensorflow:global_step/sec: 3.83728\n",
            "Total loss is :[0.265390635]\n",
            "INFO:tensorflow:global_step/sec: 3.70201\n",
            "Total loss is :[0.399777472]\n",
            "INFO:tensorflow:global_step/sec: 4.02354\n",
            "Total loss is :[0.203383833]\n",
            "INFO:tensorflow:global_step/sec: 3.74796\n",
            "Total loss is :[0.218712524]\n",
            "INFO:tensorflow:global_step/sec: 3.92645\n",
            "Total loss is :[0.194082513]\n",
            "INFO:tensorflow:global_step/sec: 3.64626\n",
            "Total loss is :[0.18582648]\n",
            "INFO:tensorflow:global_step/sec: 3.79529\n",
            "Total loss is :[0.314243346]\n",
            "INFO:tensorflow:global_step/sec: 3.81833\n",
            "Total loss is :[0.290251255]\n",
            "INFO:tensorflow:global_step/sec: 3.73115\n",
            "Total loss is :[0.224406183]\n",
            "INFO:tensorflow:global_step/sec: 3.77225\n",
            "Total loss is :[0.192618236]\n",
            "INFO:tensorflow:global_step/sec: 3.62893\n",
            "Total loss is :[0.276508]\n",
            "INFO:tensorflow:global_step/sec: 3.56585\n",
            "Total loss is :[0.220236927]\n",
            "INFO:tensorflow:global_step/sec: 3.79855\n",
            "Total loss is :[0.297499895]\n",
            "INFO:tensorflow:global_step/sec: 3.71493\n",
            "Total loss is :[0.182581216]\n",
            "INFO:tensorflow:global_step/sec: 3.65487\n",
            "Total loss is :[0.18142955]\n",
            "INFO:tensorflow:global_step/sec: 3.76809\n",
            "Total loss is :[0.244743571]\n",
            "INFO:tensorflow:global_step/sec: 3.77195\n",
            "Total loss is :[0.169624925]\n",
            "INFO:tensorflow:global_step/sec: 3.75001\n",
            "Total loss is :[0.211804956]\n",
            "INFO:tensorflow:global_step/sec: 3.8327\n",
            "Total loss is :[0.209926546]\n",
            "INFO:tensorflow:global_step/sec: 3.93971\n",
            "Total loss is :[0.195419833]\n",
            "INFO:tensorflow:global_step/sec: 4.03952\n",
            "Total loss is :[0.226577833]\n",
            "INFO:tensorflow:global_step/sec: 3.88408\n",
            "Total loss is :[0.303346068]\n",
            "INFO:tensorflow:global_step/sec: 3.77978\n",
            "Total loss is :[0.191038474]\n",
            "INFO:tensorflow:global_step/sec: 3.72286\n",
            "Total loss is :[0.164399579]\n",
            "INFO:tensorflow:global_step/sec: 3.81926\n",
            "Total loss is :[0.303054]\n",
            "INFO:tensorflow:global_step/sec: 3.88627\n",
            "Total loss is :[0.232009321]\n",
            "INFO:tensorflow:global_step/sec: 3.90581\n",
            "Total loss is :[0.196021318]\n",
            "INFO:tensorflow:global_step/sec: 3.8\n",
            "Total loss is :[0.192407]\n",
            "INFO:tensorflow:global_step/sec: 3.62601\n",
            "Total loss is :[0.24406302]\n",
            "INFO:tensorflow:global_step/sec: 3.90853\n",
            "Total loss is :[0.232213855]\n",
            "INFO:tensorflow:global_step/sec: 3.68142\n",
            "Total loss is :[0.234554991]\n",
            "INFO:tensorflow:global_step/sec: 3.77706\n",
            "Total loss is :[0.400023878]\n",
            "INFO:tensorflow:global_step/sec: 3.6442\n",
            "Total loss is :[0.265583038]\n",
            "INFO:tensorflow:global_step/sec: 3.6762\n",
            "Total loss is :[0.229249075]\n",
            "INFO:tensorflow:global_step/sec: 3.80946\n",
            "Total loss is :[0.211573929]\n",
            "INFO:tensorflow:global_step/sec: 3.8604\n",
            "Total loss is :[0.235274628]\n",
            "INFO:tensorflow:global_step/sec: 3.92083\n",
            "Total loss is :[0.22714445]\n",
            "INFO:tensorflow:global_step/sec: 3.57277\n",
            "Total loss is :[0.328397632]\n",
            "INFO:tensorflow:global_step/sec: 3.76494\n",
            "Total loss is :[0.213208959]\n",
            "INFO:tensorflow:global_step/sec: 3.79731\n",
            "Total loss is :[0.175487921]\n",
            "INFO:tensorflow:global_step/sec: 3.82845\n",
            "Total loss is :[0.244739488]\n",
            "INFO:tensorflow:global_step/sec: 3.41505\n",
            "Total loss is :[0.203049362]\n",
            "INFO:tensorflow:global_step/sec: 3.55374\n",
            "Total loss is :[0.239690632]\n",
            "INFO:tensorflow:global_step/sec: 3.7609\n",
            "Total loss is :[0.187093765]\n",
            "INFO:tensorflow:global_step/sec: 3.87006\n",
            "Total loss is :[0.228639722]\n",
            "INFO:tensorflow:global_step/sec: 3.77383\n",
            "Total loss is :[0.183390602]\n",
            "INFO:tensorflow:global_step/sec: 3.73585\n",
            "Total loss is :[0.235517383]\n",
            "INFO:tensorflow:global_step/sec: 3.85139\n",
            "Total loss is :[0.283598065]\n",
            "INFO:tensorflow:global_step/sec: 3.65547\n",
            "Total loss is :[0.25004068]\n",
            "INFO:tensorflow:global_step/sec: 3.73226\n",
            "Total loss is :[0.202148795]\n",
            "INFO:tensorflow:global_step/sec: 3.80916\n",
            "Total loss is :[0.224341393]\n",
            "INFO:tensorflow:global_step/sec: 3.81303\n",
            "Total loss is :[0.260112405]\n",
            "INFO:tensorflow:global_step/sec: 4.03146\n",
            "Total loss is :[0.187502503]\n",
            "INFO:tensorflow:global_step/sec: 3.86011\n",
            "Total loss is :[0.214396611]\n",
            "INFO:tensorflow:global_step/sec: 3.87005\n",
            "Total loss is :[0.189175636]\n",
            "INFO:tensorflow:global_step/sec: 3.85081\n",
            "Total loss is :[0.210323572]\n",
            "INFO:tensorflow:global_step/sec: 3.90663\n",
            "Total loss is :[0.251043558]\n",
            "INFO:tensorflow:global_step/sec: 3.90957\n",
            "Total loss is :[0.224240065]\n",
            "INFO:tensorflow:global_step/sec: 3.92291\n",
            "Total loss is :[0.191780895]\n",
            "INFO:tensorflow:global_step/sec: 3.84064\n",
            "Total loss is :[0.153294101]\n",
            "INFO:tensorflow:global_step/sec: 3.60351\n",
            "Total loss is :[0.253067881]\n",
            "INFO:tensorflow:global_step/sec: 3.78445\n",
            "Total loss is :[0.301220536]\n",
            "INFO:tensorflow:global_step/sec: 3.89142\n",
            "Total loss is :[0.189401597]\n",
            "INFO:tensorflow:global_step/sec: 3.69836\n",
            "Total loss is :[0.281976134]\n",
            "INFO:tensorflow:global_step/sec: 3.7687\n",
            "Total loss is :[0.284164667]\n",
            "INFO:tensorflow:global_step/sec: 4.04399\n",
            "Total loss is :[0.30336991]\n",
            "INFO:tensorflow:global_step/sec: 3.79092\n",
            "Total loss is :[0.205085739]\n",
            "INFO:tensorflow:global_step/sec: 3.85048\n",
            "Total loss is :[0.225930214]\n",
            "INFO:tensorflow:global_step/sec: 3.73971\n",
            "Total loss is :[0.288160831]\n",
            "INFO:tensorflow:global_step/sec: 3.68979\n",
            "Total loss is :[0.516890764]\n",
            "INFO:tensorflow:global_step/sec: 3.86611\n",
            "Total loss is :[0.191113651]\n",
            "INFO:tensorflow:global_step/sec: 3.42572\n",
            "Total loss is :[0.401043]\n",
            "INFO:tensorflow:global_step/sec: 3.45618\n",
            "Total loss is :[0.276690722]\n",
            "INFO:tensorflow:global_step/sec: 3.59626\n",
            "Total loss is :[0.216108531]\n",
            "INFO:tensorflow:global_step/sec: 3.7976\n",
            "Total loss is :[0.15938887]\n",
            "INFO:tensorflow:global_step/sec: 3.82873\n",
            "Total loss is :[0.176120147]\n",
            "INFO:tensorflow:global_step/sec: 4.01367\n",
            "Total loss is :[0.213753358]\n",
            "INFO:tensorflow:global_step/sec: 3.76803\n",
            "Total loss is :[0.243134081]\n",
            "INFO:tensorflow:global_step/sec: 3.5506\n",
            "Total loss is :[0.239200324]\n",
            "INFO:tensorflow:global_step/sec: 3.54257\n",
            "Total loss is :[0.173226327]\n",
            "INFO:tensorflow:global_step/sec: 3.73994\n",
            "Total loss is :[0.190887228]\n",
            "INFO:tensorflow:global_step/sec: 3.65318\n",
            "Total loss is :[0.206387162]\n",
            "INFO:tensorflow:global_step/sec: 3.68474\n",
            "Total loss is :[0.189415723]\n",
            "INFO:tensorflow:global_step/sec: 3.7586\n",
            "Total loss is :[0.201882079]\n",
            "INFO:tensorflow:global_step/sec: 3.576\n",
            "Total loss is :[0.165439606]\n",
            "INFO:tensorflow:global_step/sec: 3.81302\n",
            "Total loss is :[0.212897494]\n",
            "INFO:tensorflow:global_step/sec: 3.7946\n",
            "Total loss is :[0.180677682]\n",
            "INFO:tensorflow:global_step/sec: 3.78896\n",
            "Total loss is :[0.24568288]\n",
            "INFO:tensorflow:global_step/sec: 3.83048\n",
            "Total loss is :[0.21255289]\n",
            "INFO:tensorflow:global_step/sec: 3.93726\n",
            "Total loss is :[0.270413488]\n",
            "INFO:tensorflow:global_step/sec: 3.70705\n",
            "Total loss is :[0.215509355]\n",
            "INFO:tensorflow:global_step/sec: 3.83742\n",
            "Total loss is :[0.244074613]\n",
            "INFO:tensorflow:global_step/sec: 3.87541\n",
            "Total loss is :[0.167086214]\n",
            "INFO:tensorflow:global_step/sec: 3.91037\n",
            "Total loss is :[0.19947575]\n",
            "INFO:tensorflow:global_step/sec: 3.76751\n",
            "Total loss is :[0.194812179]\n",
            "INFO:tensorflow:global_step/sec: 3.50434\n",
            "Total loss is :[0.287158489]\n",
            "INFO:tensorflow:global_step/sec: 3.41707\n",
            "Total loss is :[0.232774571]\n",
            "INFO:tensorflow:global_step/sec: 3.77205\n",
            "Total loss is :[0.236292064]\n",
            "INFO:tensorflow:global_step/sec: 3.67483\n",
            "Total loss is :[0.222671121]\n",
            "INFO:tensorflow:global_step/sec: 3.84881\n",
            "Total loss is :[0.182302624]\n",
            "INFO:tensorflow:global_step/sec: 3.80084\n",
            "Total loss is :[0.19216913]\n",
            "INFO:tensorflow:global_step/sec: 3.80654\n",
            "Total loss is :[0.187065274]\n",
            "INFO:tensorflow:global_step/sec: 3.59865\n",
            "Total loss is :[0.218502462]\n",
            "INFO:tensorflow:global_step/sec: 3.60757\n",
            "Total loss is :[0.305691332]\n",
            "INFO:tensorflow:global_step/sec: 3.58197\n",
            "Total loss is :[0.215967402]\n",
            "INFO:tensorflow:global_step/sec: 3.8484\n",
            "Total loss is :[0.240218371]\n",
            "INFO:tensorflow:global_step/sec: 3.8134\n",
            "Total loss is :[0.221932262]\n",
            "INFO:tensorflow:global_step/sec: 3.91141\n",
            "Total loss is :[0.233070761]\n",
            "INFO:tensorflow:global_step/sec: 3.71369\n",
            "Total loss is :[0.229094014]\n",
            "INFO:tensorflow:global_step/sec: 3.79527\n",
            "Total loss is :[0.232282802]\n",
            "INFO:tensorflow:global_step/sec: 3.9456\n",
            "Total loss is :[0.18798542]\n",
            "INFO:tensorflow:global_step/sec: 3.84976\n",
            "Total loss is :[0.223438412]\n",
            "INFO:tensorflow:global_step/sec: 3.6823\n",
            "Total loss is :[0.21023795]\n",
            "INFO:tensorflow:global_step/sec: 3.79635\n",
            "Total loss is :[0.2752617]\n",
            "INFO:tensorflow:global_step/sec: 3.81042\n",
            "Total loss is :[0.211712718]\n",
            "INFO:tensorflow:global_step/sec: 3.8415\n",
            "Total loss is :[0.225919902]\n",
            "INFO:tensorflow:global_step/sec: 3.91897\n",
            "Total loss is :[0.266757667]\n",
            "INFO:tensorflow:global_step/sec: 3.75367\n",
            "Total loss is :[0.187696129]\n",
            "INFO:tensorflow:global_step/sec: 3.66419\n",
            "Total loss is :[0.170724109]\n",
            "INFO:tensorflow:global_step/sec: 3.66501\n",
            "Total loss is :[0.23033309]\n",
            "INFO:tensorflow:global_step/sec: 3.84426\n",
            "Total loss is :[0.163624063]\n",
            "INFO:tensorflow:global_step/sec: 3.842\n",
            "Total loss is :[0.24541764]\n",
            "INFO:tensorflow:global_step/sec: 3.88559\n",
            "Total loss is :[0.20844698]\n",
            "INFO:tensorflow:global_step/sec: 3.72715\n",
            "Total loss is :[0.212029099]\n",
            "INFO:tensorflow:global_step/sec: 3.81575\n",
            "Total loss is :[0.298654675]\n",
            "INFO:tensorflow:global_step/sec: 3.75254\n",
            "Total loss is :[0.204989046]\n",
            "INFO:tensorflow:global_step/sec: 3.81827\n",
            "Total loss is :[0.260065019]\n",
            "INFO:tensorflow:global_step/sec: 3.69936\n",
            "Total loss is :[0.188789263]\n",
            "INFO:tensorflow:global_step/sec: 3.91069\n",
            "Total loss is :[0.300786078]\n",
            "INFO:tensorflow:global_step/sec: 3.73226\n",
            "Total loss is :[0.213937074]\n",
            "INFO:tensorflow:global_step/sec: 3.82836\n",
            "Total loss is :[0.608234823]\n",
            "INFO:tensorflow:global_step/sec: 3.82669\n",
            "Total loss is :[0.183456734]\n",
            "INFO:tensorflow:global_step/sec: 3.57005\n",
            "Total loss is :[0.278577387]\n",
            "INFO:tensorflow:global_step/sec: 3.65671\n",
            "Total loss is :[0.33803308]\n",
            "INFO:tensorflow:global_step/sec: 3.59012\n",
            "Total loss is :[0.187211663]\n",
            "INFO:tensorflow:global_step/sec: 3.85745\n",
            "Total loss is :[0.210390195]\n",
            "INFO:tensorflow:global_step/sec: 3.88163\n",
            "Total loss is :[0.34919852]\n",
            "INFO:tensorflow:global_step/sec: 3.8637\n",
            "Total loss is :[0.210450009]\n",
            "INFO:tensorflow:global_step/sec: 3.7469\n",
            "Total loss is :[0.175253198]\n",
            "INFO:tensorflow:global_step/sec: 3.70536\n",
            "Total loss is :[0.186314926]\n",
            "INFO:tensorflow:global_step/sec: 3.55811\n",
            "Total loss is :[0.358294755]\n",
            "INFO:tensorflow:global_step/sec: 3.89432\n",
            "Total loss is :[0.298154652]\n",
            "INFO:tensorflow:global_step/sec: 3.87422\n",
            "Total loss is :[0.323022842]\n",
            "INFO:tensorflow:global_step/sec: 4.03946\n",
            "Total loss is :[0.173002139]\n",
            "INFO:tensorflow:global_step/sec: 3.81923\n",
            "Total loss is :[0.179146424]\n",
            "INFO:tensorflow:global_step/sec: 3.69539\n",
            "Total loss is :[0.203839645]\n",
            "INFO:tensorflow:global_step/sec: 3.86811\n",
            "Total loss is :[0.204567105]\n",
            "INFO:tensorflow:global_step/sec: 3.77321\n",
            "Total loss is :[0.215692014]\n",
            "INFO:tensorflow:global_step/sec: 3.98765\n",
            "Total loss is :[0.192730367]\n",
            "INFO:tensorflow:global_step/sec: 3.81781\n",
            "Total loss is :[0.238688067]\n",
            "INFO:tensorflow:global_step/sec: 3.85518\n",
            "Total loss is :[0.239287198]\n",
            "INFO:tensorflow:global_step/sec: 3.80136\n",
            "Total loss is :[0.163181916]\n",
            "INFO:tensorflow:global_step/sec: 3.90224\n",
            "Total loss is :[0.214704096]\n",
            "INFO:tensorflow:global_step/sec: 3.91972\n",
            "Total loss is :[0.192806244]\n",
            "INFO:tensorflow:global_step/sec: 3.61773\n",
            "Total loss is :[0.351746351]\n",
            "INFO:tensorflow:global_step/sec: 3.90967\n",
            "Total loss is :[0.199589834]\n",
            "INFO:tensorflow:global_step/sec: 3.61854\n",
            "Total loss is :[0.254569173]\n",
            "INFO:tensorflow:global_step/sec: 3.85572\n",
            "Total loss is :[0.191569582]\n",
            "INFO:tensorflow:global_step/sec: 3.68798\n",
            "Total loss is :[0.303552568]\n",
            "INFO:tensorflow:global_step/sec: 3.81191\n",
            "Total loss is :[0.310999393]\n",
            "INFO:tensorflow:global_step/sec: 3.815\n",
            "Total loss is :[0.202451423]\n",
            "INFO:tensorflow:global_step/sec: 3.81051\n",
            "Total loss is :[0.249970183]\n",
            "INFO:tensorflow:global_step/sec: 3.72173\n",
            "Total loss is :[0.163631871]\n",
            "INFO:tensorflow:global_step/sec: 3.63409\n",
            "Total loss is :[0.221875429]\n",
            "INFO:tensorflow:global_step/sec: 3.67599\n",
            "Total loss is :[0.175353497]\n",
            "INFO:tensorflow:global_step/sec: 3.69412\n",
            "Total loss is :[0.200282067]\n",
            "INFO:tensorflow:global_step/sec: 3.74062\n",
            "Total loss is :[0.243264765]\n",
            "INFO:tensorflow:global_step/sec: 3.88052\n",
            "Total loss is :[0.260279983]\n",
            "INFO:tensorflow:global_step/sec: 3.7641\n",
            "Total loss is :[0.174007818]\n",
            "INFO:tensorflow:global_step/sec: 3.69402\n",
            "Total loss is :[0.175351709]\n",
            "INFO:tensorflow:global_step/sec: 3.83646\n",
            "Total loss is :[0.227334648]\n",
            "INFO:tensorflow:global_step/sec: 3.64581\n",
            "Total loss is :[0.173526868]\n",
            "INFO:tensorflow:global_step/sec: 3.70012\n",
            "Total loss is :[0.237425804]\n",
            "INFO:tensorflow:global_step/sec: 3.63956\n",
            "Total loss is :[0.263339877]\n",
            "INFO:tensorflow:global_step/sec: 3.61904\n",
            "Total loss is :[0.168469578]\n",
            "INFO:tensorflow:global_step/sec: 3.74046\n",
            "Total loss is :[0.221225992]\n",
            "INFO:tensorflow:global_step/sec: 3.90233\n",
            "Total loss is :[0.197252393]\n",
            "INFO:tensorflow:global_step/sec: 3.6305\n",
            "Total loss is :[0.21532613]\n",
            "INFO:tensorflow:global_step/sec: 3.91473\n",
            "Total loss is :[0.248640865]\n",
            "INFO:tensorflow:global_step/sec: 3.82511\n",
            "Total loss is :[0.18544665]\n",
            "INFO:tensorflow:global_step/sec: 4.01675\n",
            "Total loss is :[0.19169426]\n",
            "INFO:tensorflow:global_step/sec: 3.88518\n",
            "Total loss is :[0.187024087]\n",
