QCANet.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "QCANet.ipynb",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Timokleia/QCANet/blob/master/QCANet.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RtKmLg-gZ9BJ"
      },
      "source": [
        "QCANet \n",
        "\n",
        "Note: not possible to run all at once\n",
        "1. the configuration files need to be changed manually once\n",
        "2. everytime you run the model you need to change the paths in the results part "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_IuWdGnyvtmB"
      },
      "source": [
        "Setup "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ISM0pd2vb_mn",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "59fe2900-fd4d-42df-e28e-eb6e93314103"
      },
      "source": [
        "# Mount drive:\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Bs_KoUErT-Qp"
      },
      "source": [
        "Installations and Importations\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vb3bZeNmcfrH",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0e24973b-acac-410b-8ff1-6aa59fdd9c4c"
      },
      "source": [
        "import skimage\n",
        "import matplotlib\n",
        "import numpy\n",
        "import os\n",
        "from shutil import copyfile\n",
        "\n",
        "!pip install 'cupy-cuda101>=7.7.0,<8.0.0'\n",
        "!pip install chainer\n",
        "import chainer\n",
        "\n",
        "from skimage import io\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from tifffile import imwrite\n",
        "import pandas as pd\n",
        "\n",
        "import tensorflow as tf\n",
        "import numpy as np"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting cupy-cuda101<8.0.0,>=7.7.0\n",
            "  Downloading cupy_cuda101-7.8.0-cp37-cp37m-manylinux1_x86_64.whl (370.3 MB)\n",
            "\u001b[K     |████████████████████████████████| 370.3 MB 34 kB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy>=1.9.0 in /usr/local/lib/python3.7/dist-packages (from cupy-cuda101<8.0.0,>=7.7.0) (1.19.5)\n",
            "Requirement already satisfied: fastrlock>=0.3 in /usr/local/lib/python3.7/dist-packages (from cupy-cuda101<8.0.0,>=7.7.0) (0.6)\n",
            "Requirement already satisfied: six>=1.9.0 in /usr/local/lib/python3.7/dist-packages (from cupy-cuda101<8.0.0,>=7.7.0) (1.15.0)\n",
            "Installing collected packages: cupy-cuda101\n",
            "  Attempting uninstall: cupy-cuda101\n",
            "    Found existing installation: cupy-cuda101 9.1.0\n",
            "    Uninstalling cupy-cuda101-9.1.0:\n",
            "      Successfully uninstalled cupy-cuda101-9.1.0\n",
            "Successfully installed cupy-cuda101-7.8.0\n",
            "Collecting chainer\n",
            "  Downloading chainer-7.8.0.tar.gz (1.0 MB)\n",
            "\u001b[K     |████████████████████████████████| 1.0 MB 7.3 MB/s \n",
            "\u001b[?25hRequirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from chainer) (57.2.0)\n",
            "Requirement already satisfied: typing_extensions in /usr/local/lib/python3.7/dist-packages (from chainer) (3.7.4.3)\n",
            "Requirement already satisfied: filelock in /usr/local/lib/python3.7/dist-packages (from chainer) (3.0.12)\n",
            "Requirement already satisfied: numpy>=1.9.0 in /usr/local/lib/python3.7/dist-packages (from chainer) (1.19.5)\n",
            "Requirement already satisfied: protobuf>=3.0.0 in /usr/local/lib/python3.7/dist-packages (from chainer) (3.17.3)\n",
            "Requirement already satisfied: six>=1.9.0 in /usr/local/lib/python3.7/dist-packages (from chainer) (1.15.0)\n",
            "Building wheels for collected packages: chainer\n",
            "  Building wheel for chainer (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for chainer: filename=chainer-7.8.0-py3-none-any.whl size=967734 sha256=e5781618d2106b720b90045a8eea1148f599d0e00db1f1b21377db80b8687da2\n",
            "  Stored in directory: /root/.cache/pip/wheels/8d/45/4c/c961846626d4db2511e4acbd5056eefb2aeedcec8208a3ca42\n",
            "Successfully built chainer\n",
            "Installing collected packages: chainer\n",
            "Successfully installed chainer-7.8.0\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/chainer/_environment_check.py:75: UserWarning: \n",
            "--------------------------------------------------------------------------------\n",
            "CuPy (cupy-cuda101) version 9.1.0 may not be compatible with this version of Chainer.\n",
            "Please consider installing the supported version by running:\n",
            "  $ pip install 'cupy-cuda101>=7.7.0,<8.0.0'\n",
            "\n",
            "See the following page for more details:\n",
            "  https://docs.cupy.dev/en/latest/install.html\n",
            "--------------------------------------------------------------------------------\n",
            "\n",
            "  requirement=requirement, help=help))\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wTcCsEcwctCT",
        "outputId": "477e5a5a-c077-4650-9f45-f9536c3e1bbe"
      },
      "source": [
        "!python --version"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Python 3.7.11\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "47i_GpH2UY_w",
        "outputId": "fa4d500c-b89f-4a4c-e717-d204c9d95c59"
      },
      "source": [
        "print(tf.test.gpu_device_name())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/device:GPU:0\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_7qWovyLUHbG"
      },