            "INFO:tensorflow:global_step/sec: 3.73654\n",
            "Total loss is :[0.183317736]\n",
            "INFO:tensorflow:global_step/sec: 3.73852\n",
            "Total loss is :[0.223914117]\n",
            "INFO:tensorflow:global_step/sec: 3.73952\n",
            "Total loss is :[0.275109023]\n",
            "INFO:tensorflow:global_step/sec: 3.69624\n",
            "Total loss is :[0.23579441]\n",
            "INFO:tensorflow:global_step/sec: 3.75246\n",
            "Total loss is :[0.24530755]\n",
            "INFO:tensorflow:global_step/sec: 3.73458\n",
            "Total loss is :[0.251597762]\n",
            "INFO:tensorflow:global_step/sec: 4.08692\n",
            "Total loss is :[0.19766587]\n",
            "INFO:tensorflow:global_step/sec: 3.81971\n",
            "Total loss is :[0.242896244]\n",
            "INFO:tensorflow:global_step/sec: 3.83341\n",
            "Total loss is :[0.18894586]\n",
            "INFO:tensorflow:global_step/sec: 3.64828\n",
            "Total loss is :[0.194921255]\n",
            "INFO:tensorflow:global_step/sec: 3.60123\n",
            "Total loss is :[0.184841201]\n",
            "INFO:tensorflow:global_step/sec: 4.03642\n",
            "Total loss is :[0.169284]\n",
            "INFO:tensorflow:global_step/sec: 3.63897\n",
            "Total loss is :[0.259310484]\n",
            "INFO:tensorflow:global_step/sec: 3.68192\n",
            "Total loss is :[0.216203079]\n",
            "INFO:tensorflow:global_step/sec: 3.92257\n",
            "Total loss is :[0.241889745]\n",
            "INFO:tensorflow:global_step/sec: 3.78759\n",
            "Total loss is :[0.23924008]\n",
            "INFO:tensorflow:global_step/sec: 3.78202\n",
            "Total loss is :[0.191224545]\n",
            "INFO:tensorflow:global_step/sec: 3.61911\n",
            "Total loss is :[0.295775771]\n",
            "INFO:tensorflow:global_step/sec: 3.84829\n",
            "Total loss is :[0.26889205]\n",
            "INFO:tensorflow:global_step/sec: 3.77811\n",
            "Total loss is :[0.213576883]\n",
            "INFO:tensorflow:global_step/sec: 3.69928\n",
            "Total loss is :[0.224179476]\n",
            "INFO:tensorflow:global_step/sec: 3.68199\n",
            "Total loss is :[0.226922467]\n",
            "INFO:tensorflow:global_step/sec: 3.65268\n",
            "Total loss is :[0.186168447]\n",
            "INFO:tensorflow:global_step/sec: 3.88438\n",
            "Total loss is :[0.231491983]\n",
            "INFO:tensorflow:global_step/sec: 3.8795\n",
            "Total loss is :[0.183231726]\n",
            "INFO:tensorflow:global_step/sec: 3.72095\n",
            "Total loss is :[0.190107539]\n",
            "INFO:tensorflow:global_step/sec: 3.94579\n",
            "Total loss is :[0.181698382]\n",
            "INFO:tensorflow:global_step/sec: 3.71\n",
            "Total loss is :[0.325492263]\n",
            "INFO:tensorflow:global_step/sec: 3.71384\n",
            "Total loss is :[0.209942967]\n",
            "INFO:tensorflow:global_step/sec: 3.64558\n",
            "Total loss is :[0.18925938]\n",
            "INFO:tensorflow:global_step/sec: 3.60638\n",
            "Total loss is :[0.235622793]\n",
            "INFO:tensorflow:global_step/sec: 3.60474\n",
            "Total loss is :[0.201715887]\n",
            "INFO:tensorflow:global_step/sec: 3.51815\n",
            "Total loss is :[0.184630603]\n",
            "INFO:tensorflow:global_step/sec: 3.77673\n",
            "Total loss is :[0.235254779]\n",
            "INFO:tensorflow:global_step/sec: 3.64926\n",
            "Total loss is :[0.1671931]\n",
            "INFO:tensorflow:global_step/sec: 3.71855\n",
            "Total loss is :[0.224963203]\n",
            "INFO:tensorflow:global_step/sec: 3.68542\n",
            "Total loss is :[0.188265]\n",
            "INFO:tensorflow:global_step/sec: 3.71082\n",
            "Total loss is :[0.309140474]\n",
            "INFO:tensorflow:global_step/sec: 3.74746\n",
            "Total loss is :[0.27146554]\n",
            "INFO:tensorflow:global_step/sec: 3.7678\n",
            "Total loss is :[0.16971752]\n",
            "INFO:tensorflow:global_step/sec: 3.67723\n",
            "Total loss is :[0.242156923]\n",
            "INFO:tensorflow:global_step/sec: 3.65587\n",
            "Total loss is :[0.171336487]\n",
            "INFO:tensorflow:global_step/sec: 3.64339\n",
            "Total loss is :[0.178979054]\n",
            "INFO:tensorflow:global_step/sec: 3.7785\n",
            "Total loss is :[0.212654695]\n",
            "INFO:tensorflow:global_step/sec: 3.85977\n",
            "Total loss is :[0.219989419]\n",
            "INFO:tensorflow:global_step/sec: 3.92044\n",
            "Total loss is :[0.174713016]\n",
            "INFO:tensorflow:global_step/sec: 3.88533\n",
            "Total loss is :[0.206397384]\n",
            "INFO:tensorflow:global_step/sec: 3.8372\n",
            "Total loss is :[0.238432765]\n",
            "INFO:tensorflow:global_step/sec: 3.72776\n",
            "Total loss is :[0.283972114]\n",
            "INFO:tensorflow:global_step/sec: 3.76896\n",
            "Total loss is :[0.191026375]\n",
            "INFO:tensorflow:global_step/sec: 3.74727\n",
            "Total loss is :[0.218048468]\n",
            "INFO:tensorflow:global_step/sec: 3.8897\n",
            "Total loss is :[0.206091434]\n",
            "INFO:tensorflow:global_step/sec: 3.97715\n",
            "Total loss is :[0.266801059]\n",
            "INFO:tensorflow:global_step/sec: 3.91212\n",
            "Total loss is :[0.179753289]\n",
            "INFO:tensorflow:global_step/sec: 3.60723\n",
            "Total loss is :[0.206579]\n",
            "INFO:tensorflow:global_step/sec: 3.77437\n",
            "Total loss is :[0.24145788]\n",
            "INFO:tensorflow:global_step/sec: 4.01664\n",
            "Total loss is :[0.195511356]\n",
            "INFO:tensorflow:global_step/sec: 3.88535\n",
            "Total loss is :[0.371446371]\n",
            "INFO:tensorflow:global_step/sec: 3.64194\n",
            "Total loss is :[0.173925027]\n",
            "INFO:tensorflow:global_step/sec: 3.9538\n",
            "Total loss is :[0.520899236]\n",
            "INFO:tensorflow:global_step/sec: 3.64618\n",
            "Total loss is :[0.194637835]\n",
            "INFO:tensorflow:global_step/sec: 3.83885\n",
            "Total loss is :[0.196484402]\n",
            "INFO:tensorflow:global_step/sec: 3.69098\n",
            "Total loss is :[0.370076656]\n",
            "INFO:tensorflow:global_step/sec: 3.89676\n",
            "Total loss is :[0.215785116]\n",
            "INFO:tensorflow:global_step/sec: 3.74618\n",
            "Total loss is :[0.194240749]\n",
            "INFO:tensorflow:global_step/sec: 3.82198\n",
            "Total loss is :[0.24560599]\n",
            "INFO:tensorflow:global_step/sec: 3.95175\n",
            "Total loss is :[0.258910537]\n",
            "INFO:tensorflow:global_step/sec: 3.87684\n",
            "Total loss is :[0.278436273]\n",
            "INFO:tensorflow:global_step/sec: 3.61845\n",
            "Total loss is :[0.247569859]\n",
            "INFO:tensorflow:global_step/sec: 4.06441\n",
            "Total loss is :[0.252988]\n",
            "INFO:tensorflow:global_step/sec: 3.96\n",
            "Total loss is :[0.228451923]\n",
            "INFO:tensorflow:global_step/sec: 3.73378\n",
            "Total loss is :[0.206240013]\n",
            "INFO:tensorflow:global_step/sec: 3.59895\n",
            "Total loss is :[0.230164915]\n",
            "INFO:tensorflow:global_step/sec: 3.69982\n",
            "Total loss is :[0.242072523]\n",
            "INFO:tensorflow:global_step/sec: 3.85151\n",
            "Total loss is :[0.213561907]\n",
            "INFO:tensorflow:global_step/sec: 3.83034\n",
            "Total loss is :[0.200867712]\n",
            "INFO:tensorflow:global_step/sec: 3.78153\n",
            "Total loss is :[0.212672248]\n",
            "INFO:tensorflow:global_step/sec: 3.70845\n",
            "Total loss is :[0.401419282]\n",
            "INFO:tensorflow:global_step/sec: 3.73541\n",
            "Total loss is :[0.191384986]\n",
            "INFO:tensorflow:global_step/sec: 3.82315\n",
            "Total loss is :[0.172466218]\n",
            "INFO:tensorflow:global_step/sec: 3.64035\n",
            "Total loss is :[0.192606866]\n",
            "INFO:tensorflow:global_step/sec: 3.79841\n",
            "Total loss is :[0.182208925]\n",
            "INFO:tensorflow:global_step/sec: 3.82004\n",
            "Total loss is :[0.200151682]\n",
            "INFO:tensorflow:global_step/sec: 3.81911\n",
            "Total loss is :[0.203035265]\n",
            "INFO:tensorflow:global_step/sec: 3.924\n",
            "Total loss is :[0.236942172]\n",
            "INFO:tensorflow:global_step/sec: 3.87218\n",
            "Total loss is :[0.170152619]\n",
            "INFO:tensorflow:global_step/sec: 3.68911\n",
            "Total loss is :[0.161384508]\n",
            "INFO:tensorflow:global_step/sec: 3.93326\n",
            "Total loss is :[0.177308872]\n",
            "INFO:tensorflow:global_step/sec: 3.82668\n",
            "Total loss is :[0.201919794]\n",
            "INFO:tensorflow:global_step/sec: 3.97945\n",
            "Total loss is :[0.302294612]\n",
            "INFO:tensorflow:global_step/sec: 3.79229\n",
            "Total loss is :[0.245903105]\n",
            "INFO:tensorflow:global_step/sec: 3.77724\n",
            "Total loss is :[0.248992562]\n",
            "INFO:tensorflow:global_step/sec: 3.73647\n",
            "Total loss is :[0.209296256]\n",
            "INFO:tensorflow:global_step/sec: 3.62789\n",
            "Total loss is :[0.228508011]\n",
            "INFO:tensorflow:global_step/sec: 3.80613\n",
            "Total loss is :[0.185714692]\n",
            "INFO:tensorflow:global_step/sec: 3.68448\n",
            "Total loss is :[0.188452899]\n",
            "INFO:tensorflow:global_step/sec: 3.78347\n",
            "Total loss is :[0.227794856]\n",
            "INFO:tensorflow:global_step/sec: 3.88431\n",
            "Total loss is :[0.223686203]\n",
            "INFO:tensorflow:global_step/sec: 3.78644\n",
            "Total loss is :[0.242130548]\n",
            "INFO:tensorflow:global_step/sec: 3.66626\n",
            "Total loss is :[0.185412392]\n",
            "INFO:tensorflow:global_step/sec: 3.73308\n",
            "Total loss is :[0.190005615]\n",
            "INFO:tensorflow:global_step/sec: 3.75732\n",
            "Total loss is :[0.223852068]\n",
            "INFO:tensorflow:global_step/sec: 3.88333\n",
            "Total loss is :[0.198214471]\n",
            "INFO:tensorflow:global_step/sec: 3.61585\n",
            "Total loss is :[0.348533213]\n",
            "INFO:tensorflow:global_step/sec: 3.75729\n",
            "Total loss is :[0.196133196]\n",
            "INFO:tensorflow:global_step/sec: 3.81149\n",
            "Total loss is :[0.189418867]\n",
            "INFO:tensorflow:global_step/sec: 4.02506\n",
            "Total loss is :[0.210920498]\n",
            "INFO:tensorflow:global_step/sec: 3.73647\n",
            "Total loss is :[0.205003515]\n",
            "INFO:tensorflow:global_step/sec: 3.7659\n",
            "Total loss is :[0.187087119]\n",
            "INFO:tensorflow:global_step/sec: 3.69629\n",
            "Total loss is :[0.224435255]\n",
            "INFO:tensorflow:global_step/sec: 3.80496\n",
            "Total loss is :[0.217551932]\n",
            "INFO:tensorflow:global_step/sec: 3.80058\n",
            "Total loss is :[0.269537121]\n",
            "INFO:tensorflow:global_step/sec: 4.03187\n",
            "Total loss is :[0.226489931]\n",
            "INFO:tensorflow:global_step/sec: 3.79329\n",
            "Total loss is :[0.169045046]\n",
            "INFO:tensorflow:global_step/sec: 3.83433\n",
            "Total loss is :[0.358354837]\n",
            "INFO:tensorflow:global_step/sec: 3.91528\n",
            "Total loss is :[0.196143031]\n",
            "INFO:tensorflow:global_step/sec: 3.72082\n",
            "Total loss is :[0.257281363]\n",
            "INFO:tensorflow:global_step/sec: 3.83924\n",
            "Total loss is :[0.189605176]\n",
            "INFO:tensorflow:global_step/sec: 3.63186\n",
            "Total loss is :[0.17273359]\n",
            "INFO:tensorflow:global_step/sec: 3.61029\n",
            "Total loss is :[0.192579269]\n",
            "INFO:tensorflow:global_step/sec: 3.85391\n",
            "Total loss is :[0.160250887]\n",
            "INFO:tensorflow:global_step/sec: 4.01296\n",
            "Total loss is :[0.248304546]\n",
            "INFO:tensorflow:global_step/sec: 3.67506\n",
            "Total loss is :[0.258895487]\n",
            "INFO:tensorflow:global_step/sec: 3.71422\n",
            "Total loss is :[0.183793709]\n",
            "INFO:tensorflow:global_step/sec: 3.85151\n",
            "Total loss is :[0.200029776]\n",
            "INFO:tensorflow:global_step/sec: 3.67685\n",
            "Total loss is :[0.172161251]\n",
            "INFO:tensorflow:global_step/sec: 3.59135\n",
            "Total loss is :[0.372072399]\n",
            "INFO:tensorflow:global_step/sec: 3.57981\n",
            "Total loss is :[0.294784]\n",
            "INFO:tensorflow:global_step/sec: 3.6563\n",
            "Total loss is :[0.265850067]\n",
            "INFO:tensorflow:global_step/sec: 3.72784\n",
            "Total loss is :[0.212187111]\n",
            "INFO:tensorflow:global_step/sec: 3.76943\n",
            "Total loss is :[0.401399255]\n",
            "INFO:tensorflow:global_step/sec: 3.71437\n",
            "Total loss is :[0.212991595]\n",
            "INFO:tensorflow:global_step/sec: 3.68971\n",
            "Total loss is :[0.189063624]\n",
            "INFO:tensorflow:global_step/sec: 3.71326\n",
            "Total loss is :[0.17358683]\n",
            "INFO:tensorflow:global_step/sec: 3.66876\n",
            "Total loss is :[0.210159421]\n",
            "INFO:tensorflow:global_step/sec: 3.7949\n",
            "Total loss is :[0.192385033]\n",
            "INFO:tensorflow:global_step/sec: 3.63402\n",
            "Total loss is :[0.207479358]\n",
            "INFO:tensorflow:global_step/sec: 3.7733\n",
            "Total loss is :[0.198717654]\n",
            "INFO:tensorflow:global_step/sec: 3.77278\n",
            "Total loss is :[0.248420477]\n",
            "INFO:tensorflow:global_step/sec: 3.74082\n",
            "Total loss is :[0.478751123]\n",
            "INFO:tensorflow:global_step/sec: 3.60417\n",
            "Total loss is :[0.159342602]\n",
            "INFO:tensorflow:global_step/sec: 3.69767\n",
            "Total loss is :[0.172490358]\n",
            "INFO:tensorflow:global_step/sec: 3.82104\n",
            "Total loss is :[0.242881715]\n",
            "INFO:tensorflow:global_step/sec: 3.91627\n",
            "Total loss is :[0.178707108]\n",
            "INFO:tensorflow:global_step/sec: 3.90665\n",
            "Total loss is :[0.187193334]\n",
            "INFO:tensorflow:global_step/sec: 3.85714\n",
            "Total loss is :[0.187904969]\n",
            "INFO:tensorflow:global_step/sec: 3.62938\n",
            "Total loss is :[0.211021364]\n",
            "INFO:tensorflow:global_step/sec: 3.86374\n",
            "Total loss is :[0.180728465]\n",
            "INFO:tensorflow:global_step/sec: 3.72884\n",
            "Total loss is :[0.226273119]\n",
            "INFO:tensorflow:global_step/sec: 3.40728\n",
            "Total loss is :[0.170069158]\n",
            "INFO:tensorflow:global_step/sec: 3.79515\n",
            "Total loss is :[0.211857095]\n",
            "INFO:tensorflow:global_step/sec: 3.92507\n",
            "Total loss is :[0.198775768]\n",
            "INFO:tensorflow:global_step/sec: 3.82237\n",
            "Total loss is :[0.239717767]\n",
            "INFO:tensorflow:global_step/sec: 3.74263\n",
            "Total loss is :[0.332659811]\n",
            "INFO:tensorflow:global_step/sec: 3.76912\n",
            "Total loss is :[0.250337899]\n",
            "INFO:tensorflow:global_step/sec: 3.88931\n",
            "Total loss is :[0.505641162]\n",
            "INFO:tensorflow:global_step/sec: 3.94567\n",
            "Total loss is :[0.267782956]\n",
            "INFO:tensorflow:global_step/sec: 3.79425\n",
            "Total loss is :[0.218567759]\n",
            "INFO:tensorflow:global_step/sec: 3.67747\n",
            "Total loss is :[0.202268347]\n",
            "INFO:tensorflow:global_step/sec: 3.80577\n",
            "Total loss is :[0.357466668]\n",
            "INFO:tensorflow:global_step/sec: 3.93549\n",
            "Total loss is :[0.186502188]\n",
            "INFO:tensorflow:global_step/sec: 3.9943\n",
            "Total loss is :[0.176311761]\n",
            "INFO:tensorflow:global_step/sec: 3.56313\n",
            "Total loss is :[0.178152502]\n",
            "INFO:tensorflow:global_step/sec: 3.64272\n",
            "Total loss is :[0.224140033]\n",
            "INFO:tensorflow:global_step/sec: 3.45435\n",
            "Total loss is :[0.21727477]\n",
            "INFO:tensorflow:global_step/sec: 3.70835\n",
            "Total loss is :[0.23927334]\n",
            "INFO:tensorflow:global_step/sec: 3.81549\n",
            "Total loss is :[0.22111246]\n",
            "INFO:tensorflow:global_step/sec: 3.78707\n",
            "Total loss is :[0.269637674]\n",
            "INFO:tensorflow:global_step/sec: 3.58948\n",
            "Total loss is :[0.191950142]\n",
            "INFO:tensorflow:global_step/sec: 3.69247\n",
            "Total loss is :[0.185709804]\n",
            "INFO:tensorflow:global_step/sec: 3.91328\n",
            "Total loss is :[0.156683937]\n",
            "INFO:tensorflow:global_step/sec: 3.87974\n",
            "INFO:tensorflow:Saving checkpoints for 13513 into /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt.\n",
            "Total loss is :[0.215065077]\n",
            "INFO:tensorflow:global_step/sec: 2.74957\n",
            "Total loss is :[0.183438361]\n",
            "INFO:tensorflow:global_step/sec: 3.85661\n",
            "Total loss is :[0.219606608]\n",
            "INFO:tensorflow:global_step/sec: 3.83189\n",
            "Total loss is :[0.415853351]\n",
            "INFO:tensorflow:global_step/sec: 3.83514\n",
            "Total loss is :[0.283078849]\n",
            "INFO:tensorflow:global_step/sec: 3.61332\n",
            "Total loss is :[0.251105756]\n",
            "INFO:tensorflow:global_step/sec: 3.66002\n",
            "Total loss is :[0.187071592]\n",
            "INFO:tensorflow:global_step/sec: 3.85837\n",
            "Total loss is :[0.247208267]\n",
            "INFO:tensorflow:global_step/sec: 3.77264\n",
            "Total loss is :[0.218811929]\n",
            "INFO:tensorflow:global_step/sec: 3.79781\n",
            "Total loss is :[0.217912242]\n",