      "source": [
        "Clone the code from github\n",
        "(should be done once)"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f8rEytjndDXv",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "8c5a2405-015a-4ad1-b8b4-07b4b1793588"
      },
      "source": [
        "os.chdir(\"/content/drive/My Drive/\")\n",
        "\n",
        "!if [ -d '/content/drive/My Drive/QCANet' ]; then echo \"Directory QCANet already exists\"; else echo \"Directory QCANet created\" && git clone https://github.com/funalab/QCANet.git; fi\n",
        "\n",
        "#NDN model\n",
        "if os.path.exists(\"/content/drive/MyDrive/QCANet/models/learned_ndn.npz\")==False:\n",
        "  !curl -o QCANet/models/learned_ndn.npz https://fun.bio.keio.ac.jp/software/QCANet/learned_ndn.npz\n",
        "else: print(\"model learned_ndn already exists\")\n",
        "\n",
        "#NSN model\n",
        "if os.path.exists(\"/content/drive/MyDrive/QCANet/models/learned_nsn.npz\")==False:\n",
        "  !curl -o QCANet/models/learned_nsn.npz https://fun.bio.keio.ac.jp/software/QCANet/learned_nsn.npz\n",
        "else: print(\"model learned_nsn already exists\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Directory QCANet already exists\n",
            "model learned_ndn already exists\n",
            "model learned_nsn already exists\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Sekyp4eradCd"
      },
      "source": [
        "Definition:\n",
        "Show_layers to check the input and output"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HxIngXlUdTr3"
      },
      "source": [
        "def show_layers(data):\n",
        "  for i in range(data.shape[0]):\n",
        "    print(\"Layer\", i)\n",
        "    io.imshow(data[i])\n",
        "    io.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QCd-dieaUnb1"
      },
      "source": [
        "\n",
        "\n",
        "Set the paths to the data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9g1EjQmEdW2M"
      },
      "source": [
        "# Folder path:\n",
        "folder_path = \"/content/drive/MyDrive/QCANet\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Uo3zmh1BU0vT"
      },
      "source": [
        "#First sample\n",
        "train_raw = \"/content/drive/MyDrive/Annotation/Brain1-Sample1/Brain1-Sample1-Raw.tif\" # first train volume, raw data\n",
        "train_lab = \"/content/drive/MyDrive/Annotation/Brain1-Sample1/Brain1-Sample1-labels.tif\" # first train volume, labels\n",
        "\n",
        "#Second sample\n",
        "train2_raw = \"/content/drive/MyDrive/Annotation/Brain1-Sample3/Brain1-Sample3-Raw.tif\" # second train volume, raw data\n",
        "train2_lab = \"/content/drive/MyDrive/Annotation/Brain1-Sample3/Brain1-Sample3-labels.tif\" # second train volume, labels\n",
        "\n",
        "#Validation sample\n",
        "test_raw = \"/content/drive/MyDrive/Annotation/Brain1-Sample2/Brain1-Sample2-Raw.tif\" # test volume, raw data\n",
        "test_lab = \"/content/drive/MyDrive/Annotation/Brain1-Sample2/Brain1-Sample2-labels.tif\" # test volume, labels"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CUi9BSCGVSlW"
      },
      "source": [
        "Create the configuration files"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KHuxGBq8da-N",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3fae3076-6322-4806-e001-588e61bba02f"
      },
      "source": [
        "orig_cfg = folder_path+\"/confs/example_train_nsn.cfg\" \n",
        "dest_cfg = folder_path+\"/confs/CFOS_nsn.cfg\"\n",
        "\n",
        "if os.path.exists(dest_cfg):\n",
        "  print(\"cfg nsn file already exists\")\n",
        "elif os.path.exists(dest_cfg)==False:\n",
        "  copyfile(orig_cfg, dest_cfg)\n",
        "\n",
        "\n",
        "orig_cfg = folder_path+\"/confs/example_train_ndn.cfg\" \n",
        "dest_cfg = folder_path+\"/confs/CFOS_ndn.cfg\"\n",
        "\n",
        "if os.path.exists(dest_cfg):\n",
        "  print(\"cfg ndn file already exists\")\n",
        "elif os.path.exists(dest_cfg)==False:\n",
        "  copyfile(orig_cfg, dest_cfg)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "cfg nsn file already exists\n",
            "cfg ndn file already exists\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9qZwhkTElf7g"
      },
      "source": [
        "After creating the configuration files you need to manually change:\n",
        "\n",
        "[Dataset]\n",
        "\n",
        "root_path = your_dataset\n",
        "\n",
        "split_list_train = your_dataset/split_list/train.txt\n",
        "\n",
        "split_list_validation = your_dataset/split_list/validation.txt\n",
        "\n",
        "input_format = tif\n",
        "\n",
        "image_dtype = uint16\n",
        "\n",
        "resolution = (1.0, 1.0, 1.0)\n",
        "\n",
        "[Runtime]\n",
        "\n",
        "save_dir = results/result_train_nsn\n",
        "\n",
        "batchsize = 1\n",
        "\n",
        "val_batchsize = 1\n",
        "\n",
        "epoch = 400\n",
        "\n",
        "gpu = 0\n",
        "\n",
        "patch_size = (64, 64, 64)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uRZzDrWkVZGp"
      },
      "source": [
        "Prepare the right directories in the needed structure"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "osIkKD5sVs3-"
      },
      "source": [
        "# Needed structure should be:\n",
        "\n",
        "# your_dataset/\n",
        "#        +-- images_raw/  (3D fluorescence microscopic image)\n",
        "#        |           +-- image1.tif\n",
        "#        |           +-- image2.tif\n",
        "#        |           +-- image3.tif\n",
        "#        +-- images_nsn/  (the ground truth of segmentation)\n",
        "#        |           +-- image1.tif\n",
        "#        |           +-- image2.tif\n",
        "#        |           +-- image3.tif\n",
        "#        +-- images_ndn/  (the ground truth of detection)\n",
        "#        |           +-- image1.tif\n",
        "#        |           +-- image2.tif\n",
        "#        |           +-- image3.tif"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j79wFBX_Vl36",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 351
        },
        "outputId": "1bd1b52f-5b6c-4a08-e618-03c91d959832"
      },
      "source": [
        "os.chdir(folder_path)\n",
        "\n",
        "os.makedirs(folder_path + \"/your_dataset/\")\n",
        "os.makedirs(folder_path + \"/your_dataset/images_raw/\")\n",
        "os.makedirs(folder_path + \"/your_dataset/images_nsn/\")\n",
        "os.makedirs(folder_path + \"/your_dataset/images_ndn/\")\n",
        "os.makedirs(folder_path + \"/your_dataset/input_gt/\")\n",
        "os.makedirs(folder_path + \"/your_dataset/input/\")\n",
        "os.makedirs(folder_path + \"/your_dataset/split_list/\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "FileExistsError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mFileExistsError\u001b[0m                           Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-11-bb07cf07f985>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mchdir\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfolder_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m \u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmakedirs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfolder_path\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m\"/your_dataset/\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m \u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmakedirs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfolder_path\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m\"/your_dataset/images_raw/\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmakedirs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfolder_path\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m\"/your_dataset/images_nsn/\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/lib/python3.7/os.py\u001b[0m in \u001b[0;36mmakedirs\u001b[0;34m(name, mode, exist_ok)\u001b[0m\n\u001b[1;32m    221\u001b[0m             \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    222\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 223\u001b[0;31m         \u001b[0mmkdir\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmode\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    224\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0mOSError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    225\u001b[0m         \u001b[0;31m# Cannot rely on checking for EEXIST, since the operating system\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mFileExistsError\u001b[0m: [Errno 17] File exists: '/content/drive/MyDrive/QCANet/your_dataset/'"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ANLz1s5JY_AF"
      },
      "source": [
        "**Prepare** **and save the training and validation datasets**\n",
        "\n",
        "Convert data into 'uint16' /\n",
        "Crop the data /\n",
        "Save in the correct structure /\n",
        "Check if correct"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LG-i-GH1fWDF"
      },
      "source": [
        "#first raw data sample\n",
        "first_sample=io.imread(train_raw)\n",
        "raw1=first_sample.astype(\"uint16\")\n",
        "raw1=raw1[300:364,15:79,15:79]\n",
        "imwrite(folder_path + '/your_dataset/images_raw/Image1.tif', raw1, dtype=raw1.dtype)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "d2vRenqdbL1i",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "09d44c4b-7ae3-41f4-e52b-c4145d7daa84"
      },
      "source": [
        "raw1.shape\n",
        "raw1.mean(), raw1.min(), raw1.max()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(346.88773345947266, 222, 2212)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v2sXgAB1bEUI"
      },
      "source": [
        "show_layers(raw1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YOeddphqgZpU"
      },
      "source": [
        "#first labels for nsn model\n",
        "labels1=io.imread(train_lab)\n",
        "nsn_labels1=labels1.astype(\"uint16\")\n",
        "nsn_labels1=nsn_labels1[300:364,15:79,15:79]\n",
        "imwrite(folder_path + '/your_dataset/images_nsn/Image1.tif', nsn_labels1, dtype=nsn_labels1.dtype)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mg95E7EngcLz",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "52dd5783-828b-4f6a-8646-3ce0c338964d"
      },
      "source": [
        "nsn_labels1.shape\n",
        "nsn_labels1.mean(), nsn_labels1.max(), nsn_labels1.min()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0.15538787841796875, 149, 0)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GNWX36R_a8hb"
      },
      "source": [
        "show_layers(nsn_labels1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LF6kwXE_gkIP"
      },
      "source": [
        "#first labels for ndn model\n",
        "labels1=io.imread(train_lab)\n",
        "ndn_labels1=labels1.astype(\"uint16\")\n",
        "ndn_labels1=ndn_labels1[300:364,15:79,15:79]\n",