            "INFO:tensorflow:global_step/sec: 3.71892\n",
            "Total loss is :[0.245418325]\n",
            "INFO:tensorflow:global_step/sec: 3.68465\n",
            "Total loss is :[0.226621464]\n",
            "INFO:tensorflow:global_step/sec: 3.89536\n",
            "Total loss is :[0.235601395]\n",
            "INFO:tensorflow:global_step/sec: 3.83398\n",
            "Total loss is :[0.174839407]\n",
            "INFO:tensorflow:global_step/sec: 3.82635\n",
            "Total loss is :[0.303195149]\n",
            "INFO:tensorflow:global_step/sec: 3.71667\n",
            "Total loss is :[0.216830894]\n",
            "INFO:tensorflow:global_step/sec: 3.80245\n",
            "Total loss is :[0.225031063]\n",
            "INFO:tensorflow:global_step/sec: 3.79073\n",
            "Total loss is :[0.218945146]\n",
            "INFO:tensorflow:global_step/sec: 3.93957\n",
            "Total loss is :[0.174389094]\n",
            "INFO:tensorflow:global_step/sec: 3.78067\n",
            "Total loss is :[0.219669968]\n",
            "INFO:tensorflow:global_step/sec: 3.73807\n",
            "Total loss is :[0.287739]\n",
            "INFO:tensorflow:global_step/sec: 3.67273\n",
            "Total loss is :[0.188415468]\n",
            "INFO:tensorflow:global_step/sec: 3.71573\n",
            "Total loss is :[0.237256631]\n",
            "INFO:tensorflow:global_step/sec: 3.67705\n",
            "Total loss is :[0.213034093]\n",
            "INFO:tensorflow:global_step/sec: 3.86096\n",
            "Total loss is :[0.164392099]\n",
            "INFO:tensorflow:global_step/sec: 3.59482\n",
            "Total loss is :[0.205320016]\n",
            "INFO:tensorflow:global_step/sec: 3.6793\n",
            "Total loss is :[0.211326271]\n",
            "INFO:tensorflow:global_step/sec: 3.79665\n",
            "Total loss is :[0.19980891]\n",
            "INFO:tensorflow:global_step/sec: 3.74489\n",
            "Total loss is :[0.182668984]\n",
            "INFO:tensorflow:global_step/sec: 3.67557\n",
            "Total loss is :[0.332436293]\n",
            "INFO:tensorflow:global_step/sec: 3.71542\n",
            "Total loss is :[0.214218542]\n",
            "INFO:tensorflow:global_step/sec: 3.76564\n",
            "Total loss is :[0.177186757]\n",
            "INFO:tensorflow:global_step/sec: 3.77843\n",
            "Total loss is :[0.19743593]\n",
            "INFO:tensorflow:global_step/sec: 3.75903\n",
            "Total loss is :[0.18691799]\n",
            "INFO:tensorflow:global_step/sec: 3.75322\n",
            "Total loss is :[0.168432713]\n",
            "INFO:tensorflow:global_step/sec: 3.75486\n",
            "Total loss is :[0.228489801]\n",
            "INFO:tensorflow:global_step/sec: 3.73737\n",
            "Total loss is :[1.21493602]\n",
            "INFO:tensorflow:global_step/sec: 3.68435\n",
            "Total loss is :[0.209934935]\n",
            "INFO:tensorflow:global_step/sec: 3.76325\n",
            "Total loss is :[0.243238032]\n",
            "INFO:tensorflow:global_step/sec: 3.88424\n",
            "Total loss is :[0.183431938]\n",
            "INFO:tensorflow:global_step/sec: 3.78301\n",
            "Total loss is :[0.221186966]\n",
            "INFO:tensorflow:global_step/sec: 3.96073\n",
            "Total loss is :[0.16267243]\n",
            "INFO:tensorflow:global_step/sec: 3.89716\n",
            "Total loss is :[0.225846469]\n",
            "INFO:tensorflow:global_step/sec: 3.70645\n",
            "Total loss is :[0.182894558]\n",
            "INFO:tensorflow:global_step/sec: 3.94479\n",
            "Total loss is :[0.235972732]\n",
            "INFO:tensorflow:global_step/sec: 3.66055\n",
            "Total loss is :[0.224611908]\n",
            "INFO:tensorflow:global_step/sec: 3.83404\n",
            "Total loss is :[0.191368774]\n",
            "INFO:tensorflow:global_step/sec: 3.79623\n",
            "Total loss is :[0.188322306]\n",
            "INFO:tensorflow:global_step/sec: 3.90737\n",
            "Total loss is :[0.214441821]\n",
            "INFO:tensorflow:global_step/sec: 3.71765\n",
            "Total loss is :[0.183644637]\n",
            "INFO:tensorflow:global_step/sec: 3.83049\n",
            "Total loss is :[0.172668174]\n",
            "INFO:tensorflow:global_step/sec: 3.59856\n",
            "Total loss is :[0.172506973]\n",
            "INFO:tensorflow:global_step/sec: 3.70155\n",
            "Total loss is :[0.174281567]\n",
            "INFO:tensorflow:global_step/sec: 3.67371\n",
            "Total loss is :[0.270489246]\n",
            "INFO:tensorflow:global_step/sec: 3.63243\n",
            "Total loss is :[0.186858118]\n",
            "INFO:tensorflow:global_step/sec: 3.76306\n",
            "Total loss is :[0.209134787]\n",
            "INFO:tensorflow:global_step/sec: 3.97675\n",
            "Total loss is :[0.170863748]\n",
            "INFO:tensorflow:global_step/sec: 3.86725\n",
            "Total loss is :[0.205653295]\n",
            "INFO:tensorflow:global_step/sec: 3.77183\n",
            "Total loss is :[0.246623233]\n",
            "INFO:tensorflow:global_step/sec: 3.8483\n",
            "Total loss is :[0.207692772]\n",
            "INFO:tensorflow:global_step/sec: 3.82689\n",
            "Total loss is :[0.260945886]\n",
            "INFO:tensorflow:global_step/sec: 3.78336\n",
            "Total loss is :[0.173754469]\n",
            "INFO:tensorflow:global_step/sec: 3.83642\n",
            "Total loss is :[0.205512926]\n",
            "INFO:tensorflow:global_step/sec: 3.88809\n",
            "Total loss is :[0.199731559]\n",
            "INFO:tensorflow:global_step/sec: 3.9615\n",
            "Total loss is :[0.21224983]\n",
            "INFO:tensorflow:global_step/sec: 3.79049\n",
            "Total loss is :[0.205871731]\n",
            "INFO:tensorflow:global_step/sec: 3.81353\n",
            "Total loss is :[0.180755943]\n",
            "INFO:tensorflow:global_step/sec: 3.89464\n",
            "Total loss is :[0.190313101]\n",
            "INFO:tensorflow:global_step/sec: 3.89238\n",
            "Total loss is :[0.194191232]\n",
            "INFO:tensorflow:global_step/sec: 3.8521\n",
            "Total loss is :[0.296078742]\n",
            "INFO:tensorflow:global_step/sec: 3.7091\n",
            "Total loss is :[0.308990151]\n",
            "INFO:tensorflow:global_step/sec: 3.66941\n",
            "Total loss is :[0.197204]\n",
            "INFO:tensorflow:global_step/sec: 3.82674\n",
            "Total loss is :[0.18882142]\n",
            "INFO:tensorflow:global_step/sec: 3.7822\n",
            "Total loss is :[0.197458342]\n",
            "INFO:tensorflow:global_step/sec: 3.77148\n",
            "Total loss is :[0.186047971]\n",
            "INFO:tensorflow:global_step/sec: 3.85646\n",
            "Total loss is :[0.450030684]\n",
            "INFO:tensorflow:global_step/sec: 3.93269\n",
            "Total loss is :[0.276396036]\n",
            "INFO:tensorflow:global_step/sec: 3.7623\n",
            "Total loss is :[0.218934834]\n",
            "INFO:tensorflow:global_step/sec: 3.85625\n",
            "Total loss is :[0.208628446]\n",
            "INFO:tensorflow:global_step/sec: 3.8413\n",
            "Total loss is :[0.25938189]\n",
            "INFO:tensorflow:global_step/sec: 3.72976\n",
            "Total loss is :[0.171546549]\n",
            "INFO:tensorflow:global_step/sec: 3.74045\n",
            "Total loss is :[0.184971988]\n",
            "INFO:tensorflow:global_step/sec: 3.75544\n",
            "Total loss is :[0.233880013]\n",
            "INFO:tensorflow:global_step/sec: 3.64976\n",
            "Total loss is :[0.256135315]\n",
            "INFO:tensorflow:global_step/sec: 3.35437\n",
            "Total loss is :[0.260121733]\n",
            "INFO:tensorflow:global_step/sec: 3.82568\n",
            "Total loss is :[0.187237978]\n",
            "INFO:tensorflow:global_step/sec: 3.80803\n",
            "Total loss is :[0.188039035]\n",
            "INFO:tensorflow:global_step/sec: 3.76779\n",
            "Total loss is :[0.205637544]\n",
            "INFO:tensorflow:global_step/sec: 3.72922\n",
            "Total loss is :[0.221104115]\n",
            "INFO:tensorflow:global_step/sec: 3.57056\n",
            "Total loss is :[0.272668272]\n",
            "INFO:tensorflow:global_step/sec: 3.8431\n",
            "Total loss is :[0.196025506]\n",
            "INFO:tensorflow:global_step/sec: 3.75777\n",
            "Total loss is :[0.183670565]\n",
            "INFO:tensorflow:global_step/sec: 3.63903\n",
            "Total loss is :[0.196049109]\n",
            "INFO:tensorflow:global_step/sec: 3.83402\n",
            "Total loss is :[0.170190334]\n",
            "INFO:tensorflow:global_step/sec: 3.47316\n",
            "Total loss is :[0.33668378]\n",
            "INFO:tensorflow:global_step/sec: 3.56333\n",
            "Total loss is :[0.195963427]\n",
            "INFO:tensorflow:global_step/sec: 3.624\n",
            "Total loss is :[0.197968021]\n",
            "INFO:tensorflow:global_step/sec: 3.87825\n",
            "Total loss is :[0.220211953]\n",
            "INFO:tensorflow:global_step/sec: 3.74913\n",
            "Total loss is :[0.222635537]\n",
            "INFO:tensorflow:global_step/sec: 3.96707\n",
            "Total loss is :[0.172664195]\n",
            "INFO:tensorflow:global_step/sec: 3.66985\n",
            "Total loss is :[0.230621397]\n",
            "INFO:tensorflow:global_step/sec: 3.84427\n",
            "Total loss is :[0.198401928]\n",
            "INFO:tensorflow:global_step/sec: 3.8639\n",
            "Total loss is :[0.228527099]\n",
            "INFO:tensorflow:global_step/sec: 3.76429\n",
            "Total loss is :[0.191071987]\n",
            "INFO:tensorflow:global_step/sec: 3.83279\n",
            "Total loss is :[0.18282491]\n",
            "INFO:tensorflow:global_step/sec: 3.6748\n",
            "Total loss is :[0.229646608]\n",
            "INFO:tensorflow:global_step/sec: 3.8334\n",
            "Total loss is :[0.181949198]\n",
            "INFO:tensorflow:global_step/sec: 3.81333\n",
            "Total loss is :[0.198619947]\n",
            "INFO:tensorflow:global_step/sec: 3.62348\n",
            "Total loss is :[0.207766786]\n",
            "INFO:tensorflow:global_step/sec: 3.86027\n",
            "Total loss is :[0.189455494]\n",
            "INFO:tensorflow:global_step/sec: 3.73079\n",
            "Total loss is :[0.193952203]\n",
            "INFO:tensorflow:global_step/sec: 3.87446\n",
            "Total loss is :[0.207339659]\n",
            "INFO:tensorflow:global_step/sec: 3.69028\n",
            "Total loss is :[0.213443473]\n",
            "INFO:tensorflow:global_step/sec: 3.66969\n",
            "Total loss is :[0.223872617]\n",
            "INFO:tensorflow:global_step/sec: 3.63746\n",
            "Total loss is :[0.240299672]\n",
            "INFO:tensorflow:global_step/sec: 3.82543\n",
            "Total loss is :[0.183856189]\n",
            "INFO:tensorflow:global_step/sec: 3.75664\n",
            "Total loss is :[0.256342]\n",
            "INFO:tensorflow:global_step/sec: 3.77114\n",
            "Total loss is :[0.222131222]\n",
            "INFO:tensorflow:global_step/sec: 3.7811\n",
            "Total loss is :[0.18180123]\n",
            "INFO:tensorflow:global_step/sec: 3.7184\n",
            "Total loss is :[0.197616875]\n",
            "INFO:tensorflow:global_step/sec: 3.90556\n",
            "Total loss is :[0.456070781]\n",
            "INFO:tensorflow:global_step/sec: 3.8675\n",
            "Total loss is :[0.18415843]\n",
            "INFO:tensorflow:global_step/sec: 3.84286\n",
            "Total loss is :[0.216147229]\n",
            "INFO:tensorflow:global_step/sec: 3.80913\n",
            "Total loss is :[0.23392725]\n",
            "INFO:tensorflow:global_step/sec: 3.69947\n",
            "Total loss is :[0.198366821]\n",
            "INFO:tensorflow:global_step/sec: 3.78486\n",
            "Total loss is :[0.21200864]\n",
            "INFO:tensorflow:global_step/sec: 3.75899\n",
            "Total loss is :[0.206001818]\n",
            "INFO:tensorflow:global_step/sec: 3.69641\n",
            "Total loss is :[0.203934252]\n",
            "INFO:tensorflow:global_step/sec: 3.72738\n",
            "Total loss is :[0.164964586]\n",
            "INFO:tensorflow:global_step/sec: 3.66962\n",
            "Total loss is :[0.223077819]\n",
            "INFO:tensorflow:global_step/sec: 3.90149\n",
            "Total loss is :[0.185740933]\n",
            "INFO:tensorflow:global_step/sec: 3.64785\n",
            "Total loss is :[0.187253579]\n",
            "INFO:tensorflow:global_step/sec: 3.66384\n",
            "Total loss is :[0.220431373]\n",
            "INFO:tensorflow:global_step/sec: 3.7343\n",
            "Total loss is :[0.330614328]\n",
            "INFO:tensorflow:global_step/sec: 3.88726\n",
            "Total loss is :[0.183776036]\n",
            "INFO:tensorflow:global_step/sec: 3.99912\n",
            "Total loss is :[0.216458052]\n",
            "INFO:tensorflow:global_step/sec: 3.63518\n",
            "Total loss is :[0.24406223]\n",
            "INFO:tensorflow:global_step/sec: 3.80983\n",
            "Total loss is :[0.175847083]\n",
            "INFO:tensorflow:global_step/sec: 3.74828\n",
            "Total loss is :[0.198657349]\n",
            "INFO:tensorflow:global_step/sec: 3.73251\n",
            "Total loss is :[0.210857391]\n",
            "INFO:tensorflow:global_step/sec: 3.72902\n",
            "Total loss is :[0.21878612]\n",
            "INFO:tensorflow:global_step/sec: 3.93431\n",
            "Total loss is :[0.208367139]\n",
            "INFO:tensorflow:global_step/sec: 3.49943\n",
            "Total loss is :[0.211609349]\n",
            "INFO:tensorflow:global_step/sec: 3.65955\n",
            "Total loss is :[0.190339908]\n",
            "INFO:tensorflow:global_step/sec: 3.14607\n",
            "Total loss is :[0.168307692]\n",
            "INFO:tensorflow:global_step/sec: 3.80726\n",
            "Total loss is :[0.636974216]\n",
            "INFO:tensorflow:global_step/sec: 3.66422\n",
            "Total loss is :[0.257145494]\n",
            "INFO:tensorflow:global_step/sec: 3.67345\n",
            "Total loss is :[0.191629007]\n",
            "INFO:tensorflow:global_step/sec: 3.8478\n",
            "INFO:tensorflow:Saving checkpoints for 15000 into /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt.\n",
            "*********** Rizan Eval kortase ************\n",
            "\n",
            "WARNING: The TensorFlow contrib module will not be included in TensorFlow 2.0.\n",
            "For more information, please see:\n",
            "  * https://github.com/tensorflow/community/blob/master/rfcs/20180907-contrib-sunset.md\n",
            "  * https://github.com/tensorflow/addons\n",
            "If you depend on functionality not listed there, please file an issue.\n",
            "\n",
            "INFO:tensorflow:Evaluating on val set\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/control_flow_ops.py:423: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n",
            "INFO:tensorflow:Performing single-scale test.\n",
            "WARNING:tensorflow:From /root/models/research/deeplab/core/feature_extractor.py:196: to_float (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/metrics_impl.py:259: to_int64 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/confusion_matrix.py:194: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/metrics_impl.py:1155: div (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Deprecated in favor of operator or tf.math.divide.\n",
            "WARNING:tensorflow:From /root/models/research/deeplab/eval.py:170: print_model_analysis (from tensorflow.contrib.tfprof.model_analyzer) is deprecated and will be removed after 2018-01-01.\n",
            "Instructions for updating:\n",
            "Use `tf.profiler.profile(graph, run_meta, op_log, cmd, options)`. Build `options` with `tf.profiler.ProfileOptionBuilder`. See README.md for details\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/profiler/internal/flops_registry.py:142: tensor_shape_from_node_def_name (from tensorflow.python.framework.graph_util_impl) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.compat.v1.graph_util.remove_training_nodes\n",
            "4 ops no flops stats due to incomplete shapes.\n",
            "Parsing Inputs...\n",
            "\n",
            "=========================Options=============================\n",
            "-max_depth                  10000\n",
            "-min_bytes                  0\n",
            "-min_peak_bytes             0\n",
            "-min_residual_bytes         0\n",
            "-min_output_bytes           0\n",
            "-min_micros                 0\n",
            "-min_accelerator_micros     0\n",
            "-min_cpu_micros             0\n",
            "-min_params                 0\n",
            "-min_float_ops              0\n",
            "-min_occurrence             0\n",
            "-step                       -1\n",
            "-order_by                   name\n",
            "-account_type_regexes       _trainable_variables\n",
            "-start_name_regexes         .*\n",
            "-trim_name_regexes          \n",
            "-show_name_regexes          .*\n",
            "-hide_name_regexes          \n",
            "-account_displayed_op_only  true\n",
            "-select                     params\n",
            "-output                     stdout:\n",
            "\n",
            "==================Model Analysis Report======================\n",
            "\n",
            "Doc:\n",
            "scope: The nodes in the model graph are organized by their names, which is hierarchical like filesystem.\n",
            "param: Number of parameters (in the Variable).\n",
            "\n",
            "Profile:\n",
            "node name | # parameters\n",
            "_TFProfRoot (--/2.11m params)\n",
            "  MobilenetV2 (--/1.81m params)\n",
            "    MobilenetV2/Conv (--/928 params)\n",
            "      MobilenetV2/Conv/BatchNorm (--/64 params)\n",
            "        MobilenetV2/Conv/BatchNorm/beta (32, 32/32 params)\n",
            "        MobilenetV2/Conv/BatchNorm/gamma (32, 32/32 params)\n",
            "      MobilenetV2/Conv/weights (3x3x3x32, 864/864 params)\n",
            "    MobilenetV2/expanded_conv (--/896 params)\n",
            "      MobilenetV2/expanded_conv/depthwise (--/352 params)\n",
            "        MobilenetV2/expanded_conv/depthwise/BatchNorm (--/64 params)\n",
            "          MobilenetV2/expanded_conv/depthwise/BatchNorm/beta (32, 32/32 params)\n",
            "          MobilenetV2/expanded_conv/depthwise/BatchNorm/gamma (32, 32/32 params)\n",