        "imwrite(folder_path + '/your_dataset/images_ndn/Image1.tif', ndn_labels1, dtype=ndn_labels1.dtype)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wx5iBJ7xf7qM"
      },
      "source": [
        "ndn.shape\n",
        "ndn_labels1.mean(), ndn_labels1.max(), ndn_labels1.min()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FaG_x4daceOW"
      },
      "source": [
        "show_layers(ndn_labels1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q3M_i4oBgu9W"
      },
      "source": [
        "#second raw data sample\n",
        "second_sample=io.imread(train2_raw)\n",
        "raw2=second_sample.astype(\"uint16\")\n",
        "raw2=raw2[300:364,15:79,15:79]\n",
        "imwrite(folder_path + '/your_dataset/images_raw/Image2.tif', raw2, dtype=raw2.dtype)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1uaCibrngAvd"
      },
      "source": [
        "raw2.shape\n",
        "raw2.mean(), raw2.min(), raw2.max()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zarlgzuacydx"
      },
      "source": [
        "show_layers(raw2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xLH-yjwhgw3F"
      },
      "source": [
        "#second labels for nsn model\n",
        "labels2=io.imread(train2_lab)\n",
        "nsn_labels2=labels2.astype(\"uint16\")\n",
        "nsn_labels2=nsn_labels2[300:364,15:79,15:79]\n",
        "imwrite(folder_path + '/your_dataset/images_nsn/Image2.tif', nsn_labels2, dtype=nsn_labels2.dtype)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oDeIoJa5gC-Y"
      },
      "source": [
        "nsn_labels2.shape\n",
        "nsn_labels2.mean(), nsn_labels2.max(), nsn_labels2.min()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OtNedydNdPeC"
      },
      "source": [
        "show_layers(nsn_labels2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nSoXF0eHg0_0"
      },
      "source": [
        "#second labels for ndn model\n",
        "labels2=io.imread(train2_lab)\n",
        "ndn_labels2=labels2.astype(\"uint16\")\n",
        "ndn_labels2=ndn_labels2[300:364,15:79,15:79]\n",
        "imwrite(folder_path + '/your_dataset/images_ndn/Image2.tif', ndn_labels2, dtype=ndn_labels2.dtype)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LLscUeTjd4N9"
      },
      "source": [
        "ndn_labels2.shape\n",
        "ndn_labels2.mean(), ndn_labels2.max(), ndn_labels2.min()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0B7lnGGrd_Ig"
      },
      "source": [
        "show_layers(ndn_labels2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZZx7EEKug4gY"
      },
      "source": [
        "#validation raw data\n",
        "validation=io.imread(test_raw)\n",
        "testing_data=validation.astype(\"uint16\")\n",
        "testing_data=testing_data[300:364,15:79,15:79]\n",
        "imwrite(folder_path+'/your_dataset/input/CFOS_noisy2.tif', testing_data, dtype=testing_data.dtype)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nLLHM8b_enZr"
      },
      "source": [
        "testing_data.shape\n",
        "testing_data.mean(), testing_data.max(), testing_data.min()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z40et1DeetHC"
      },
      "source": [
        "show_layers(testing_data)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "22WHsfs4g8DG"
      },
      "source": [
        "#validation labels\n",
        "validation_labels=io.imread(test_lab)\n",
        "testing_labels=validation_labels.astype(\"uint16\")\n",
        "testing_labels=testing_labels[300:364,15:79,15:79]\n",
        "imwrite(folder_path + '/your_dataset/input_gt/CFOS_noisy2.tif', testing_labels, dtype=testing_labels.dtype)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LtPU3uLSesAC"
      },
      "source": [
        "testing_labels.shape\n",
        "testing_labels.mean(), testing_labels.max(), testing_labels.min()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KFATQpwRlVVd"
      },
      "source": [
        "show_layers(testing_labels)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8V3FCLOxZKyS"
      },
      "source": [
        "Create the split_list files"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9tbo0U9Sg_gB",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 232
        },
        "outputId": "64ac4dcf-7861-41c3-8738-5a1fc004c6b2"
      },
      "source": [
        "f = open(folder_path+\"/your_dataset/split_list/train.txt\", \"x\")\n",
        "f.write(\"Image1.tif\")\n",
        "f.close()\n",
        "\n",
        "f = open(folder_path+\"/your_dataset/split_list/validation.txt\", \"x\")\n",
        "f.write(\"Image2.tif\")\n",
        "f.close()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "FileExistsError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mFileExistsError\u001b[0m                           Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-32-8ea4125aca46>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfolder_path\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;34m\"/your_dataset/split_list/train.txt\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"x\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwrite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"Image1.tif\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclose\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0mf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfolder_path\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;34m\"/your_dataset/split_list/validation.txt\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"x\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mFileExistsError\u001b[0m: [Errno 17] File exists: '/content/drive/MyDrive/QCANet/your_dataset/split_list/train.txt'"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oFYOZ1l9ZRY4"
      },
      "source": [
        "Training 🏋️"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5YFHdz8XhFZH",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7d27a1a8-97a1-4e0f-9285-2ff2127ecd36"
      },
      "source": [
        "os.chdir(folder_path)\n",
        "!python src/tools/train.py --conf_file confs/CFOS_nsn.cfg --gpu_id 0"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Output Directory not exist! Create...\n",
            "Output Directory: ./results/result_train_nsn_Aug2Mon_2021_134931\n",
            "============================\n",
            "[Dataset]\n",
            "root_path = your_dataset\n",
            "input_format = tif\n",
            "image_dtype = uint16\n",
            "resolution = (1.0, 1.0, 2.18)\n",
            "split_list_train = your_dataset/split_list/train.txt\n",
            "split_list_validation = your_dataset/split_list/validation.txt\n",
            "\n",
            "[Model]\n",
            "model = NSN\n",
            "ndim = 3\n",
            "lossfun = F.softmax_cross_entropy\n",
            "ch_in = 12\n",
            "ch_base = 16\n",
            "ch_out = 12\n",
            "\n",
            "[Runtime]\n",
            "save_dir = results/result_train_nsn\n",
            "batchsize = 1\n",
            "val_batchsize = 1\n",
            "epoch = 400\n",
            "optimizer = Adam\n",
            "init_lr = 0.002\n",
            "weight_decay = 0.0001\n",
            "gpu = 0\n",
            "patch_size = (64, 64, 64)\n",
            "normalization = 0\n",
            "augmentation = 0\n",
            "class_weight = (1, 1)\n",
            "scaling = 1\n",
            "============================\n",
            "\n",
            "Loading datasets...\n",
            "-- train_dataset.size = 1\n",
            "-- validation_dataset.size = 1\n",
            "Initializing models...\n",
            "[epoch 1]\n",
            "train mean loss=2.6861534118652344\n",
            "train accuracy=0.7531356811523438, train recall=0.4939550949913644\n",
            "train precision=0.004419923655864126, specificity=0.7537094030164586\n",
            "train F-measure=0.008761449621664676, IoU=0.0044\n",
            "test mean loss=12.025519609451294\n",
            "test accuracy=0.9306306907598921, recall=0.13685123329907503\n",
            "test precision=0.25241395651916354, specificity=0.9765518163981212\n",
            "test F-measure=0.17747880958827084, IoU=0.09738093061522991\n",
            "[epoch 2]\n",
            "train mean loss=2.522644519805908\n",
            "train accuracy=0.8222999572753906, train recall=1.0\n",
            "train precision=0.003721367923519473, specificity=0.8221819292285376\n",
            "train F-measure=0.00741514137776736, IoU=0.003721367923519473\n",
            "test mean loss=11.29642105102539\n",
            "test accuracy=0.9340223133992805, recall=0.14558710174717368\n",
            "test precision=0.29256486381825225, specificity=0.9796342677923777\n",
            "test F-measure=0.19442419043534206, IoU=0.10767988217687721\n",
            "[epoch 3]\n",
            "train mean loss=2.3901243209838867\n",
            "train accuracy=0.8458595275878906, train recall=0.7631578947368421\n",
            "train precision=0.0007173423701981349, specificity=0.8458715176302717\n",
            "train F-measure=0.0014333374521191152, IoU=0.0007171827084775942\n",
            "test mean loss=10.840156555175781\n",
            "test accuracy=0.9349145683453237, recall=0.14269655704008222\n",
            "test precision=0.3000810482236931, specificity=0.9807453623877758\n",
            "test F-measure=0.19341778764529188, IoU=0.10706282079086243\n",
            "[epoch 4]\n",
            "train mean loss=2.19427490234375\n",
            "train accuracy=0.8863601684570312, train recall=1.0\n",
            "train precision=0.010331882661705592, specificity=0.8862251893382423\n",
            "train F-measure=0.020452452979087203, IoU=0.010331882661705592\n",
            "test mean loss=10.304790019989014\n",
            "test accuracy=0.9349128119379496, recall=0.14019141829393628\n",
            "test precision=0.29793188178281343, specificity=0.9808884297520661\n",
            "test F-measure=0.19066547273243495, IoU=0.10537878422094539\n",
            "[epoch 5]\n",
            "train mean loss=2.0042479038238525\n",
            "train accuracy=0.9193687438964844, train recall=0.9956709956709957\n",
            "train precision=0.03161656891495601, specificity=0.9191664977376258\n",
            "train F-measure=0.06128702757916241, IoU=0.03161222339304531\n",
            "test mean loss=9.7584890127182\n",
            "test accuracy=0.9348056710881295, recall=0.13630524152106885\n",
            "test precision=0.29329647546648235, specificity=0.9809999108151495\n",
            "test F-measure=0.1861158619479893, IoU=0.10260625695082443\n",
            "[epoch 6]\n",
            "train mean loss=1.8618531227111816\n",
            "train accuracy=0.9485015869140625, train recall=0.9904397705544933\n",
            "train precision=0.07132039102299326, specificity=0.9483335758987047\n",
            "train F-measure=0.13305933727202673, IoU=0.07127132636213539\n",
            "test mean loss=9.35281229019165\n",
            "test accuracy=0.9346897482014388, recall=0.13196942446043167\n",
            "test precision=0.28802747791952893, specificity=0.9811281140376955\n",