            "        MobilenetV2/expanded_conv/depthwise/depthwise_weights (3x3x32x1, 288/288 params)\n",
            "      MobilenetV2/expanded_conv/project (--/544 params)\n",
            "        MobilenetV2/expanded_conv/project/BatchNorm (--/32 params)\n",
            "          MobilenetV2/expanded_conv/project/BatchNorm/beta (16, 16/16 params)\n",
            "          MobilenetV2/expanded_conv/project/BatchNorm/gamma (16, 16/16 params)\n",
            "        MobilenetV2/expanded_conv/project/weights (1x1x32x16, 512/512 params)\n",
            "    MobilenetV2/expanded_conv_1 (--/5.14k params)\n",
            "      MobilenetV2/expanded_conv_1/depthwise (--/1.06k params)\n",
            "        MobilenetV2/expanded_conv_1/depthwise/BatchNorm (--/192 params)\n",
            "          MobilenetV2/expanded_conv_1/depthwise/BatchNorm/beta (96, 96/96 params)\n",
            "          MobilenetV2/expanded_conv_1/depthwise/BatchNorm/gamma (96, 96/96 params)\n",
            "        MobilenetV2/expanded_conv_1/depthwise/depthwise_weights (3x3x96x1, 864/864 params)\n",
            "      MobilenetV2/expanded_conv_1/expand (--/1.73k params)\n",
            "        MobilenetV2/expanded_conv_1/expand/BatchNorm (--/192 params)\n",
            "          MobilenetV2/expanded_conv_1/expand/BatchNorm/beta (96, 96/96 params)\n",
            "          MobilenetV2/expanded_conv_1/expand/BatchNorm/gamma (96, 96/96 params)\n",
            "        MobilenetV2/expanded_conv_1/expand/weights (1x1x16x96, 1.54k/1.54k params)\n",
            "      MobilenetV2/expanded_conv_1/project (--/2.35k params)\n",
            "        MobilenetV2/expanded_conv_1/project/BatchNorm (--/48 params)\n",
            "          MobilenetV2/expanded_conv_1/project/BatchNorm/beta (24, 24/24 params)\n",
            "          MobilenetV2/expanded_conv_1/project/BatchNorm/gamma (24, 24/24 params)\n",
            "        MobilenetV2/expanded_conv_1/project/weights (1x1x96x24, 2.30k/2.30k params)\n",
            "    MobilenetV2/expanded_conv_10 (--/66.62k params)\n",
            "      MobilenetV2/expanded_conv_10/depthwise (--/4.22k params)\n",
            "        MobilenetV2/expanded_conv_10/depthwise/BatchNorm (--/768 params)\n",
            "          MobilenetV2/expanded_conv_10/depthwise/BatchNorm/beta (384, 384/384 params)\n",
            "          MobilenetV2/expanded_conv_10/depthwise/BatchNorm/gamma (384, 384/384 params)\n",
            "        MobilenetV2/expanded_conv_10/depthwise/depthwise_weights (3x3x384x1, 3.46k/3.46k params)\n",
            "      MobilenetV2/expanded_conv_10/expand (--/25.34k params)\n",
            "        MobilenetV2/expanded_conv_10/expand/BatchNorm (--/768 params)\n",
            "          MobilenetV2/expanded_conv_10/expand/BatchNorm/beta (384, 384/384 params)\n",
            "          MobilenetV2/expanded_conv_10/expand/BatchNorm/gamma (384, 384/384 params)\n",
            "        MobilenetV2/expanded_conv_10/expand/weights (1x1x64x384, 24.58k/24.58k params)\n",
            "      MobilenetV2/expanded_conv_10/project (--/37.06k params)\n",
            "        MobilenetV2/expanded_conv_10/project/BatchNorm (--/192 params)\n",
            "          MobilenetV2/expanded_conv_10/project/BatchNorm/beta (96, 96/96 params)\n",
            "          MobilenetV2/expanded_conv_10/project/BatchNorm/gamma (96, 96/96 params)\n",
            "        MobilenetV2/expanded_conv_10/project/weights (1x1x384x96, 36.86k/36.86k params)\n",
            "    MobilenetV2/expanded_conv_11 (--/118.27k params)\n",
            "      MobilenetV2/expanded_conv_11/depthwise (--/6.34k params)\n",
            "        MobilenetV2/expanded_conv_11/depthwise/BatchNorm (--/1.15k params)\n",
            "          MobilenetV2/expanded_conv_11/depthwise/BatchNorm/beta (576, 576/576 params)\n",
            "          MobilenetV2/expanded_conv_11/depthwise/BatchNorm/gamma (576, 576/576 params)\n",
            "        MobilenetV2/expanded_conv_11/depthwise/depthwise_weights (3x3x576x1, 5.18k/5.18k params)\n",
            "      MobilenetV2/expanded_conv_11/expand (--/56.45k params)\n",
            "        MobilenetV2/expanded_conv_11/expand/BatchNorm (--/1.15k params)\n",
            "          MobilenetV2/expanded_conv_11/expand/BatchNorm/beta (576, 576/576 params)\n",
            "          MobilenetV2/expanded_conv_11/expand/BatchNorm/gamma (576, 576/576 params)\n",
            "        MobilenetV2/expanded_conv_11/expand/weights (1x1x96x576, 55.30k/55.30k params)\n",
            "      MobilenetV2/expanded_conv_11/project (--/55.49k params)\n",
            "        MobilenetV2/expanded_conv_11/project/BatchNorm (--/192 params)\n",
            "          MobilenetV2/expanded_conv_11/project/BatchNorm/beta (96, 96/96 params)\n",
            "          MobilenetV2/expanded_conv_11/project/BatchNorm/gamma (96, 96/96 params)\n",
            "        MobilenetV2/expanded_conv_11/project/weights (1x1x576x96, 55.30k/55.30k params)\n",
            "    MobilenetV2/expanded_conv_12 (--/118.27k params)\n",
            "      MobilenetV2/expanded_conv_12/depthwise (--/6.34k params)\n",
            "        MobilenetV2/expanded_conv_12/depthwise/BatchNorm (--/1.15k params)\n",
            "          MobilenetV2/expanded_conv_12/depthwise/BatchNorm/beta (576, 576/576 params)\n",
            "          MobilenetV2/expanded_conv_12/depthwise/BatchNorm/gamma (576, 576/576 params)\n",
            "        MobilenetV2/expanded_conv_12/depthwise/depthwise_weights (3x3x576x1, 5.18k/5.18k params)\n",
            "      MobilenetV2/expanded_conv_12/expand (--/56.45k params)\n",
            "        MobilenetV2/expanded_conv_12/expand/BatchNorm (--/1.15k params)\n",
            "          MobilenetV2/expanded_conv_12/expand/BatchNorm/beta (576, 576/576 params)\n",
            "          MobilenetV2/expanded_conv_12/expand/BatchNorm/gamma (576, 576/576 params)\n",
            "        MobilenetV2/expanded_conv_12/expand/weights (1x1x96x576, 55.30k/55.30k params)\n",
            "      MobilenetV2/expanded_conv_12/project (--/55.49k params)\n",
            "        MobilenetV2/expanded_conv_12/project/BatchNorm (--/192 params)\n",
            "          MobilenetV2/expanded_conv_12/project/BatchNorm/beta (96, 96/96 params)\n",
            "          MobilenetV2/expanded_conv_12/project/BatchNorm/gamma (96, 96/96 params)\n",
            "        MobilenetV2/expanded_conv_12/project/weights (1x1x576x96, 55.30k/55.30k params)\n",
            "    MobilenetV2/expanded_conv_13 (--/155.26k params)\n",
            "      MobilenetV2/expanded_conv_13/depthwise (--/6.34k params)\n",
            "        MobilenetV2/expanded_conv_13/depthwise/BatchNorm (--/1.15k params)\n",
            "          MobilenetV2/expanded_conv_13/depthwise/BatchNorm/beta (576, 576/576 params)\n",
            "          MobilenetV2/expanded_conv_13/depthwise/BatchNorm/gamma (576, 576/576 params)\n",
            "        MobilenetV2/expanded_conv_13/depthwise/depthwise_weights (3x3x576x1, 5.18k/5.18k params)\n",
            "      MobilenetV2/expanded_conv_13/expand (--/56.45k params)\n",
            "        MobilenetV2/expanded_conv_13/expand/BatchNorm (--/1.15k params)\n",
            "          MobilenetV2/expanded_conv_13/expand/BatchNorm/beta (576, 576/576 params)\n",
            "          MobilenetV2/expanded_conv_13/expand/BatchNorm/gamma (576, 576/576 params)\n",
            "        MobilenetV2/expanded_conv_13/expand/weights (1x1x96x576, 55.30k/55.30k params)\n",
            "      MobilenetV2/expanded_conv_13/project (--/92.48k params)\n",
            "        MobilenetV2/expanded_conv_13/project/BatchNorm (--/320 params)\n",
            "          MobilenetV2/expanded_conv_13/project/BatchNorm/beta (160, 160/160 params)\n",
            "          MobilenetV2/expanded_conv_13/project/BatchNorm/gamma (160, 160/160 params)\n",
            "        MobilenetV2/expanded_conv_13/project/weights (1x1x576x160, 92.16k/92.16k params)\n",
            "    MobilenetV2/expanded_conv_14 (--/320.00k params)\n",
            "      MobilenetV2/expanded_conv_14/depthwise (--/10.56k params)\n",
            "        MobilenetV2/expanded_conv_14/depthwise/BatchNorm (--/1.92k params)\n",
            "          MobilenetV2/expanded_conv_14/depthwise/BatchNorm/beta (960, 960/960 params)\n",
            "          MobilenetV2/expanded_conv_14/depthwise/BatchNorm/gamma (960, 960/960 params)\n",
            "        MobilenetV2/expanded_conv_14/depthwise/depthwise_weights (3x3x960x1, 8.64k/8.64k params)\n",
            "      MobilenetV2/expanded_conv_14/expand (--/155.52k params)\n",
            "        MobilenetV2/expanded_conv_14/expand/BatchNorm (--/1.92k params)\n",
            "          MobilenetV2/expanded_conv_14/expand/BatchNorm/beta (960, 960/960 params)\n",
            "          MobilenetV2/expanded_conv_14/expand/BatchNorm/gamma (960, 960/960 params)\n",
            "        MobilenetV2/expanded_conv_14/expand/weights (1x1x160x960, 153.60k/153.60k params)\n",
            "      MobilenetV2/expanded_conv_14/project (--/153.92k params)\n",
            "        MobilenetV2/expanded_conv_14/project/BatchNorm (--/320 params)\n",
            "          MobilenetV2/expanded_conv_14/project/BatchNorm/beta (160, 160/160 params)\n",
            "          MobilenetV2/expanded_conv_14/project/BatchNorm/gamma (160, 160/160 params)\n",
            "        MobilenetV2/expanded_conv_14/project/weights (1x1x960x160, 153.60k/153.60k params)\n",
            "    MobilenetV2/expanded_conv_15 (--/320.00k params)\n",
            "      MobilenetV2/expanded_conv_15/depthwise (--/10.56k params)\n",
            "        MobilenetV2/expanded_conv_15/depthwise/BatchNorm (--/1.92k params)\n",
            "          MobilenetV2/expanded_conv_15/depthwise/BatchNorm/beta (960, 960/960 params)\n",
            "          MobilenetV2/expanded_conv_15/depthwise/BatchNorm/gamma (960, 960/960 params)\n",
            "        MobilenetV2/expanded_conv_15/depthwise/depthwise_weights (3x3x960x1, 8.64k/8.64k params)\n",
            "      MobilenetV2/expanded_conv_15/expand (--/155.52k params)\n",
            "        MobilenetV2/expanded_conv_15/expand/BatchNorm (--/1.92k params)\n",
            "          MobilenetV2/expanded_conv_15/expand/BatchNorm/beta (960, 960/960 params)\n",
            "          MobilenetV2/expanded_conv_15/expand/BatchNorm/gamma (960, 960/960 params)\n",
            "        MobilenetV2/expanded_conv_15/expand/weights (1x1x160x960, 153.60k/153.60k params)\n",
            "      MobilenetV2/expanded_conv_15/project (--/153.92k params)\n",
            "        MobilenetV2/expanded_conv_15/project/BatchNorm (--/320 params)\n",
            "          MobilenetV2/expanded_conv_15/project/BatchNorm/beta (160, 160/160 params)\n",
            "          MobilenetV2/expanded_conv_15/project/BatchNorm/gamma (160, 160/160 params)\n",
            "        MobilenetV2/expanded_conv_15/project/weights (1x1x960x160, 153.60k/153.60k params)\n",
            "    MobilenetV2/expanded_conv_16 (--/473.92k params)\n",
            "      MobilenetV2/expanded_conv_16/depthwise (--/10.56k params)\n",
            "        MobilenetV2/expanded_conv_16/depthwise/BatchNorm (--/1.92k params)\n",
            "          MobilenetV2/expanded_conv_16/depthwise/BatchNorm/beta (960, 960/960 params)\n",
            "          MobilenetV2/expanded_conv_16/depthwise/BatchNorm/gamma (960, 960/960 params)\n",
            "        MobilenetV2/expanded_conv_16/depthwise/depthwise_weights (3x3x960x1, 8.64k/8.64k params)\n",
            "      MobilenetV2/expanded_conv_16/expand (--/155.52k params)\n",
            "        MobilenetV2/expanded_conv_16/expand/BatchNorm (--/1.92k params)\n",
            "          MobilenetV2/expanded_conv_16/expand/BatchNorm/beta (960, 960/960 params)\n",
            "          MobilenetV2/expanded_conv_16/expand/BatchNorm/gamma (960, 960/960 params)\n",
            "        MobilenetV2/expanded_conv_16/expand/weights (1x1x160x960, 153.60k/153.60k params)\n",
            "      MobilenetV2/expanded_conv_16/project (--/307.84k params)\n",
            "        MobilenetV2/expanded_conv_16/project/BatchNorm (--/640 params)\n",
            "          MobilenetV2/expanded_conv_16/project/BatchNorm/beta (320, 320/320 params)\n",
            "          MobilenetV2/expanded_conv_16/project/BatchNorm/gamma (320, 320/320 params)\n",
            "        MobilenetV2/expanded_conv_16/project/weights (1x1x960x320, 307.20k/307.20k params)\n",
            "    MobilenetV2/expanded_conv_2 (--/8.83k params)\n",
            "      MobilenetV2/expanded_conv_2/depthwise (--/1.58k params)\n",
            "        MobilenetV2/expanded_conv_2/depthwise/BatchNorm (--/288 params)\n",
            "          MobilenetV2/expanded_conv_2/depthwise/BatchNorm/beta (144, 144/144 params)\n",
            "          MobilenetV2/expanded_conv_2/depthwise/BatchNorm/gamma (144, 144/144 params)\n",
            "        MobilenetV2/expanded_conv_2/depthwise/depthwise_weights (3x3x144x1, 1.30k/1.30k params)\n",
            "      MobilenetV2/expanded_conv_2/expand (--/3.74k params)\n",
            "        MobilenetV2/expanded_conv_2/expand/BatchNorm (--/288 params)\n",
            "          MobilenetV2/expanded_conv_2/expand/BatchNorm/beta (144, 144/144 params)\n",
            "          MobilenetV2/expanded_conv_2/expand/BatchNorm/gamma (144, 144/144 params)\n",
            "        MobilenetV2/expanded_conv_2/expand/weights (1x1x24x144, 3.46k/3.46k params)\n",
            "      MobilenetV2/expanded_conv_2/project (--/3.50k params)\n",
            "        MobilenetV2/expanded_conv_2/project/BatchNorm (--/48 params)\n",
            "          MobilenetV2/expanded_conv_2/project/BatchNorm/beta (24, 24/24 params)\n",
            "          MobilenetV2/expanded_conv_2/project/BatchNorm/gamma (24, 24/24 params)\n",
            "        MobilenetV2/expanded_conv_2/project/weights (1x1x144x24, 3.46k/3.46k params)\n",
            "    MobilenetV2/expanded_conv_3 (--/10.00k params)\n",
            "      MobilenetV2/expanded_conv_3/depthwise (--/1.58k params)\n",
            "        MobilenetV2/expanded_conv_3/depthwise/BatchNorm (--/288 params)\n",
            "          MobilenetV2/expanded_conv_3/depthwise/BatchNorm/beta (144, 144/144 params)\n",
            "          MobilenetV2/expanded_conv_3/depthwise/BatchNorm/gamma (144, 144/144 params)\n",
            "        MobilenetV2/expanded_conv_3/depthwise/depthwise_weights (3x3x144x1, 1.30k/1.30k params)\n",
            "      MobilenetV2/expanded_conv_3/expand (--/3.74k params)\n",
            "        MobilenetV2/expanded_conv_3/expand/BatchNorm (--/288 params)\n",
            "          MobilenetV2/expanded_conv_3/expand/BatchNorm/beta (144, 144/144 params)\n",
            "          MobilenetV2/expanded_conv_3/expand/BatchNorm/gamma (144, 144/144 params)\n",
            "        MobilenetV2/expanded_conv_3/expand/weights (1x1x24x144, 3.46k/3.46k params)\n",
            "      MobilenetV2/expanded_conv_3/project (--/4.67k params)\n",
            "        MobilenetV2/expanded_conv_3/project/BatchNorm (--/64 params)\n",
            "          MobilenetV2/expanded_conv_3/project/BatchNorm/beta (32, 32/32 params)\n",
            "          MobilenetV2/expanded_conv_3/project/BatchNorm/gamma (32, 32/32 params)\n",
            "        MobilenetV2/expanded_conv_3/project/weights (1x1x144x32, 4.61k/4.61k params)\n",
            "    MobilenetV2/expanded_conv_4 (--/14.85k params)\n",
            "      MobilenetV2/expanded_conv_4/depthwise (--/2.11k params)\n",
            "        MobilenetV2/expanded_conv_4/depthwise/BatchNorm (--/384 params)\n",
            "          MobilenetV2/expanded_conv_4/depthwise/BatchNorm/beta (192, 192/192 params)\n",
            "          MobilenetV2/expanded_conv_4/depthwise/BatchNorm/gamma (192, 192/192 params)\n",
            "        MobilenetV2/expanded_conv_4/depthwise/depthwise_weights (3x3x192x1, 1.73k/1.73k params)\n",
            "      MobilenetV2/expanded_conv_4/expand (--/6.53k params)\n",
            "        MobilenetV2/expanded_conv_4/expand/BatchNorm (--/384 params)\n",
            "          MobilenetV2/expanded_conv_4/expand/BatchNorm/beta (192, 192/192 params)\n",
            "          MobilenetV2/expanded_conv_4/expand/BatchNorm/gamma (192, 192/192 params)\n",
            "        MobilenetV2/expanded_conv_4/expand/weights (1x1x32x192, 6.14k/6.14k params)\n",
            "      MobilenetV2/expanded_conv_4/project (--/6.21k params)\n",
            "        MobilenetV2/expanded_conv_4/project/BatchNorm (--/64 params)\n",
            "          MobilenetV2/expanded_conv_4/project/BatchNorm/beta (32, 32/32 params)\n",
            "          MobilenetV2/expanded_conv_4/project/BatchNorm/gamma (32, 32/32 params)\n",
            "        MobilenetV2/expanded_conv_4/project/weights (1x1x192x32, 6.14k/6.14k params)\n",
            "    MobilenetV2/expanded_conv_5 (--/14.85k params)\n",
            "      MobilenetV2/expanded_conv_5/depthwise (--/2.11k params)\n",
            "        MobilenetV2/expanded_conv_5/depthwise/BatchNorm (--/384 params)\n",
            "          MobilenetV2/expanded_conv_5/depthwise/BatchNorm/beta (192, 192/192 params)\n",
            "          MobilenetV2/expanded_conv_5/depthwise/BatchNorm/gamma (192, 192/192 params)\n",
            "        MobilenetV2/expanded_conv_5/depthwise/depthwise_weights (3x3x192x1, 1.73k/1.73k params)\n",
            "      MobilenetV2/expanded_conv_5/expand (--/6.53k params)\n",
            "        MobilenetV2/expanded_conv_5/expand/BatchNorm (--/384 params)\n",
            "          MobilenetV2/expanded_conv_5/expand/BatchNorm/beta (192, 192/192 params)\n",
            "          MobilenetV2/expanded_conv_5/expand/BatchNorm/gamma (192, 192/192 params)\n",
            "        MobilenetV2/expanded_conv_5/expand/weights (1x1x32x192, 6.14k/6.14k params)\n",
            "      MobilenetV2/expanded_conv_5/project (--/6.21k params)\n",
            "        MobilenetV2/expanded_conv_5/project/BatchNorm (--/64 params)\n",
            "          MobilenetV2/expanded_conv_5/project/BatchNorm/beta (32, 32/32 params)\n",
            "          MobilenetV2/expanded_conv_5/project/BatchNorm/gamma (32, 32/32 params)\n",
            "        MobilenetV2/expanded_conv_5/project/weights (1x1x192x32, 6.14k/6.14k params)\n",
            "    MobilenetV2/expanded_conv_6 (--/21.06k params)\n",
            "      MobilenetV2/expanded_conv_6/depthwise (--/2.11k params)\n",
            "        MobilenetV2/expanded_conv_6/depthwise/BatchNorm (--/384 params)\n",
            "          MobilenetV2/expanded_conv_6/depthwise/BatchNorm/beta (192, 192/192 params)\n",
            "          MobilenetV2/expanded_conv_6/depthwise/BatchNorm/gamma (192, 192/192 params)\n",
            "        MobilenetV2/expanded_conv_6/depthwise/depthwise_weights (3x3x192x1, 1.73k/1.73k params)\n",
            "      MobilenetV2/expanded_conv_6/expand (--/6.53k params)\n",
            "        MobilenetV2/expanded_conv_6/expand/BatchNorm (--/384 params)\n",
            "          MobilenetV2/expanded_conv_6/expand/BatchNorm/beta (192, 192/192 params)\n",
            "          MobilenetV2/expanded_conv_6/expand/BatchNorm/gamma (192, 192/192 params)\n",
            "        MobilenetV2/expanded_conv_6/expand/weights (1x1x32x192, 6.14k/6.14k params)\n",
            "      MobilenetV2/expanded_conv_6/project (--/12.42k params)\n",
            "        MobilenetV2/expanded_conv_6/project/BatchNorm (--/128 params)\n",
            "          MobilenetV2/expanded_conv_6/project/BatchNorm/beta (64, 64/64 params)\n",
            "          MobilenetV2/expanded_conv_6/project/BatchNorm/gamma (64, 64/64 params)\n",
            "        MobilenetV2/expanded_conv_6/project/weights (1x1x192x64, 12.29k/12.29k params)\n",
            "    MobilenetV2/expanded_conv_7 (--/54.27k params)\n",
            "      MobilenetV2/expanded_conv_7/depthwise (--/4.22k params)\n",
            "        MobilenetV2/expanded_conv_7/depthwise/BatchNorm (--/768 params)\n",
            "          MobilenetV2/expanded_conv_7/depthwise/BatchNorm/beta (384, 384/384 params)\n",
            "          MobilenetV2/expanded_conv_7/depthwise/BatchNorm/gamma (384, 384/384 params)\n",
            "        MobilenetV2/expanded_conv_7/depthwise/depthwise_weights (3x3x384x1, 3.46k/3.46k params)\n",
            "      MobilenetV2/expanded_conv_7/expand (--/25.34k params)\n",
            "        MobilenetV2/expanded_conv_7/expand/BatchNorm (--/768 params)\n",
            "          MobilenetV2/expanded_conv_7/expand/BatchNorm/beta (384, 384/384 params)\n",
            "          MobilenetV2/expanded_conv_7/expand/BatchNorm/gamma (384, 384/384 params)\n",
            "        MobilenetV2/expanded_conv_7/expand/weights (1x1x64x384, 24.58k/24.58k params)\n",
            "      MobilenetV2/expanded_conv_7/project (--/24.70k params)\n",
            "        MobilenetV2/expanded_conv_7/project/BatchNorm (--/128 params)\n",
            "          MobilenetV2/expanded_conv_7/project/BatchNorm/beta (64, 64/64 params)\n",
            "          MobilenetV2/expanded_conv_7/project/BatchNorm/gamma (64, 64/64 params)\n",
            "        MobilenetV2/expanded_conv_7/project/weights (1x1x384x64, 24.58k/24.58k params)\n",
            "    MobilenetV2/expanded_conv_8 (--/54.27k params)\n",
            "      MobilenetV2/expanded_conv_8/depthwise (--/4.22k params)\n",
            "        MobilenetV2/expanded_conv_8/depthwise/BatchNorm (--/768 params)\n",
            "          MobilenetV2/expanded_conv_8/depthwise/BatchNorm/beta (384, 384/384 params)\n",
            "          MobilenetV2/expanded_conv_8/depthwise/BatchNorm/gamma (384, 384/384 params)\n",
            "        MobilenetV2/expanded_conv_8/depthwise/depthwise_weights (3x3x384x1, 3.46k/3.46k params)\n",
            "      MobilenetV2/expanded_conv_8/expand (--/25.34k params)\n",
            "        MobilenetV2/expanded_conv_8/expand/BatchNorm (--/768 params)\n",
            "          MobilenetV2/expanded_conv_8/expand/BatchNorm/beta (384, 384/384 params)\n",
            "          MobilenetV2/expanded_conv_8/expand/BatchNorm/gamma (384, 384/384 params)\n",
            "        MobilenetV2/expanded_conv_8/expand/weights (1x1x64x384, 24.58k/24.58k params)\n",
            "      MobilenetV2/expanded_conv_8/project (--/24.70k params)\n",
            "        MobilenetV2/expanded_conv_8/project/BatchNorm (--/128 params)\n",
            "          MobilenetV2/expanded_conv_8/project/BatchNorm/beta (64, 64/64 params)\n",
            "          MobilenetV2/expanded_conv_8/project/BatchNorm/gamma (64, 64/64 params)\n",
            "        MobilenetV2/expanded_conv_8/project/weights (1x1x384x64, 24.58k/24.58k params)\n",
            "    MobilenetV2/expanded_conv_9 (--/54.27k params)\n",
            "      MobilenetV2/expanded_conv_9/depthwise (--/4.22k params)\n",
            "        MobilenetV2/expanded_conv_9/depthwise/BatchNorm (--/768 params)\n",
            "          MobilenetV2/expanded_conv_9/depthwise/BatchNorm/beta (384, 384/384 params)\n",
            "          MobilenetV2/expanded_conv_9/depthwise/BatchNorm/gamma (384, 384/384 params)\n",
            "        MobilenetV2/expanded_conv_9/depthwise/depthwise_weights (3x3x384x1, 3.46k/3.46k params)\n",
            "      MobilenetV2/expanded_conv_9/expand (--/25.34k params)\n",
            "        MobilenetV2/expanded_conv_9/expand/BatchNorm (--/768 params)\n",
            "          MobilenetV2/expanded_conv_9/expand/BatchNorm/beta (384, 384/384 params)\n",
            "          MobilenetV2/expanded_conv_9/expand/BatchNorm/gamma (384, 384/384 params)\n",
            "        MobilenetV2/expanded_conv_9/expand/weights (1x1x64x384, 24.58k/24.58k params)\n",
            "      MobilenetV2/expanded_conv_9/project (--/24.70k params)\n",
            "        MobilenetV2/expanded_conv_9/project/BatchNorm (--/128 params)\n",
            "          MobilenetV2/expanded_conv_9/project/BatchNorm/beta (64, 64/64 params)\n",
            "          MobilenetV2/expanded_conv_9/project/BatchNorm/gamma (64, 64/64 params)\n",
            "        MobilenetV2/expanded_conv_9/project/weights (1x1x384x64, 24.58k/24.58k params)\n",
            "  aspp0 (--/82.43k params)\n",
            "    aspp0/BatchNorm (--/512 params)\n",
            "      aspp0/BatchNorm/beta (256, 256/256 params)\n",
            "      aspp0/BatchNorm/gamma (256, 256/256 params)\n",
            "    aspp0/weights (1x1x320x256, 81.92k/81.92k params)\n",
            "  concat_projection (--/131.58k params)\n",
            "    concat_projection/BatchNorm (--/512 params)\n",
            "      concat_projection/BatchNorm/beta (256, 256/256 params)\n",
            "      concat_projection/BatchNorm/gamma (256, 256/256 params)\n",
            "    concat_projection/weights (1x1x512x256, 131.07k/131.07k params)\n",
            "  image_pooling (--/82.43k params)\n",
            "    image_pooling/BatchNorm (--/512 params)\n",
            "      image_pooling/BatchNorm/beta (256, 256/256 params)\n",
            "      image_pooling/BatchNorm/gamma (256, 256/256 params)\n",
            "    image_pooling/weights (1x1x320x256, 81.92k/81.92k params)\n",
            "  logits (--/514 params)\n",
            "    logits/semantic (--/514 params)\n",
            "      logits/semantic/biases (2, 2/2 params)\n",
            "      logits/semantic/weights (1x1x256x2, 512/512 params)\n",
            "\n",
            "======================End of Report==========================\n",
            "4 ops no flops stats due to incomplete shapes.\n",
            "Parsing Inputs...\n",
            "\n",
            "=========================Options=============================\n",
            "-max_depth                  10000\n",
            "-min_bytes                  0\n",
            "-min_peak_bytes             0\n",
            "-min_residual_bytes         0\n",
            "-min_output_bytes           0\n",
            "-min_micros                 0\n",
            "-min_accelerator_micros     0\n",
            "-min_cpu_micros             0\n",
            "-min_params                 0\n",
            "-min_float_ops              1\n",
            "-min_occurrence             0\n",
            "-step                       -1\n",
            "-order_by                   float_ops\n",
            "-account_type_regexes       .*\n",
            "-start_name_regexes         .*\n",
            "-trim_name_regexes          \n",
            "-show_name_regexes          .*\n",
            "-hide_name_regexes          \n",
            "-account_displayed_op_only  true\n",
            "-select                     float_ops\n",
            "-output                     stdout:\n",
            "\n",
            "==================Model Analysis Report======================\n",
            "\n",
            "Doc:\n",
            "scope: The nodes in the model graph are organized by their names, which is hierarchical like filesystem.\n",
            "flops: Number of float operations. Note: Please read the implementation for the math behind it.\n",
            "\n",
            "Profile:\n",
            "node name | # float_ops\n",
            "_TFProfRoot (--/5.49b flops)\n",
            "  MobilenetV2/expanded_conv_16/project/Conv2D (669.08m/669.08m flops)\n",
            "  MobilenetV2/expanded_conv_15/expand/Conv2D (334.54m/334.54m flops)\n",
            "  MobilenetV2/expanded_conv_14/expand/Conv2D (334.54m/334.54m flops)\n",
            "  MobilenetV2/expanded_conv_16/expand/Conv2D (334.54m/334.54m flops)\n",
            "  MobilenetV2/expanded_conv_14/project/Conv2D (334.54m/334.54m flops)\n",
            "  MobilenetV2/expanded_conv_15/project/Conv2D (334.54m/334.54m flops)\n",
            "  concat_projection/Conv2D (285.47m/285.47m flops)\n",
            "  MobilenetV2/expanded_conv_1/expand/Conv2D (202.90m/202.90m flops)\n",
            "  MobilenetV2/expanded_conv_13/project/Conv2D (200.72m/200.72m flops)\n",
            "  aspp0/Conv2D (178.42m/178.42m flops)\n",
            "  MobilenetV2/expanded_conv_12/project/Conv2D (120.43m/120.43m flops)\n",
            "  MobilenetV2/expanded_conv_12/expand/Conv2D (120.43m/120.43m flops)\n",
            "  MobilenetV2/expanded_conv_13/expand/Conv2D (120.43m/120.43m flops)\n",
            "  MobilenetV2/expanded_conv_11/project/Conv2D (120.43m/120.43m flops)\n",
            "  MobilenetV2/expanded_conv_11/expand/Conv2D (120.43m/120.43m flops)\n",
            "  MobilenetV2/expanded_conv_3/expand/Conv2D (115.02m/115.02m flops)\n",
            "  MobilenetV2/expanded_conv_2/project/Conv2D (115.02m/115.02m flops)\n",
            "  MobilenetV2/expanded_conv_2/expand/Conv2D (115.02m/115.02m flops)\n",
            "  MobilenetV2/Conv/Conv2D (114.13m/114.13m flops)\n",
            "  MobilenetV2/expanded_conv_10/project/Conv2D (80.29m/80.29m flops)\n",
            "  MobilenetV2/expanded_conv_1/project/Conv2D (76.68m/76.68m flops)\n",
            "  MobilenetV2/expanded_conv/project/Conv2D (67.63m/67.63m flops)\n",
            "  MobilenetV2/expanded_conv_7/project/Conv2D (53.53m/53.53m flops)\n",
            "  MobilenetV2/expanded_conv_7/expand/Conv2D (53.53m/53.53m flops)\n",
            "  MobilenetV2/expanded_conv_9/project/Conv2D (53.53m/53.53m flops)\n",
            "  MobilenetV2/expanded_conv_9/expand/Conv2D (53.53m/53.53m flops)\n",
            "  MobilenetV2/expanded_conv_8/expand/Conv2D (53.53m/53.53m flops)\n",
            "  MobilenetV2/expanded_conv_8/project/Conv2D (53.53m/53.53m flops)\n",
            "  MobilenetV2/expanded_conv_10/expand/Conv2D (53.53m/53.53m flops)\n",
            "  MobilenetV2/expanded_conv_5/expand/Conv2D (51.92m/51.92m flops)\n",
            "  MobilenetV2/expanded_conv_4/expand/Conv2D (51.92m/51.92m flops)\n",
            "  MobilenetV2/expanded_conv_4/project/Conv2D (51.92m/51.92m flops)\n",
            "  MobilenetV2/expanded_conv_5/project/Conv2D (51.92m/51.92m flops)\n",
            "  MobilenetV2/expanded_conv_6/expand/Conv2D (51.92m/51.92m flops)\n",
            "  MobilenetV2/expanded_conv_2/depthwise/depthwise (43.13m/43.13m flops)\n",
            "  MobilenetV2/expanded_conv_3/project/Conv2D (38.94m/38.94m flops)\n",
            "  MobilenetV2/expanded_conv/depthwise/depthwise (38.04m/38.04m flops)\n",
            "  MobilenetV2/expanded_conv_1/depthwise/depthwise (28.76m/28.76m flops)\n",
            "  MobilenetV2/expanded_conv_6/project/Conv2D (26.76m/26.76m flops)\n",
            "  MobilenetV2/expanded_conv_16/depthwise/depthwise (19.98m/19.98m flops)\n",
            "  MobilenetV2/expanded_conv_14/depthwise/depthwise (19.98m/19.98m flops)\n",
            "  MobilenetV2/expanded_conv_15/depthwise/depthwise (19.98m/19.98m flops)\n",
            "  MobilenetV2/expanded_conv_4/depthwise/depthwise (14.60m/14.60m flops)\n",
            "  MobilenetV2/expanded_conv_5/depthwise/depthwise (14.60m/14.60m flops)\n",
            "  MobilenetV2/expanded_conv_13/depthwise/depthwise (11.29m/11.29m flops)\n",
            "  MobilenetV2/expanded_conv_12/depthwise/depthwise (11.29m/11.29m flops)\n",
            "  MobilenetV2/expanded_conv_11/depthwise/depthwise (11.29m/11.29m flops)\n",
            "  MobilenetV2/expanded_conv_3/depthwise/depthwise (10.95m/10.95m flops)\n",
            "  MobilenetV2/expanded_conv_7/depthwise/depthwise (7.53m/7.53m flops)\n",
            "  MobilenetV2/expanded_conv_8/depthwise/depthwise (7.53m/7.53m flops)\n",
            "  MobilenetV2/expanded_conv_10/depthwise/depthwise (7.53m/7.53m flops)\n",
            "  MobilenetV2/expanded_conv_9/depthwise/depthwise (7.53m/7.53m flops)\n",
            "  MobilenetV2/expanded_conv_6/depthwise/depthwise (3.76m/3.76m flops)\n",
            "  logits/semantic/Conv2D (1.12m/1.12m flops)\n",
            "  MobilenetV2/expanded_conv_16/project/kernel/Regularizer/l2_regularizer (1/921.60k flops)\n",
            "    MobilenetV2/expanded_conv_16/project/kernel/Regularizer/l2_regularizer/L2Loss (921.60k/921.60k flops)\n",
            "  sub (789.51k/789.51k flops)\n",
            "  mul (789.51k/789.51k flops)\n",
            "  MobilenetV2/expanded_conv_16/project/weights/Initializer/truncated_normal (307.20k/614.40k flops)\n",
            "    MobilenetV2/expanded_conv_16/project/weights/Initializer/truncated_normal/mul (307.20k/307.20k flops)\n",
            "  MobilenetV2/expanded_conv_14/expand/kernel/Regularizer/l2_regularizer (1/460.80k flops)\n",
            "    MobilenetV2/expanded_conv_14/expand/kernel/Regularizer/l2_regularizer/L2Loss (460.80k/460.80k flops)\n",
            "  MobilenetV2/expanded_conv_16/expand/kernel/Regularizer/l2_regularizer (1/460.80k flops)\n",
            "    MobilenetV2/expanded_conv_16/expand/kernel/Regularizer/l2_regularizer/L2Loss (460.80k/460.80k flops)\n",
            "  MobilenetV2/expanded_conv_14/project/kernel/Regularizer/l2_regularizer (1/460.80k flops)\n",
            "    MobilenetV2/expanded_conv_14/project/kernel/Regularizer/l2_regularizer/L2Loss (460.80k/460.80k flops)\n",
            "  MobilenetV2/expanded_conv_15/project/kernel/Regularizer/l2_regularizer (1/460.80k flops)\n",
            "    MobilenetV2/expanded_conv_15/project/kernel/Regularizer/l2_regularizer/L2Loss (460.80k/460.80k flops)\n",
            "  MobilenetV2/expanded_conv_15/expand/kernel/Regularizer/l2_regularizer (1/460.80k flops)\n",
            "    MobilenetV2/expanded_conv_15/expand/kernel/Regularizer/l2_regularizer/L2Loss (460.80k/460.80k flops)\n",
            "  MobilenetV2/expanded_conv_2/add (399.38k/399.38k flops)\n",
            "  concat_projection/kernel/Regularizer/l2_regularizer (1/393.22k flops)\n",
            "    concat_projection/kernel/Regularizer/l2_regularizer/L2Loss (393.21k/393.21k flops)\n",
            "  AvgPool2D/AvgPool (348.48k/348.48k flops)\n",
            "  MobilenetV2/expanded_conv_15/project/weights/Initializer/truncated_normal (153.60k/307.20k flops)\n",
            "    MobilenetV2/expanded_conv_15/project/weights/Initializer/truncated_normal/mul (153.60k/153.60k flops)\n",
            "  MobilenetV2/expanded_conv_14/project/weights/Initializer/truncated_normal (153.60k/307.20k flops)\n",
            "    MobilenetV2/expanded_conv_14/project/weights/Initializer/truncated_normal/mul (153.60k/153.60k flops)\n",
            "  MobilenetV2/expanded_conv_14/expand/weights/Initializer/truncated_normal (153.60k/307.20k flops)\n",
            "    MobilenetV2/expanded_conv_14/expand/weights/Initializer/truncated_normal/mul (153.60k/153.60k flops)\n",
            "  MobilenetV2/expanded_conv_16/expand/weights/Initializer/truncated_normal (153.60k/307.20k flops)\n",
            "    MobilenetV2/expanded_conv_16/expand/weights/Initializer/truncated_normal/mul (153.60k/153.60k flops)\n",
            "  MobilenetV2/expanded_conv_15/expand/weights/Initializer/truncated_normal (153.60k/307.20k flops)\n",
            "    MobilenetV2/expanded_conv_15/expand/weights/Initializer/truncated_normal/mul (153.60k/153.60k flops)\n",
            "  MobilenetV2/expanded_conv_13/project/kernel/Regularizer/l2_regularizer (1/276.48k flops)\n",
            "    MobilenetV2/expanded_conv_13/project/kernel/Regularizer/l2_regularizer/L2Loss (276.48k/276.48k flops)\n",
            "  ArgMax (263.17k/263.17k flops)\n",
            "  mean_iou/confusion_matrix/assert_non_negative/assert_less_equal/LessEqual (263.17k/263.17k flops)\n",
            "  mean_iou/confusion_matrix/assert_less/Less (263.17k/263.17k flops)\n",
            "  concat_projection/weights/Initializer/random_uniform (131.07k/262.14k flops)\n",
            "    concat_projection/weights/Initializer/random_uniform/mul (131.07k/131.07k flops)\n",
            "    concat_projection/weights/Initializer/random_uniform/sub (1/1 flops)\n",
            "  aspp0/kernel/Regularizer/l2_regularizer (1/245.76k flops)\n",
            "    aspp0/kernel/Regularizer/l2_regularizer/L2Loss (245.76k/245.76k flops)\n",
            "  image_pooling/kernel/Regularizer/l2_regularizer (1/245.76k flops)\n",
            "    image_pooling/kernel/Regularizer/l2_regularizer/L2Loss (245.76k/245.76k flops)\n",
            "  MobilenetV2/expanded_conv_13/project/weights/Initializer/truncated_normal (92.16k/184.32k flops)\n",
            "    MobilenetV2/expanded_conv_13/project/weights/Initializer/truncated_normal/mul (92.16k/92.16k flops)\n",
            "  MobilenetV2/expanded_conv_14/add (174.24k/174.24k flops)\n",
            "  MobilenetV2/expanded_conv_15/add (174.24k/174.24k flops)\n",
            "  MobilenetV2/expanded_conv_11/project/kernel/Regularizer/l2_regularizer (1/165.89k flops)\n",
            "    MobilenetV2/expanded_conv_11/project/kernel/Regularizer/l2_regularizer/L2Loss (165.89k/165.89k flops)\n",
            "  MobilenetV2/expanded_conv_12/expand/kernel/Regularizer/l2_regularizer (1/165.89k flops)\n",
            "    MobilenetV2/expanded_conv_12/expand/kernel/Regularizer/l2_regularizer/L2Loss (165.89k/165.89k flops)\n",
            "  MobilenetV2/expanded_conv_12/project/kernel/Regularizer/l2_regularizer (1/165.89k flops)\n",
            "    MobilenetV2/expanded_conv_12/project/kernel/Regularizer/l2_regularizer/L2Loss (165.89k/165.89k flops)\n",
            "  MobilenetV2/expanded_conv_11/expand/kernel/Regularizer/l2_regularizer (1/165.89k flops)\n",
            "    MobilenetV2/expanded_conv_11/expand/kernel/Regularizer/l2_regularizer/L2Loss (165.89k/165.89k flops)\n",