            "test F-measure=0.18100524205982116, IoU=0.09950839125275471\n",
            "[epoch 7]\n",
            "train mean loss=2.063690662384033\n",
            "train accuracy=0.8672523498535156, train recall=1.0\n",
            "train precision=0.00246524294109216, specificity=0.8672087858413023\n",
            "train F-measure=0.004918360927625747, IoU=0.00246524294109216\n",
            "test mean loss=9.098035097122192\n",
            "test accuracy=0.9353606958183454, recall=0.1281795991778006\n",
            "test precision=0.2924237983587339, specificity=0.982057122896724\n",
            "test F-measure=0.17823329760628795, IoU=0.09783541293849435\n",
            "[epoch 8]\n",
            "train mean loss=1.8569384813308716\n",
            "train accuracy=0.9179916381835938, train recall=0.9924812030075187\n",
            "train precision=0.01213235294117647, specificity=0.9179159761415621\n",
            "train F-measure=0.023971669844728954, IoU=0.01213123793768955\n",
            "test mean loss=8.906283378601074\n",
            "test accuracy=0.9363952197616906, recall=0.12686279547790338\n",
            "test precision=0.3043846805887339, specificity=0.9832276740590998\n",
            "test F-measure=0.17908553034252941, IoU=0.09834922690038095\n",
            "[epoch 9]\n",
            "train mean loss=1.686119794845581\n",
            "train accuracy=0.9598617553710938, train recall=0.9617590822179732\n",
            "train precision=0.08756963788300835, specificity=0.9598541543788156\n",
            "train F-measure=0.16052337641614806, IoU=0.08726578764746704\n",
            "test mean loss=8.676988244056702\n",
            "test accuracy=0.9371118339703237, recall=0.12381166495375129\n",
            "test precision=0.3114144922853219, specificity=0.9841622569712825\n",
            "test F-measure=0.17718028266115135, IoU=0.09720121028744327\n",
            "[epoch 10]\n",
            "train mean loss=1.7023696899414062\n",
            "train accuracy=0.9565048217773438, train recall=0.9852941176470589\n",
            "train precision=0.01732758620689655, specificity=0.9564824005497442\n",
            "train F-measure=0.03405625211792613, IoU=0.017323106093251744\n",
            "test mean loss=8.453057289123535\n",
            "test accuracy=0.9380040889163669, recall=0.11925102774922919\n",
            "test precision=0.320445326659187, specificity=0.9853699684880195\n",
            "test F-measure=0.17381738173817382, IoU=0.09518072289156626\n",
            "[epoch 11]\n",
            "train mean loss=1.541485071182251\n",
            "train accuracy=0.9780807495117188, train recall=0.8066893424036281\n",
            "train precision=0.208406561218512, specificity=0.9792418772563177\n",
            "train F-measure=0.33123836126629425, IoU=0.19849351373971266\n",
            "test mean loss=8.212957262992859\n",
            "test accuracy=0.9388225747526978, recall=0.11664953751284686\n",
            "test precision=0.3314171001003741, specificity=0.9863863041797967\n",
            "test F-measure=0.17256206200261312, IoU=0.09442841172035463\n",
            "[epoch 12]\n",
            "train mean loss=1.63546621799469\n",
            "train accuracy=0.9607276916503906, train recall=0.9736842105263158\n",
            "train precision=0.0035814538766818314, specificity=0.9607258132206054\n",
            "train F-measure=0.007136657344006173, IoU=0.0035811072396438252\n",
            "test mean loss=8.037197828292847\n",
            "test accuracy=0.9395479709982014, recall=0.11260277492291881\n",
            "test precision=0.3405867495628521, specificity=0.9873877757298294\n",
            "test F-measure=0.1692493362297852, IoU=0.09244805400274232\n",
            "[epoch 13]\n",
            "train mean loss=1.445078730583191\n",
            "train accuracy=0.9854278564453125, train recall=0.9821073558648111\n",
            "train precision=0.11475029036004646, specificity=0.985434240046476\n",
            "train F-measure=0.20549084858569053, IoU=0.11451089476124246\n",
            "test mean loss=7.865413308143616\n",
            "test accuracy=0.9399519446942446, recall=0.10974434737923947\n",
            "test precision=0.34564029941331176, specificity=0.9879804833818895\n",
            "test F-measure=0.1665935351762469, IoU=0.09086557638611886\n",
            "[epoch 14]\n",
            "train mean loss=1.4053047895431519\n",
            "train accuracy=0.9850387573242188, train recall=0.8946395563770795\n",
            "train precision=0.11128995171303747, specificity=0.9852257045981888\n",
            "train F-measure=0.19795501022494888, IoU=0.10985020426690877\n",
            "test mean loss=7.691778779029846\n",
            "test accuracy=0.9401592007643885, recall=0.10775308324768756\n",
            "test precision=0.3478846951472418, specificity=0.9883149265711397\n",
            "test F-measure=0.16454144188327613, IoU=0.08964595858383434\n",
            "[epoch 15]\n",
            "train mean loss=1.513547420501709\n",
            "train accuracy=0.9516754150390625, train recall=0.9924242424242424\n",
            "train precision=0.010235974370995469, specificity=0.9516548860357541\n",
            "train F-measure=0.020262954369682912, IoU=0.010235174623017424\n",
            "test mean loss=7.530701279640198\n",
            "test accuracy=0.9404560336106115, recall=0.1048625385405961\n",
            "test precision=0.35126412049488975, specificity=0.9887961531601165\n",
            "test F-measure=0.16150973263090201, IoU=0.0878491094010655\n",