            "  MobilenetV2/expanded_conv_13/expand/kernel/Regularizer/l2_regularizer (1/165.89k flops)\n",
            "    MobilenetV2/expanded_conv_13/expand/kernel/Regularizer/l2_regularizer/L2Loss (165.89k/165.89k flops)\n",
            "  image_pooling/weights/Initializer/random_uniform (81.92k/163.84k flops)\n",
            "    image_pooling/weights/Initializer/random_uniform/mul (81.92k/81.92k flops)\n",
            "    image_pooling/weights/Initializer/random_uniform/sub (1/1 flops)\n",
            "  aspp0/weights/Initializer/random_uniform (81.92k/163.84k flops)\n",
            "    aspp0/weights/Initializer/random_uniform/mul (81.92k/81.92k flops)\n",
            "    aspp0/weights/Initializer/random_uniform/sub (1/1 flops)\n",
            "  image_pooling/Conv2D (163.84k/163.84k flops)\n",
            "  MobilenetV2/expanded_conv_5/add (135.20k/135.20k flops)\n",
            "  MobilenetV2/expanded_conv_4/add (135.20k/135.20k flops)\n",
            "  MobilenetV2/expanded_conv_12/project/weights/Initializer/truncated_normal (55.30k/110.59k flops)\n",
            "    MobilenetV2/expanded_conv_12/project/weights/Initializer/truncated_normal/mul (55.30k/55.30k flops)\n",
            "  MobilenetV2/expanded_conv_12/expand/weights/Initializer/truncated_normal (55.30k/110.59k flops)\n",
            "    MobilenetV2/expanded_conv_12/expand/weights/Initializer/truncated_normal/mul (55.30k/55.30k flops)\n",
            "  MobilenetV2/expanded_conv_11/expand/weights/Initializer/truncated_normal (55.30k/110.59k flops)\n",
            "    MobilenetV2/expanded_conv_11/expand/weights/Initializer/truncated_normal/mul (55.30k/55.30k flops)\n",
            "  MobilenetV2/expanded_conv_13/expand/weights/Initializer/truncated_normal (55.30k/110.59k flops)\n",
            "    MobilenetV2/expanded_conv_13/expand/weights/Initializer/truncated_normal/mul (55.30k/55.30k flops)\n",
            "  MobilenetV2/expanded_conv_10/project/kernel/Regularizer/l2_regularizer (1/110.59k flops)\n",
            "    MobilenetV2/expanded_conv_10/project/kernel/Regularizer/l2_regularizer/L2Loss (110.59k/110.59k flops)\n",
            "  MobilenetV2/expanded_conv_11/project/weights/Initializer/truncated_normal (55.30k/110.59k flops)\n",
            "    MobilenetV2/expanded_conv_11/project/weights/Initializer/truncated_normal/mul (55.30k/55.30k flops)\n",
            "  MobilenetV2/expanded_conv_11/add (104.54k/104.54k flops)\n",
            "  MobilenetV2/expanded_conv_12/add (104.54k/104.54k flops)\n",
            "  MobilenetV2/expanded_conv_7/project/kernel/Regularizer/l2_regularizer (1/73.73k flops)\n",
            "    MobilenetV2/expanded_conv_7/project/kernel/Regularizer/l2_regularizer/L2Loss (73.73k/73.73k flops)\n",
            "  MobilenetV2/expanded_conv_10/project/weights/Initializer/truncated_normal (36.86k/73.73k flops)\n",
            "    MobilenetV2/expanded_conv_10/project/weights/Initializer/truncated_normal/mul (36.86k/36.86k flops)\n",
            "  MobilenetV2/expanded_conv_7/expand/kernel/Regularizer/l2_regularizer (1/73.73k flops)\n",
            "    MobilenetV2/expanded_conv_7/expand/kernel/Regularizer/l2_regularizer/L2Loss (73.73k/73.73k flops)\n",
            "  MobilenetV2/expanded_conv_10/expand/kernel/Regularizer/l2_regularizer (1/73.73k flops)\n",
            "    MobilenetV2/expanded_conv_10/expand/kernel/Regularizer/l2_regularizer/L2Loss (73.73k/73.73k flops)\n",
            "  MobilenetV2/expanded_conv_8/expand/kernel/Regularizer/l2_regularizer (1/73.73k flops)\n",
            "    MobilenetV2/expanded_conv_8/expand/kernel/Regularizer/l2_regularizer/L2Loss (73.73k/73.73k flops)\n",
            "  MobilenetV2/expanded_conv_8/project/kernel/Regularizer/l2_regularizer (1/73.73k flops)\n",
            "    MobilenetV2/expanded_conv_8/project/kernel/Regularizer/l2_regularizer/L2Loss (73.73k/73.73k flops)\n",
            "  MobilenetV2/expanded_conv_9/expand/kernel/Regularizer/l2_regularizer (1/73.73k flops)\n",
            "    MobilenetV2/expanded_conv_9/expand/kernel/Regularizer/l2_regularizer/L2Loss (73.73k/73.73k flops)\n",
            "  MobilenetV2/expanded_conv_9/project/kernel/Regularizer/l2_regularizer (1/73.73k flops)\n",
            "    MobilenetV2/expanded_conv_9/project/kernel/Regularizer/l2_regularizer/L2Loss (73.73k/73.73k flops)\n",
            "  MobilenetV2/expanded_conv_8/add (69.70k/69.70k flops)\n",
            "  MobilenetV2/expanded_conv_9/add (69.70k/69.70k flops)\n",
            "  MobilenetV2/expanded_conv_7/add (69.70k/69.70k flops)\n",
            "  MobilenetV2/expanded_conv_8/project/weights/Initializer/truncated_normal (24.58k/49.15k flops)\n",
            "    MobilenetV2/expanded_conv_8/project/weights/Initializer/truncated_normal/mul (24.58k/24.58k flops)\n",
            "  MobilenetV2/expanded_conv_7/expand/weights/Initializer/truncated_normal (24.58k/49.15k flops)\n",
            "    MobilenetV2/expanded_conv_7/expand/weights/Initializer/truncated_normal/mul (24.58k/24.58k flops)\n",
            "  MobilenetV2/expanded_conv_10/expand/weights/Initializer/truncated_normal (24.58k/49.15k flops)\n",
            "    MobilenetV2/expanded_conv_10/expand/weights/Initializer/truncated_normal/mul (24.58k/24.58k flops)\n",
            "  MobilenetV2/expanded_conv_7/project/weights/Initializer/truncated_normal (24.58k/49.15k flops)\n",
            "    MobilenetV2/expanded_conv_7/project/weights/Initializer/truncated_normal/mul (24.58k/24.58k flops)\n",
            "  MobilenetV2/expanded_conv_8/expand/weights/Initializer/truncated_normal (24.58k/49.15k flops)\n",
            "    MobilenetV2/expanded_conv_8/expand/weights/Initializer/truncated_normal/mul (24.58k/24.58k flops)\n",
            "  MobilenetV2/expanded_conv_9/expand/weights/Initializer/truncated_normal (24.58k/49.15k flops)\n",
            "    MobilenetV2/expanded_conv_9/expand/weights/Initializer/truncated_normal/mul (24.58k/24.58k flops)\n",
            "  MobilenetV2/expanded_conv_9/project/weights/Initializer/truncated_normal (24.58k/49.15k flops)\n",
            "    MobilenetV2/expanded_conv_9/project/weights/Initializer/truncated_normal/mul (24.58k/24.58k flops)\n",
            "  MobilenetV2/expanded_conv_6/project/kernel/Regularizer/l2_regularizer (1/36.86k flops)\n",
            "    MobilenetV2/expanded_conv_6/project/kernel/Regularizer/l2_regularizer/L2Loss (36.86k/36.86k flops)\n",
            "  MobilenetV2/expanded_conv_6/project/weights/Initializer/truncated_normal (12.29k/24.58k flops)\n",
            "    MobilenetV2/expanded_conv_6/project/weights/Initializer/truncated_normal/mul (12.29k/12.29k flops)\n",
            "  MobilenetV2/expanded_conv_6/expand/kernel/Regularizer/l2_regularizer (1/18.43k flops)\n",
            "    MobilenetV2/expanded_conv_6/expand/kernel/Regularizer/l2_regularizer/L2Loss (18.43k/18.43k flops)\n",
            "  MobilenetV2/expanded_conv_5/expand/kernel/Regularizer/l2_regularizer (1/18.43k flops)\n",
            "    MobilenetV2/expanded_conv_5/expand/kernel/Regularizer/l2_regularizer/L2Loss (18.43k/18.43k flops)\n",
            "  MobilenetV2/expanded_conv_4/project/kernel/Regularizer/l2_regularizer (1/18.43k flops)\n",
            "    MobilenetV2/expanded_conv_4/project/kernel/Regularizer/l2_regularizer/L2Loss (18.43k/18.43k flops)\n",
            "  MobilenetV2/expanded_conv_4/expand/kernel/Regularizer/l2_regularizer (1/18.43k flops)\n",
            "    MobilenetV2/expanded_conv_4/expand/kernel/Regularizer/l2_regularizer/L2Loss (18.43k/18.43k flops)\n",
            "  MobilenetV2/expanded_conv_5/project/kernel/Regularizer/l2_regularizer (1/18.43k flops)\n",
            "    MobilenetV2/expanded_conv_5/project/kernel/Regularizer/l2_regularizer/L2Loss (18.43k/18.43k flops)\n",
            "  MobilenetV2/expanded_conv_15/depthwise/depthwise_weights/Initializer/truncated_normal (8.64k/17.28k flops)\n",
            "    MobilenetV2/expanded_conv_15/depthwise/depthwise_weights/Initializer/truncated_normal/mul (8.64k/8.64k flops)\n",
            "  MobilenetV2/expanded_conv_14/depthwise/depthwise_weights/Initializer/truncated_normal (8.64k/17.28k flops)\n",
            "    MobilenetV2/expanded_conv_14/depthwise/depthwise_weights/Initializer/truncated_normal/mul (8.64k/8.64k flops)\n",
            "  MobilenetV2/expanded_conv_16/depthwise/depthwise_weights/Initializer/truncated_normal (8.64k/17.28k flops)\n",
            "    MobilenetV2/expanded_conv_16/depthwise/depthwise_weights/Initializer/truncated_normal/mul (8.64k/8.64k flops)\n",
            "  MobilenetV2/expanded_conv_3/project/kernel/Regularizer/l2_regularizer (1/13.82k flops)\n",
            "    MobilenetV2/expanded_conv_3/project/kernel/Regularizer/l2_regularizer/L2Loss (13.82k/13.82k flops)\n",
            "  MobilenetV2/expanded_conv_4/expand/weights/Initializer/truncated_normal (6.14k/12.29k flops)\n",
            "    MobilenetV2/expanded_conv_4/expand/weights/Initializer/truncated_normal/mul (6.14k/6.14k flops)\n",
            "  MobilenetV2/expanded_conv_6/expand/weights/Initializer/truncated_normal (6.14k/12.29k flops)\n",
            "    MobilenetV2/expanded_conv_6/expand/weights/Initializer/truncated_normal/mul (6.14k/6.14k flops)\n",
            "  MobilenetV2/expanded_conv_5/project/weights/Initializer/truncated_normal (6.14k/12.29k flops)\n",
            "    MobilenetV2/expanded_conv_5/project/weights/Initializer/truncated_normal/mul (6.14k/6.14k flops)\n",
            "  MobilenetV2/expanded_conv_4/project/weights/Initializer/truncated_normal (6.14k/12.29k flops)\n",
            "    MobilenetV2/expanded_conv_4/project/weights/Initializer/truncated_normal/mul (6.14k/6.14k flops)\n",
            "  MobilenetV2/expanded_conv_5/expand/weights/Initializer/truncated_normal (6.14k/12.29k flops)\n",
            "    MobilenetV2/expanded_conv_5/expand/weights/Initializer/truncated_normal/mul (6.14k/6.14k flops)\n",
            "  MobilenetV2/expanded_conv_2/project/kernel/Regularizer/l2_regularizer (1/10.37k flops)\n",
            "    MobilenetV2/expanded_conv_2/project/kernel/Regularizer/l2_regularizer/L2Loss (10.37k/10.37k flops)\n",
            "  MobilenetV2/expanded_conv_2/expand/kernel/Regularizer/l2_regularizer (1/10.37k flops)\n",
            "    MobilenetV2/expanded_conv_2/expand/kernel/Regularizer/l2_regularizer/L2Loss (10.37k/10.37k flops)\n",
            "  MobilenetV2/expanded_conv_13/depthwise/depthwise_weights/Initializer/truncated_normal (5.18k/10.37k flops)\n",
            "    MobilenetV2/expanded_conv_13/depthwise/depthwise_weights/Initializer/truncated_normal/mul (5.18k/5.18k flops)\n",
            "  MobilenetV2/expanded_conv_11/depthwise/depthwise_weights/Initializer/truncated_normal (5.18k/10.37k flops)\n",
            "    MobilenetV2/expanded_conv_11/depthwise/depthwise_weights/Initializer/truncated_normal/mul (5.18k/5.18k flops)\n",
            "  MobilenetV2/expanded_conv_3/expand/kernel/Regularizer/l2_regularizer (1/10.37k flops)\n",
            "    MobilenetV2/expanded_conv_3/expand/kernel/Regularizer/l2_regularizer/L2Loss (10.37k/10.37k flops)\n",
            "  MobilenetV2/expanded_conv_12/depthwise/depthwise_weights/Initializer/truncated_normal (5.18k/10.37k flops)\n",
            "    MobilenetV2/expanded_conv_12/depthwise/depthwise_weights/Initializer/truncated_normal/mul (5.18k/5.18k flops)\n",
            "  MobilenetV2/expanded_conv_3/project/weights/Initializer/truncated_normal (4.61k/9.22k flops)\n",
            "    MobilenetV2/expanded_conv_3/project/weights/Initializer/truncated_normal/mul (4.61k/4.61k flops)\n",
            "  MobilenetV2/expanded_conv_1/project/kernel/Regularizer/l2_regularizer (1/6.91k flops)\n",
            "    MobilenetV2/expanded_conv_1/project/kernel/Regularizer/l2_regularizer/L2Loss (6.91k/6.91k flops)\n",
            "  MobilenetV2/expanded_conv_3/expand/weights/Initializer/truncated_normal (3.46k/6.91k flops)\n",
            "    MobilenetV2/expanded_conv_3/expand/weights/Initializer/truncated_normal/mul (3.46k/3.46k flops)\n",
            "  MobilenetV2/expanded_conv_9/depthwise/depthwise_weights/Initializer/truncated_normal (3.46k/6.91k flops)\n",
            "    MobilenetV2/expanded_conv_9/depthwise/depthwise_weights/Initializer/truncated_normal/mul (3.46k/3.46k flops)\n",
            "  MobilenetV2/expanded_conv_8/depthwise/depthwise_weights/Initializer/truncated_normal (3.46k/6.91k flops)\n",
            "    MobilenetV2/expanded_conv_8/depthwise/depthwise_weights/Initializer/truncated_normal/mul (3.46k/3.46k flops)\n",
            "  MobilenetV2/expanded_conv_10/depthwise/depthwise_weights/Initializer/truncated_normal (3.46k/6.91k flops)\n",
            "    MobilenetV2/expanded_conv_10/depthwise/depthwise_weights/Initializer/truncated_normal/mul (3.46k/3.46k flops)\n",
            "  MobilenetV2/expanded_conv_7/depthwise/depthwise_weights/Initializer/truncated_normal (3.46k/6.91k flops)\n",
            "    MobilenetV2/expanded_conv_7/depthwise/depthwise_weights/Initializer/truncated_normal/mul (3.46k/3.46k flops)\n",
            "  MobilenetV2/expanded_conv_2/expand/weights/Initializer/truncated_normal (3.46k/6.91k flops)\n",
            "    MobilenetV2/expanded_conv_2/expand/weights/Initializer/truncated_normal/mul (3.46k/3.46k flops)\n",
            "  MobilenetV2/expanded_conv_2/project/weights/Initializer/truncated_normal (3.46k/6.91k flops)\n",
            "    MobilenetV2/expanded_conv_2/project/weights/Initializer/truncated_normal/mul (3.46k/3.46k flops)\n",
            "  MobilenetV2/expanded_conv_1/expand/kernel/Regularizer/l2_regularizer (1/4.61k flops)\n",
            "    MobilenetV2/expanded_conv_1/expand/kernel/Regularizer/l2_regularizer/L2Loss (4.61k/4.61k flops)\n",
            "  MobilenetV2/expanded_conv_1/project/weights/Initializer/truncated_normal (2.30k/4.61k flops)\n",
            "    MobilenetV2/expanded_conv_1/project/weights/Initializer/truncated_normal/mul (2.30k/2.30k flops)\n",
            "  MobilenetV2/expanded_conv_4/depthwise/depthwise_weights/Initializer/truncated_normal (1.73k/3.46k flops)\n",
            "    MobilenetV2/expanded_conv_4/depthwise/depthwise_weights/Initializer/truncated_normal/mul (1.73k/1.73k flops)\n",
            "  MobilenetV2/expanded_conv_6/depthwise/depthwise_weights/Initializer/truncated_normal (1.73k/3.46k flops)\n",
            "    MobilenetV2/expanded_conv_6/depthwise/depthwise_weights/Initializer/truncated_normal/mul (1.73k/1.73k flops)\n",
            "  MobilenetV2/expanded_conv_5/depthwise/depthwise_weights/Initializer/truncated_normal (1.73k/3.46k flops)\n",
            "    MobilenetV2/expanded_conv_5/depthwise/depthwise_weights/Initializer/truncated_normal/mul (1.73k/1.73k flops)\n",
            "  MobilenetV2/expanded_conv_1/expand/weights/Initializer/truncated_normal (1.54k/3.07k flops)\n",
            "    MobilenetV2/expanded_conv_1/expand/weights/Initializer/truncated_normal/mul (1.54k/1.54k flops)\n",
            "  MobilenetV2/expanded_conv_3/depthwise/depthwise_weights/Initializer/truncated_normal (1.30k/2.59k flops)\n",
            "    MobilenetV2/expanded_conv_3/depthwise/depthwise_weights/Initializer/truncated_normal/mul (1.30k/1.30k flops)\n",
            "  MobilenetV2/expanded_conv_2/depthwise/depthwise_weights/Initializer/truncated_normal (1.30k/2.59k flops)\n",
            "    MobilenetV2/expanded_conv_2/depthwise/depthwise_weights/Initializer/truncated_normal/mul (1.30k/1.30k flops)\n",
            "  MobilenetV2/Conv/kernel/Regularizer/l2_regularizer (1/2.59k flops)\n",
            "    MobilenetV2/Conv/kernel/Regularizer/l2_regularizer/L2Loss (2.59k/2.59k flops)\n",
            "  logits/semantic/BiasAdd (2.18k/2.18k flops)\n",
            "  MobilenetV2/Conv/weights/Initializer/truncated_normal (864/1.73k flops)\n",
            "    MobilenetV2/Conv/weights/Initializer/truncated_normal/mul (864/864 flops)\n",
            "  MobilenetV2/expanded_conv_1/depthwise/depthwise_weights/Initializer/truncated_normal (864/1.73k flops)\n",
            "    MobilenetV2/expanded_conv_1/depthwise/depthwise_weights/Initializer/truncated_normal/mul (864/864 flops)\n",
            "  MobilenetV2/expanded_conv/project/kernel/Regularizer/l2_regularizer (1/1.54k flops)\n",
            "    MobilenetV2/expanded_conv/project/kernel/Regularizer/l2_regularizer/L2Loss (1.53k/1.53k flops)\n",
            "  logits/semantic/kernel/Regularizer/l2_regularizer (1/1.54k flops)\n",
            "    logits/semantic/kernel/Regularizer/l2_regularizer/L2Loss (1.53k/1.53k flops)\n",
            "  MobilenetV2/expanded_conv/project/weights/Initializer/truncated_normal (512/1.02k flops)\n",
            "    MobilenetV2/expanded_conv/project/weights/Initializer/truncated_normal/mul (512/512 flops)\n",
            "  logits/semantic/weights/Initializer/truncated_normal (512/1.02k flops)\n",
            "    logits/semantic/weights/Initializer/truncated_normal/mul (512/512 flops)\n",
            "  MobilenetV2/expanded_conv/depthwise/depthwise_weights/Initializer/truncated_normal (288/576 flops)\n",
            "    MobilenetV2/expanded_conv/depthwise/depthwise_weights/Initializer/truncated_normal/mul (288/288 flops)\n",
            "  mean_iou/AssignAdd (4/4 flops)\n",
            "  mean_iou/Sum (2/2 flops)\n",
            "  mean_iou/Sum_1 (2/2 flops)\n",
            "  mean_iou/add (2/2 flops)\n",
            "  mean_iou/NotEqual (2/2 flops)\n",
            "  mean_iou/div (2/2 flops)\n",
            "  mean_iou/sub (2/2 flops)\n",
            "  mean_iou/Greater (2/2 flops)\n",
            "  mean_iou/Sum_2 (1/1 flops)\n",
            "  mean_iou/Greater_1 (1/1 flops)\n",
            "  mean_iou/mean_iou (1/1 flops)\n",
            "  mean_iou/truediv (1/1 flops)\n",
            "\n",
            "======================End of Report==========================\n",
            "INFO:tensorflow:Waiting for new checkpoint at /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train\n",
            "INFO:tensorflow:Found new checkpoint at /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt-15000\n",
            "INFO:tensorflow:Graph was finalized.\n",
            "2019-05-18 12:54:02.581241: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2300000000 Hz\n",
            "2019-05-18 12:54:02.581491: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x56305e47f8c0 executing computations on platform Host. Devices:\n",
            "2019-05-18 12:54:02.581523: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): <undefined>, <undefined>\n",
            "2019-05-18 12:54:02.768238: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:998] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
            "2019-05-18 12:54:02.768785: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x56305e481020 executing computations on platform CUDA. Devices:\n",
            "2019-05-18 12:54:02.768818: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): Tesla T4, Compute Capability 7.5\n",
            "2019-05-18 12:54:02.769194: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties: \n",
            "name: Tesla T4 major: 7 minor: 5 memoryClockRate(GHz): 1.59\n",
            "pciBusID: 0000:00:04.0\n",
            "totalMemory: 14.73GiB freeMemory: 14.60GiB\n",
            "2019-05-18 12:54:02.769220: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 12:54:03.274099: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 12:54:03.274183: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 12:54:03.274196: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 12:54:03.274462: W tensorflow/core/common_runtime/gpu/gpu_bfc_allocator.cc:42] Overriding allow_growth setting because the TF_FORCE_GPU_ALLOW_GROWTH environment variable is set. Original config value was 0.\n",
            "2019-05-18 12:54:03.274528: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14115 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/training/saver.py:1266: checkpoint_exists (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use standard file APIs to check for files with this prefix.\n",
            "INFO:tensorflow:Restoring parameters from /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt-15000\n",
            "INFO:tensorflow:Running local_init_op.\n",
            "INFO:tensorflow:Done running local_init_op.\n",
            "INFO:tensorflow:Starting evaluation at 2019-05-18-12:54:04\n",
            "*********** Rizan Vis kortase ************\n",
            "\n",
            "WARNING: The TensorFlow contrib module will not be included in TensorFlow 2.0.\n",
            "For more information, please see:\n",
            "  * https://github.com/tensorflow/community/blob/master/rfcs/20180907-contrib-sunset.md\n",
            "  * https://github.com/tensorflow/addons\n",
            "If you depend on functionality not listed there, please file an issue.\n",
            "\n",
            "INFO:tensorflow:Visualizing on val set\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/control_flow_ops.py:423: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n",
            "INFO:tensorflow:Performing single-scale test.\n",
            "WARNING:tensorflow:From /root/models/research/deeplab/core/feature_extractor.py:196: to_float (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "INFO:tensorflow:Waiting for new checkpoint at /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train\n",
            "INFO:tensorflow:Found new checkpoint at /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt-15000\n",
            "INFO:tensorflow:Starting visualization at 2019-05-18-12:54:25\n",
            "INFO:tensorflow:Visualizing with model /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt-15000\n",
            "INFO:tensorflow:Graph was finalized.\n",
            "2019-05-18 12:54:25.591344: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2300000000 Hz\n",
            "2019-05-18 12:54:25.591553: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x5652e47b11e0 executing computations on platform Host. Devices:\n",
            "2019-05-18 12:54:25.591581: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): <undefined>, <undefined>\n",
            "2019-05-18 12:54:25.777248: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:998] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
            "2019-05-18 12:54:25.777785: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x5652e47b3de0 executing computations on platform CUDA. Devices:\n",
            "2019-05-18 12:54:25.777818: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): Tesla T4, Compute Capability 7.5\n",
            "2019-05-18 12:54:25.778187: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties: \n",
            "name: Tesla T4 major: 7 minor: 5 memoryClockRate(GHz): 1.59\n",
            "pciBusID: 0000:00:04.0\n",
            "totalMemory: 14.73GiB freeMemory: 14.60GiB\n",
            "2019-05-18 12:54:25.778212: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 12:54:26.272783: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 12:54:26.272863: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 12:54:26.272876: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 12:54:26.273148: W tensorflow/core/common_runtime/gpu/gpu_bfc_allocator.cc:42] Overriding allow_growth setting because the TF_FORCE_GPU_ALLOW_GROWTH environment variable is set. Original config value was 0.\n",
            "2019-05-18 12:54:26.273198: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14115 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/training/saver.py:1266: checkpoint_exists (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use standard file APIs to check for files with this prefix.\n",
            "INFO:tensorflow:Restoring parameters from /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt-15000\n",
            "INFO:tensorflow:Running local_init_op.\n",
            "INFO:tensorflow:Done running local_init_op.\n",
            "INFO:tensorflow:Visualizing batch 1\n",
            "INFO:tensorflow:Visualizing batch 2\n",
            "INFO:tensorflow:Visualizing batch 3\n",
            "INFO:tensorflow:Visualizing batch 4\n",
            "INFO:tensorflow:Visualizing batch 5\n",
            "INFO:tensorflow:Visualizing batch 6\n",
            "INFO:tensorflow:Visualizing batch 7\n",
            "INFO:tensorflow:Visualizing batch 8\n",
            "INFO:tensorflow:Visualizing batch 9\n",
            "INFO:tensorflow:Visualizing batch 10\n",
            "INFO:tensorflow:Visualizing batch 11\n",
            "INFO:tensorflow:Visualizing batch 12\n",
            "INFO:tensorflow:Visualizing batch 13\n",
            "INFO:tensorflow:Visualizing batch 14\n",
            "INFO:tensorflow:Visualizing batch 15\n",
            "INFO:tensorflow:Visualizing batch 16\n",
            "INFO:tensorflow:Visualizing batch 17\n",
            "INFO:tensorflow:Visualizing batch 18\n",
            "INFO:tensorflow:Visualizing batch 19\n",
            "INFO:tensorflow:Visualizing batch 20\n",
            "INFO:tensorflow:Visualizing batch 21\n",
            "INFO:tensorflow:Visualizing batch 22\n",
            "INFO:tensorflow:Visualizing batch 23\n",
            "INFO:tensorflow:Visualizing batch 24\n",
            "INFO:tensorflow:Visualizing batch 25\n",
            "INFO:tensorflow:Visualizing batch 26\n",
            "INFO:tensorflow:Visualizing batch 27\n",
            "INFO:tensorflow:Visualizing batch 28\n",
            "INFO:tensorflow:Visualizing batch 29\n",
            "INFO:tensorflow:Visualizing batch 30\n",
            "INFO:tensorflow:Visualizing batch 31\n",
            "INFO:tensorflow:Visualizing batch 32\n",
            "INFO:tensorflow:Visualizing batch 33\n",
            "INFO:tensorflow:Visualizing batch 34\n",
            "INFO:tensorflow:Visualizing batch 35\n",
            "INFO:tensorflow:Visualizing batch 36\n",
            "INFO:tensorflow:Visualizing batch 37\n",
            "INFO:tensorflow:Visualizing batch 38\n",
            "INFO:tensorflow:Visualizing batch 39\n",
            "INFO:tensorflow:Visualizing batch 40\n",
            "INFO:tensorflow:Visualizing batch 41\n",
            "INFO:tensorflow:Visualizing batch 42\n",
            "INFO:tensorflow:Visualizing batch 43\n",
            "INFO:tensorflow:Visualizing batch 44\n",
            "INFO:tensorflow:Visualizing batch 45\n",
            "INFO:tensorflow:Visualizing batch 46\n",
            "INFO:tensorflow:Visualizing batch 47\n",
            "INFO:tensorflow:Visualizing batch 48\n",
            "INFO:tensorflow:Visualizing batch 49\n",
            "INFO:tensorflow:Visualizing batch 50\n",
            "INFO:tensorflow:Visualizing batch 51\n",
            "INFO:tensorflow:Visualizing batch 52\n",
            "INFO:tensorflow:Visualizing batch 53\n",
            "INFO:tensorflow:Visualizing batch 54\n",
            "INFO:tensorflow:Visualizing batch 55\n",
            "INFO:tensorflow:Visualizing batch 56\n",
            "INFO:tensorflow:Visualizing batch 57\n",
            "INFO:tensorflow:Visualizing batch 58\n",
            "INFO:tensorflow:Visualizing batch 59\n",
            "INFO:tensorflow:Visualizing batch 60\n",
            "INFO:tensorflow:Visualizing batch 61\n",
            "INFO:tensorflow:Visualizing batch 62\n",
            "INFO:tensorflow:Visualizing batch 63\n",
            "INFO:tensorflow:Visualizing batch 64\n",
            "INFO:tensorflow:Visualizing batch 65\n",
            "INFO:tensorflow:Visualizing batch 66\n",
            "INFO:tensorflow:Visualizing batch 67\n",
            "INFO:tensorflow:Visualizing batch 68\n",
            "INFO:tensorflow:Visualizing batch 69\n",
            "INFO:tensorflow:Visualizing batch 70\n",
            "INFO:tensorflow:Visualizing batch 71\n",
            "INFO:tensorflow:Visualizing batch 72\n",
            "INFO:tensorflow:Visualizing batch 73\n",
            "INFO:tensorflow:Visualizing batch 74\n",
            "INFO:tensorflow:Visualizing batch 75\n",
            "INFO:tensorflow:Visualizing batch 76\n",
            "INFO:tensorflow:Visualizing batch 77\n",
            "INFO:tensorflow:Visualizing batch 78\n",
            "INFO:tensorflow:Visualizing batch 79\n",
            "INFO:tensorflow:Visualizing batch 80\n",
            "INFO:tensorflow:Visualizing batch 81\n",
            "INFO:tensorflow:Visualizing batch 82\n",
            "INFO:tensorflow:Visualizing batch 83\n",
            "INFO:tensorflow:Visualizing batch 84\n",
            "INFO:tensorflow:Visualizing batch 85\n",
            "INFO:tensorflow:Visualizing batch 86\n",
            "INFO:tensorflow:Visualizing batch 87\n",
            "INFO:tensorflow:Visualizing batch 88\n",
            "INFO:tensorflow:Visualizing batch 89\n",
            "INFO:tensorflow:Visualizing batch 90\n",
            "INFO:tensorflow:Visualizing batch 91\n",
            "INFO:tensorflow:Visualizing batch 92\n",
            "INFO:tensorflow:Visualizing batch 93\n",
            "INFO:tensorflow:Visualizing batch 94\n",
            "INFO:tensorflow:Visualizing batch 95\n",
            "INFO:tensorflow:Visualizing batch 96\n",
            "INFO:tensorflow:Visualizing batch 97\n",
            "INFO:tensorflow:Visualizing batch 98\n",
            "INFO:tensorflow:Visualizing batch 99\n",
            "INFO:tensorflow:Visualizing batch 100\n",
            "INFO:tensorflow:Visualizing batch 101\n",
            "INFO:tensorflow:Visualizing batch 102\n",
            "INFO:tensorflow:Visualizing batch 103\n",
            "INFO:tensorflow:Visualizing batch 104\n",
            "INFO:tensorflow:Visualizing batch 105\n",
            "INFO:tensorflow:Visualizing batch 106\n",
            "INFO:tensorflow:Visualizing batch 107\n",
            "INFO:tensorflow:Visualizing batch 108\n",
            "INFO:tensorflow:Visualizing batch 109\n",
            "INFO:tensorflow:Visualizing batch 110\n",
            "INFO:tensorflow:Visualizing batch 111\n",
            "INFO:tensorflow:Visualizing batch 112\n",
            "INFO:tensorflow:Visualizing batch 113\n",
            "INFO:tensorflow:Visualizing batch 114\n",
            "INFO:tensorflow:Visualizing batch 115\n",
            "INFO:tensorflow:Visualizing batch 116\n",
            "INFO:tensorflow:Visualizing batch 117\n",
            "INFO:tensorflow:Visualizing batch 118\n",
            "INFO:tensorflow:Visualizing batch 119\n",
            "INFO:tensorflow:Visualizing batch 120\n",
            "INFO:tensorflow:Visualizing batch 121\n",
            "INFO:tensorflow:Visualizing batch 122\n",
            "INFO:tensorflow:Visualizing batch 123\n",
            "INFO:tensorflow:Visualizing batch 124\n",
            "INFO:tensorflow:Visualizing batch 125\n",
            "INFO:tensorflow:Visualizing batch 126\n",
            "INFO:tensorflow:Visualizing batch 127\n",
            "INFO:tensorflow:Visualizing batch 128\n",
            "INFO:tensorflow:Visualizing batch 129\n",
            "INFO:tensorflow:Visualizing batch 130\n",
            "INFO:tensorflow:Visualizing batch 131\n",
            "INFO:tensorflow:Visualizing batch 132\n",
            "INFO:tensorflow:Visualizing batch 133\n",
            "INFO:tensorflow:Visualizing batch 134\n",
            "INFO:tensorflow:Visualizing batch 135\n",
            "INFO:tensorflow:Visualizing batch 136\n",
            "INFO:tensorflow:Visualizing batch 137\n",
            "INFO:tensorflow:Visualizing batch 138\n",
            "INFO:tensorflow:Visualizing batch 139\n",
            "INFO:tensorflow:Visualizing batch 140\n",
            "INFO:tensorflow:Visualizing batch 141\n",
            "INFO:tensorflow:Visualizing batch 142\n",
            "INFO:tensorflow:Visualizing batch 143\n",
            "INFO:tensorflow:Visualizing batch 144\n",
            "INFO:tensorflow:Visualizing batch 145\n",
            "INFO:tensorflow:Visualizing batch 146\n",
            "INFO:tensorflow:Visualizing batch 147\n",
            "INFO:tensorflow:Visualizing batch 148\n",
            "INFO:tensorflow:Visualizing batch 149\n",
            "INFO:tensorflow:Visualizing batch 150\n",
            "INFO:tensorflow:Visualizing batch 151\n",
            "INFO:tensorflow:Visualizing batch 152\n",
            "INFO:tensorflow:Visualizing batch 153\n",
            "INFO:tensorflow:Visualizing batch 154\n",
            "INFO:tensorflow:Visualizing batch 155\n",
            "INFO:tensorflow:Visualizing batch 156\n",
            "INFO:tensorflow:Visualizing batch 157\n",
            "INFO:tensorflow:Visualizing batch 158\n",
            "INFO:tensorflow:Visualizing batch 159\n",
            "INFO:tensorflow:Visualizing batch 160\n",
            "INFO:tensorflow:Visualizing batch 161\n",
            "INFO:tensorflow:Visualizing batch 162\n",
            "INFO:tensorflow:Visualizing batch 163\n",
            "INFO:tensorflow:Visualizing batch 164\n",
            "INFO:tensorflow:Visualizing batch 165\n",
            "INFO:tensorflow:Visualizing batch 166\n",
            "INFO:tensorflow:Visualizing batch 167\n",
            "INFO:tensorflow:Visualizing batch 168\n",
            "INFO:tensorflow:Visualizing batch 169\n",
            "INFO:tensorflow:Visualizing batch 170\n",
            "INFO:tensorflow:Visualizing batch 171\n",
            "INFO:tensorflow:Visualizing batch 172\n",
            "INFO:tensorflow:Visualizing batch 173\n",
            "INFO:tensorflow:Visualizing batch 174\n",
            "INFO:tensorflow:Visualizing batch 175\n",
            "INFO:tensorflow:Visualizing batch 176\n",
            "INFO:tensorflow:Visualizing batch 177\n",
            "INFO:tensorflow:Visualizing batch 178\n",
            "INFO:tensorflow:Visualizing batch 179\n",
            "INFO:tensorflow:Visualizing batch 180\n",
            "INFO:tensorflow:Visualizing batch 181\n",
            "INFO:tensorflow:Visualizing batch 182\n",
            "INFO:tensorflow:Visualizing batch 183\n",
            "INFO:tensorflow:Visualizing batch 184\n",
            "INFO:tensorflow:Visualizing batch 185\n",
            "INFO:tensorflow:Visualizing batch 186\n",
            "INFO:tensorflow:Visualizing batch 187\n",
            "INFO:tensorflow:Visualizing batch 188\n",
            "INFO:tensorflow:Visualizing batch 189\n",
            "INFO:tensorflow:Visualizing batch 190\n",
            "INFO:tensorflow:Visualizing batch 191\n",
            "INFO:tensorflow:Visualizing batch 192\n",
            "INFO:tensorflow:Visualizing batch 193\n",
            "INFO:tensorflow:Visualizing batch 194\n",
            "INFO:tensorflow:Visualizing batch 195\n",
            "INFO:tensorflow:Visualizing batch 196\n",
            "INFO:tensorflow:Visualizing batch 197\n",
            "INFO:tensorflow:Visualizing batch 198\n",
            "INFO:tensorflow:Visualizing batch 199\n",
            "INFO:tensorflow:Visualizing batch 200\n",
            "INFO:tensorflow:Visualizing batch 201\n",
            "INFO:tensorflow:Visualizing batch 202\n",
            "INFO:tensorflow:Visualizing batch 203\n",
            "INFO:tensorflow:Visualizing batch 204\n",
            "INFO:tensorflow:Visualizing batch 205\n",
            "INFO:tensorflow:Visualizing batch 206\n",
            "INFO:tensorflow:Visualizing batch 207\n",
            "INFO:tensorflow:Visualizing batch 208\n",
            "INFO:tensorflow:Visualizing batch 209\n",
            "INFO:tensorflow:Visualizing batch 210\n",
            "INFO:tensorflow:Visualizing batch 211\n",
            "INFO:tensorflow:Visualizing batch 212\n",
            "INFO:tensorflow:Visualizing batch 213\n",
            "INFO:tensorflow:Visualizing batch 214\n",
            "INFO:tensorflow:Visualizing batch 215\n",
            "INFO:tensorflow:Visualizing batch 216\n",
            "INFO:tensorflow:Visualizing batch 217\n",
            "INFO:tensorflow:Visualizing batch 218\n",
            "INFO:tensorflow:Visualizing batch 219\n",
            "INFO:tensorflow:Visualizing batch 220\n",
            "INFO:tensorflow:Visualizing batch 221\n",
            "INFO:tensorflow:Visualizing batch 222\n",
            "INFO:tensorflow:Visualizing batch 223\n",
            "INFO:tensorflow:Visualizing batch 224\n",
            "INFO:tensorflow:Visualizing batch 225\n",
            "INFO:tensorflow:Visualizing batch 226\n",
            "INFO:tensorflow:Visualizing batch 227\n",
            "INFO:tensorflow:Visualizing batch 228\n",
            "INFO:tensorflow:Visualizing batch 229\n",
            "INFO:tensorflow:Visualizing batch 230\n",
            "INFO:tensorflow:Visualizing batch 231\n",
            "INFO:tensorflow:Visualizing batch 232\n",
            "INFO:tensorflow:Visualizing batch 233\n",
            "INFO:tensorflow:Visualizing batch 234\n",
            "INFO:tensorflow:Visualizing batch 235\n",
            "INFO:tensorflow:Visualizing batch 236\n",
            "INFO:tensorflow:Visualizing batch 237\n",
            "INFO:tensorflow:Visualizing batch 238\n",
            "INFO:tensorflow:Visualizing batch 239\n",
            "INFO:tensorflow:Visualizing batch 240\n",
            "INFO:tensorflow:Visualizing batch 241\n",
            "INFO:tensorflow:Visualizing batch 242\n",
            "INFO:tensorflow:Visualizing batch 243\n",
            "INFO:tensorflow:Visualizing batch 244\n",
            "INFO:tensorflow:Visualizing batch 245\n",
            "INFO:tensorflow:Visualizing batch 246\n",
            "INFO:tensorflow:Visualizing batch 247\n",
            "INFO:tensorflow:Visualizing batch 248\n",
            "INFO:tensorflow:Visualizing batch 249\n",
            "INFO:tensorflow:Visualizing batch 250\n",
            "INFO:tensorflow:Visualizing batch 251\n",
            "INFO:tensorflow:Visualizing batch 252\n",