            "[epoch 16]\n",
            "train mean loss=1.3460969924926758\n",
            "train accuracy=0.9835090637207031, train recall=0.8392694063926941\n",
            "train precision=0.18140544808527437, specificity=0.984114093522672\n",
            "train F-measure=0.2983281934750852, IoU=0.17531476535673407\n",
            "test mean loss=7.3570908308029175\n",
            "test accuracy=0.9409267507868705, recall=0.10303186022610483\n",
            "test precision=0.35992370694491194, specificity=0.989400008918485\n",
            "test F-measure=0.16020375040575294, IoU=0.08707689802122635\n",
            "[epoch 17]\n",
            "train mean loss=1.3218371868133545\n",
            "train accuracy=0.9825668334960938, train recall=0.8317610062893082\n",
            "train precision=0.1954192833394902, specificity=0.9833021558465455\n",
            "train F-measure=0.3164822016153156, IoU=0.18798862828713575\n",
            "test mean loss=7.168601632118225\n",
            "test accuracy=0.941494070368705, recall=0.10094424460431654\n",
            "test precision=0.3715130023640662, specificity=0.9901209197930911\n",
            "test F-measure=0.15875340943529648, IoU=0.08622061284393602\n",
            "[epoch 18]\n",
            "train mean loss=1.2772854566574097\n",
            "train accuracy=0.9842720031738281, train recall=0.7582417582417582\n",
            "train precision=0.2383177570093458, specificity=0.9856138021542842\n",
            "train F-measure=0.3626526511052713, IoU=0.22148791540785498\n",
            "test mean loss=6.968733072280884\n",
            "test accuracy=0.9421597487634892, recall=0.09953108941418294\n",
            "test precision=0.3877142499687226, specificity=0.9909068612878292\n",
            "test F-measure=0.15839914130184773, IoU=0.08601165695253955\n",
            "[epoch 19]\n",
            "train mean loss=1.340597152709961\n",
            "train accuracy=0.9724884033203125, train recall=0.9916666666666667\n",
            "train precision=0.016234652114597545, specificity=0.9724796201874637\n",
            "train F-measure=0.03194630872483222, IoU=0.01623243759377984\n",
            "test mean loss=6.796721577644348\n",
            "test accuracy=0.94304673448741, recall=0.09416752312435765\n",
            "test precision=0.4098406485882024, specificity=0.9921554491943635\n",
            "test F-measure=0.15314703577957692, IoU=0.08292324226483398\n",
            "[epoch 20]\n",
            "train mean loss=1.2184247970581055\n",
            "train accuracy=0.9878807067871094, train recall=0.8324552160168599\n",
            "train precision=0.20745798319327732, specificity=0.9884454143456038\n",
            "train F-measure=0.33214210636956065, IoU=0.19914292916561632\n",
            "test mean loss=6.634369611740112\n",
            "test accuracy=0.9437545666591727, recall=0.08867548818088386\n",
            "test precision=0.4308004368856296, specificity=0.9932219513645282\n",
            "test F-measure=0.14707684112398456, IoU=0.07937557497700092\n",
            "[epoch 21]\n",
            "train mean loss=1.2708067893981934\n",
            "train accuracy=0.9791831970214844, train recall=1.0\n",
            "train precision=0.02046311254711901, specificity=0.979174140365607\n",
            "train F-measure=0.04010554089709762, IoU=0.02046311254711901\n",
            "test mean loss=6.512759327888489\n",
            "test accuracy=0.9443763348696043, recall=0.08218782117163412\n",
            "test precision=0.45284020527340296, specificity=0.9942550092157679\n",
            "test F-measure=0.13912523445782476, IoU=0.07476335164193058\n",
            "[epoch 22]\n",
            "train mean loss=1.2289855480194092\n",
            "train accuracy=0.9835014343261719, train recall=0.8963414634146342\n",
            "train precision=0.032996632996632996, specificity=0.9835559966409649\n",
            "train F-measure=0.06365014072310024, IoU=0.03287119856887299\n",
            "test mean loss=6.435103893280029\n",
            "test accuracy=0.9446468216052158, recall=0.0731628982528263\n",
            "test precision=0.46160081053698077, specificity=0.9950632469231226\n",
            "test F-measure=0.12630645116575642, IoU=0.06741041044003196\n",
            "[epoch 23]\n",
            "train mean loss=1.1935625076293945\n",
            "train accuracy=0.9876747131347656, train recall=0.9418604651162791\n",
            "train precision=0.024493498639250075, specificity=0.9876897480710377\n",
            "train F-measure=0.04774535809018567, IoU=0.024456521739130436\n",
            "test mean loss=6.372487902641296\n",
            "test accuracy=0.9446907317895683, recall=0.06539054470709146\n",
            "test precision=0.4600090375056484, specificity=0.9955593376538439\n",
            "test F-measure=0.11450424610539339, IoU=0.06072898645827119\n",
            "[epoch 24]\n",
            "train mean loss=1.112696647644043\n",
            "train accuracy=0.9915237426757812, train recall=0.5444862155388471\n",
            "train precision=0.3675972927241963, specificity=0.9942620937408846\n",
            "train F-measure=0.4388888888888889, IoU=0.28113879003558717\n",
            "test mean loss=6.2524484395980835\n",
            "test accuracy=0.9446995138264388, recall=0.06683581706063721\n",
            "test precision=0.46131678120150743, specificity=0.9954850169451216\n",
            "test F-measure=0.11675596824417202, IoU=0.06199725913126378\n",
            "[epoch 25]\n",