            "INFO:tensorflow:Visualizing batch 253\n",
            "INFO:tensorflow:Visualizing batch 254\n",
            "INFO:tensorflow:Visualizing batch 255\n",
            "INFO:tensorflow:Visualizing batch 256\n",
            "INFO:tensorflow:Visualizing batch 257\n",
            "INFO:tensorflow:Visualizing batch 258\n",
            "INFO:tensorflow:Visualizing batch 259\n",
            "INFO:tensorflow:Visualizing batch 260\n",
            "INFO:tensorflow:Visualizing batch 261\n",
            "INFO:tensorflow:Visualizing batch 262\n",
            "INFO:tensorflow:Visualizing batch 263\n",
            "INFO:tensorflow:Visualizing batch 264\n",
            "INFO:tensorflow:Visualizing batch 265\n",
            "INFO:tensorflow:Visualizing batch 266\n",
            "INFO:tensorflow:Visualizing batch 267\n",
            "INFO:tensorflow:Visualizing batch 268\n",
            "INFO:tensorflow:Visualizing batch 269\n",
            "INFO:tensorflow:Visualizing batch 270\n",
            "INFO:tensorflow:Visualizing batch 271\n",
            "INFO:tensorflow:Visualizing batch 272\n",
            "INFO:tensorflow:Visualizing batch 273\n",
            "INFO:tensorflow:Visualizing batch 274\n",
            "INFO:tensorflow:Visualizing batch 275\n",
            "INFO:tensorflow:Visualizing batch 276\n",
            "INFO:tensorflow:Visualizing batch 277\n",
            "INFO:tensorflow:Visualizing batch 278\n",
            "INFO:tensorflow:Visualizing batch 279\n",
            "INFO:tensorflow:Visualizing batch 280\n",
            "INFO:tensorflow:Visualizing batch 281\n",
            "INFO:tensorflow:Visualizing batch 282\n",
            "INFO:tensorflow:Visualizing batch 283\n",
            "INFO:tensorflow:Visualizing batch 284\n",
            "INFO:tensorflow:Visualizing batch 285\n",
            "INFO:tensorflow:Visualizing batch 286\n",
            "INFO:tensorflow:Visualizing batch 287\n",
            "INFO:tensorflow:Visualizing batch 288\n",
            "INFO:tensorflow:Visualizing batch 289\n",
            "INFO:tensorflow:Visualizing batch 290\n",
            "INFO:tensorflow:Visualizing batch 291\n",
            "INFO:tensorflow:Visualizing batch 292\n",
            "INFO:tensorflow:Visualizing batch 293\n",
            "INFO:tensorflow:Visualizing batch 294\n",
            "INFO:tensorflow:Visualizing batch 295\n",
            "INFO:tensorflow:Visualizing batch 296\n",
            "INFO:tensorflow:Visualizing batch 297\n",
            "INFO:tensorflow:Visualizing batch 298\n",
            "INFO:tensorflow:Visualizing batch 299\n",
            "INFO:tensorflow:Visualizing batch 300\n",
            "INFO:tensorflow:Visualizing batch 301\n",
            "INFO:tensorflow:Visualizing batch 302\n",
            "INFO:tensorflow:Visualizing batch 303\n",
            "INFO:tensorflow:Visualizing batch 304\n",
            "INFO:tensorflow:Visualizing batch 305\n",
            "INFO:tensorflow:Visualizing batch 306\n",
            "INFO:tensorflow:Visualizing batch 307\n",
            "INFO:tensorflow:Visualizing batch 308\n",
            "INFO:tensorflow:Visualizing batch 309\n",
            "INFO:tensorflow:Visualizing batch 310\n",
            "INFO:tensorflow:Visualizing batch 311\n",
            "INFO:tensorflow:Visualizing batch 312\n",
            "INFO:tensorflow:Visualizing batch 313\n",
            "INFO:tensorflow:Visualizing batch 314\n",
            "INFO:tensorflow:Visualizing batch 315\n",
            "INFO:tensorflow:Visualizing batch 316\n",
            "INFO:tensorflow:Visualizing batch 317\n",
            "INFO:tensorflow:Finished visualization at 2019-05-18-12:55:20\n",
            "*********** Rizan Export kortase ************\n",
            "INFO:tensorflow:Prepare to export model to: /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/export/frozen_inference_graph.pb\n",
            "INFO:tensorflow:Exported model performs single-scale inference.\n",
            "WARNING:tensorflow:From /root/models/research/deeplab/core/feature_extractor.py:196: to_float (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n",
            "WARNING:tensorflow:From /root/models/research/deeplab/export_model.py:163: all_variables (from tensorflow.python.ops.variables) is deprecated and will be removed after 2017-03-02.\n",
            "Instructions for updating:\n",
            "Please use tf.global_variables instead.\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/tools/freeze_graph.py:127: checkpoint_exists (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use standard file APIs to check for files with this prefix.\n",
            "2019-05-18 12:55:25.491911: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2300000000 Hz\n",
            "2019-05-18 12:55:25.492150: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x55b6781acc00 executing computations on platform Host. Devices:\n",
            "2019-05-18 12:55:25.492183: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): <undefined>, <undefined>\n",
            "2019-05-18 12:55:25.683141: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:998] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n",
            "2019-05-18 12:55:25.683689: I tensorflow/compiler/xla/service/service.cc:150] XLA service 0x55b6781acd60 executing computations on platform CUDA. Devices:\n",
            "2019-05-18 12:55:25.683723: I tensorflow/compiler/xla/service/service.cc:158]   StreamExecutor device (0): Tesla T4, Compute Capability 7.5\n",
            "2019-05-18 12:55:25.684073: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties: \n",
            "name: Tesla T4 major: 7 minor: 5 memoryClockRate(GHz): 1.59\n",
            "pciBusID: 0000:00:04.0\n",
            "totalMemory: 14.73GiB freeMemory: 14.60GiB\n",
            "2019-05-18 12:55:25.684097: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1512] Adding visible gpu devices: 0\n",
            "2019-05-18 12:55:26.185883: I tensorflow/core/common_runtime/gpu/gpu_device.cc:984] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
            "2019-05-18 12:55:26.185955: I tensorflow/core/common_runtime/gpu/gpu_device.cc:990]      0 \n",
            "2019-05-18 12:55:26.185978: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1003] 0:   N \n",
            "2019-05-18 12:55:26.186258: W tensorflow/core/common_runtime/gpu/gpu_bfc_allocator.cc:42] Overriding allow_growth setting because the TF_FORCE_GPU_ALLOW_GROWTH environment variable is set. Original config value was 0.\n",
            "2019-05-18 12:55:26.186314: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 14115 MB memory) -> physical GPU (device: 0, name: Tesla T4, pci bus id: 0000:00:04.0, compute capability: 7.5)\n",
            "INFO:tensorflow:Restoring parameters from /root/models/research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/train/model.ckpt-15000\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/tools/freeze_graph.py:232: convert_variables_to_constants (from tensorflow.python.framework.graph_util_impl) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.compat.v1.graph_util.convert_variables_to_constants\n",
            "WARNING:tensorflow:From /usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/graph_util_impl.py:245: extract_sub_graph (from tensorflow.python.framework.graph_util_impl) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.compat.v1.graph_util.extract_sub_graph\n",
            "INFO:tensorflow:Froze 272 variables.\n",
            "INFO:tensorflow:Converted 272 variables to const ops.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9CDNuW1RZi_r",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model_dir = 'research/deeplab/datasets/pascal_voc_seg/exp/train_on_trainval_set_mobilenetv2/export/'"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M0Fjco-RbsfZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "from matplotlib import pyplot as plt\n",
        "from matplotlib import gridspec\n",
        "\n",
        "class DeepLabModel(object):\n",
        "  \"\"\"Class to load deeplab model and run inference.\"\"\"\n",
        "\n",
        "  INPUT_TENSOR_NAME = 'ImageTensor:0'\n",
        "  OUTPUT_TENSOR_NAME = 'SemanticPredictions:0'\n",
        "  INPUT_SIZE = 513\n",
        "  FROZEN_GRAPH_NAME = 'frozen_inference_graph'\n",
        "\n",
        "  def __init__(self, tarball_path):\n",
        "    \"\"\"Creates and loads pretrained deeplab model.\"\"\"\n",
        "    self.graph = tf.Graph()\n",
        "\n",
        "    graph_def = None\n",
        "    # Extract frozen graph from tar archive.\n",
        "    tar_file = tarfile.open(tarball_path)\n",
        "    for tar_info in tar_file.getmembers():\n",
        "      if self.FROZEN_GRAPH_NAME in os.path.basename(tar_info.name):\n",
        "        file_handle = tar_file.extractfile(tar_info)\n",
        "        graph_def = tf.GraphDef.FromString(file_handle.read())\n",
        "        break\n",
        "\n",
        "    tar_file.close()\n",
        "\n",
        "    if graph_def is None:\n",
        "      raise RuntimeError('Cannot find inference graph in tar archive.')\n",
        "\n",
        "    with self.graph.as_default():\n",
        "      tf.import_graph_def(graph_def, name='')\n",
        "\n",
        "    self.sess = tf.Session(graph=self.graph)\n",
        "\n",
        "  def run(self, image):\n",
        "    \"\"\"Runs inference on a single image.\n",
        "\n",
        "    Args:\n",
        "      image: A PIL.Image object, raw input image.\n",
        "\n",
        "    Returns:\n",
        "      resized_image: RGB image resized from original input image.\n",
        "      seg_map: Segmentation map of `resized_image`.\n",
        "    \"\"\"\n",
        "    width, height = image.size\n",
        "    resize_ratio = 1.0 * self.INPUT_SIZE / max(width, height)\n",
        "    target_size = (int(resize_ratio * width), int(resize_ratio * height))\n",
        "    resized_image = image.convert('RGB').resize(target_size, Image.ANTIALIAS)\n",
        "    batch_seg_map = self.sess.run(\n",
        "        self.OUTPUT_TENSOR_NAME,\n",
        "        feed_dict={self.INPUT_TENSOR_NAME: [np.asarray(resized_image)]})\n",
        "    seg_map = batch_seg_map[0]\n",
        "    return resized_image, seg_map\n",
        "\n",
        "\n",
        "def create_pascal_label_colormap():\n",
        "  \"\"\"Creates a label colormap used in PASCAL VOC segmentation benchmark.\n",
        "\n",
        "  Returns:\n",
        "    A Colormap for visualizing segmentation results.\n",
        "  \"\"\"\n",
        "  colormap = np.zeros((256, 3), dtype=int)\n",
        "  ind = np.arange(256, dtype=int)\n",
        "\n",
        "  for shift in reversed(range(8)):\n",
        "    for channel in range(3):\n",
        "      colormap[:, channel] |= ((ind >> channel) & 1) << shift\n",
        "    ind >>= 3\n",
        "\n",
        "  return colormap\n",
        "\n",
        "\n",
        "def label_to_color_image(label):\n",
        "  \"\"\"Adds color defined by the dataset colormap to the label.\n",
        "\n",
        "  Args:\n",
        "    label: A 2D array with integer type, storing the segmentation label.\n",
        "\n",
        "  Returns:\n",
        "    result: A 2D array with floating type. The element of the array\n",
        "      is the color indexed by the corresponding element in the input label\n",
        "      to the PASCAL color map.\n",
        "\n",
        "  Raises:\n",
        "    ValueError: If label is not of rank 2 or its value is larger than color\n",
        "      map maximum entry.\n",
        "  \"\"\"\n",
        "  if label.ndim != 2:\n",
        "    raise ValueError('Expect 2-D input label')\n",
        "\n",
        "  colormap = create_pascal_label_colormap()\n",
        "\n",
        "  if np.max(label) >= len(colormap):\n",
        "    raise ValueError('label value too large.')\n",
        "\n",
        "  return colormap[label]\n",
        "\n",
        "\n",
        "def vis_segmentation(image, seg_map):\n",
        "  \"\"\"Visualizes input image, segmentation map and overlay view.\"\"\"\n",
        "  plt.figure(figsize=(15, 5))\n",
        "  grid_spec = gridspec.GridSpec(1, 4, width_ratios=[6, 6, 6, 1])\n",
        "\n",
        "  plt.subplot(grid_spec[0])\n",
        "  plt.imshow(image)\n",
        "  plt.axis('off')\n",
        "  plt.title('input image')\n",
        "\n",
        "  plt.subplot(grid_spec[1])\n",
        "  seg_image = label_to_color_image(seg_map).astype(np.uint8)\n",
        "  plt.imshow(seg_image)\n",
        "  plt.axis('off')\n",
        "  plt.title('segmentation map')\n",
        "\n",
        "  plt.subplot(grid_spec[2])\n",
        "  plt.imshow(image)\n",
        "  plt.imshow(seg_image, alpha=0.7)\n",
        "  plt.axis('off')\n",
        "  plt.title('segmentation overlay')\n",
        "\n",
        "  unique_labels = np.unique(seg_map)\n",
        "  ax = plt.subplot(grid_spec[3])\n",
        "  plt.imshow(\n",
        "      FULL_COLOR_MAP[unique_labels].astype(np.uint8), interpolation='nearest')\n",
        "  ax.yaxis.tick_right()\n",
        "  plt.yticks(range(len(unique_labels)), LABEL_NAMES[unique_labels])\n",
        "  plt.xticks([], [])\n",
        "  ax.tick_params(width=0.0)\n",
        "  plt.grid('off')\n",
        "  plt.show()\n",
        "\n",
        "\n",
        "LABEL_NAMES = np.asarray([\n",
        "    'background', 'leaf'\n",
        "])\n",
        "\n",
        "FULL_LABEL_MAP = np.arange(len(LABEL_NAMES)).reshape(len(LABEL_NAMES), 1)\n",
        "FULL_COLOR_MAP = label_to_color_image(FULL_LABEL_MAP)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "00ed8a0e-e528-4f47-ebbf-5eaef6b9c8a8",
        "id": "0Ui0R4DNoeY5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "%cd"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/root\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "outputId": "b9d013eb-3810-4623-a854-b824b6fd894e",
        "id": "2rvchewBodtV",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "ls"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[01;34mmodels\u001b[0m/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fMeTULOJ77u9",
        "colab_type": "code",
        "outputId": "c7070018-c336-47f3-b38e-030d0a16a084",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "%cd models/research/deeplab/datasets/pascal_voc_seg/init_models"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/root/models/research/deeplab/datasets/pascal_voc_seg/init_models\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mKdKhqb18fSb",
        "colab_type": "code",
        "outputId": "b3b02bbc-ce20-4221-f944-65dc11130257",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "ls"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[37;44mdeeplabv3_mnv2_pascal_train_aug\u001b[0m/\n",
            "deeplabv3_mnv2_pascal_train_aug_2018_01_29.tar.gz\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tGMYH06h8nwR",
        "colab_type": "code",
        "outputId": "791e2c35-6c5c-44a0-f3bd-5a1b409de277",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "%cd"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/root\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Gl6vsgJ689Zu",
        "colab_type": "code",
        "outputId": "685a6ea6-cb84-46ec-8296-359d809ddbbe",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "ls"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[0m\u001b[01;34mmodels\u001b[0m/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TQuc2DcK9HmH",
        "colab_type": "code",
        "outputId": "7709da4d-6574-4947-975d-2cf2e3ba30e1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "%cd models"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/root/models\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3deGw3Mp9uFt",
        "colab_type": "code",
        "outputId": "fae95694-e021-43e2-e4ca-ff2a106105d1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "ls"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "AUTHORS     CONTRIBUTING.md    LICENSE    README.md  \u001b[0m\u001b[01;34msamples\u001b[0m/    WORKSPACE\n",
            "CODEOWNERS  ISSUE_TEMPLATE.md  \u001b[01;34mofficial\u001b[0m/  \u001b[01;34mresearch\u001b[0m/  \u001b[01;34mtutorials\u001b[0m/\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aLMFoSAkbtjw",
        "colab_type": "code",
        "outputId": "e93a077f-c76a-49ef-db14-38016f505f44",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "import os\n",
        "import tarfile\n",
        "\n",
        "_MODEL_NAME = 'frozen_inference_graph.pb'\n",
        "_TARBALL_NAME = 'deeplabv3_mnv2_pascal_train_aug_2018_01_29.tar.gz'\n",
        "\n",
        "\n",
        "_dl_dir ='research/deeplab/datasets/pascal_voc_seg/init_models/'\n",
        "\n",
        "\n",
        "model_path = os.path.join(model_dir, _MODEL_NAME)\n",
        "download_path = os.path.join(_dl_dir, _TARBALL_NAME)\n",
        "with tarfile.open(download_path, \"w:gz\") as tar:\n",
        "        tar.add(model_path)\n",
        "\n",
        "MODEL = DeepLabModel(download_path)\n",
        "print('model loaded successfully!')"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "model loaded successfully!\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jbEOOPuHUlo3",
        "colab_type": "code",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 368
        },
        "outputId": "4e3a9498-7487-4b7f-f9f9-d2da81fd0de9"
      },
      "source": [
        "from google.colab import files\n",
        "from os import path\n",
        "from PIL import Image\n",
        "\n",
        "uploaded = files.upload()\n",
        "  \n",
        "for name, data in uploaded.items():\n",
        "  with open('img.jpg', 'wb') as f:\n",
        "    f.write(data)\n",
        "    f.close()\n",
        "    print('saved file ' + name)\n",
        "    \n",
        "im = Image.open(name)\n",
        "resized_im, seg_map = MODEL.run(im)\n",
        "\n",
        "#plt.imshow(seg_map)\n",
        "\n",
        "vis_segmentation(resized_im, seg_map)\n"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-bc0c45da-930a-4802-ba20-404eb021a9b4\" name=\"files[]\" multiple disabled />\n",
              "     <output id=\"result-bc0c45da-930a-4802-ba20-404eb021a9b4\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving re500500leaf1235.jpg to re500500leaf1235 (1).jpg\n",
            "saved file re500500leaf1235.jpg\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA68AAAEVCAYAAAAPVtPTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
            "text/plain": [
              "<Figure size 1080x360 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t0kiiY9gBao9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}