            "train mean loss=1.1480050086975098\n",
            "train accuracy=0.9874267578125, train recall=0.8095238095238095\n",
            "train precision=0.03023117960877297, specificity=0.9875123083146959\n",
            "train F-measure=0.05828571428571429, IoU=0.03001765744555621\n",
            "test mean loss=6.135919213294983\n",
            "test accuracy=0.9446889753821942, recall=0.06734969167523125\n",
            "test precision=0.4609804352604968, specificity=0.9954441405553244\n",
            "test F-measure=0.11752837326607818, IoU=0.062433011789924976\n",
            "[epoch 26]\n",
            "train mean loss=1.0601242780685425\n",
            "train accuracy=0.9933738708496094, train recall=0.662020905923345\n",
            "train precision=0.28273809523809523, specificity=0.9944657708308615\n",
            "train F-measure=0.3962460896767466, IoU=0.247074122236671\n",
            "test mean loss=6.001367211341858\n",
            "test accuracy=0.9446784369379496, recall=0.07097893114080164\n",
            "test precision=0.4622463919682075, specificity=0.9952230364468756\n",
            "test F-measure=0.12306150290948575, IoU=0.06556501616874832\n",
            "[epoch 27]\n",
            "train mean loss=1.0487340688705444\n",
            "train accuracy=0.9916763305664062, train recall=0.6318449873631002\n",
            "train precision=0.30060120240480964, specificity=0.9933130745678407\n",
            "train F-measure=0.40738728951656705, IoU=0.2557980900409277\n",
            "test mean loss=5.857917308807373\n",
            "test accuracy=0.9446362831609713, recall=0.07550745118191161\n",
            "test precision=0.46215844309023, specificity=0.9949164635233961\n",
            "test F-measure=0.1298070286834332, IoU=0.06940836088804912\n",
            "[epoch 28]\n",
            "train mean loss=1.0180784463882446\n",
            "train accuracy=0.9944915771484375, train recall=0.6693679092382496\n",
            "train precision=0.249848759830611, specificity=0.9952586157452195\n",
            "train F-measure=0.3638766519823789, IoU=0.22240172320947765\n",
            "test mean loss=5.723949432373047\n",
            "test accuracy=0.9445660268660072, recall=0.07807682425488181\n",
            "test precision=0.45980707395498394, specificity=0.9946935013972293\n",
            "test F-measure=0.1334870823380831, IoU=0.0715168274888209\n",
            "[epoch 29]\n",
            "train mean loss=0.9893079400062561\n",
            "train accuracy=0.993011474609375, train recall=0.6725860155382908\n",
            "train precision=0.28278114792347175, specificity=0.9941165887698427\n",
            "train F-measure=0.3981603153745072, IoU=0.24856439704675964\n",
            "test mean loss=5.600464701652527\n",
            "test accuracy=0.9446538472347122, recall=0.07997173689619733\n",
            "test precision=0.4649859943977591, specificity=0.9946767792377668\n",
            "test F-measure=0.1364720067961963, IoU=0.07323314020175878\n",
            "[epoch 30]\n",
            "train mean loss=0.9646947383880615\n",
            "train accuracy=0.9952507019042969, train recall=0.66728280961183\n",
            "train precision=0.25315568022440393, specificity=0.9959289457689705\n",
            "train F-measure=0.3670564311133706, IoU=0.22478206724782068\n",
            "test mean loss=5.4881510734558105\n",
            "test accuracy=0.9447855777877698, recall=0.08022867420349435\n",
            "test precision=0.4716767371601209, specificity=0.9948012664248767\n",
            "test F-measure=0.13713219148001754, IoU=0.07361348500029469\n",
            "[epoch 31]\n",
            "train mean loss=0.9481579661369324\n",
            "train accuracy=0.9918937683105469, train recall=0.6279411764705882\n",
            "train precision=0.34532955923978975, specificity=0.9937917970427633\n",
            "train F-measure=0.4456039655622228, IoU=0.28667338032896944\n",
            "test mean loss=5.377619981765747\n",
            "test accuracy=0.9449875646357914, recall=0.08054984583761562\n",
            "test precision=0.4822149586617958, specificity=0.9949963582852726\n",
            "test F-measure=0.13804111511682307, IoU=0.07413757427059624\n",
            "[epoch 32]\n",
            "train mean loss=0.9307674169540405\n",
            "train accuracy=0.9911003112792969, train recall=0.5242771412984178\n",
            "train precision=0.3967795210569777, specificity=0.9943874826649661\n",
            "train F-measure=0.4517038777908343, IoU=0.2917425622343655\n",
            "test mean loss=5.268298387527466\n",
            "test accuracy=0.9453160128147482, recall=0.08077466598150052\n",
            "test precision=0.5001988862370724, specificity=0.9953308014745229\n",
            "test F-measure=0.1390885963942042, IoU=0.07474219144699694\n",
            "[epoch 33]\n",
            "train mean loss=0.9078836441040039\n",
            "train accuracy=0.9936180114746094, train recall=0.6392896781354052\n",
            "train precision=0.2993762993762994, specificity=0.9948400531306102\n",
            "train F-measure=0.40778761061946905, IoU=0.25611382836816365\n",
            "test mean loss=5.16461968421936\n",
            "test accuracy=0.9456427045863309, recall=0.08048561151079137\n",
            "test precision=0.5194859038142621, specificity=0.9956931149295439\n",
            "test F-measure=0.13937708565072301, IoU=0.07490883003527231\n",
            "[epoch 34]\n",
            "train mean loss=0.8926457166671753\n",
            "train accuracy=0.9936943054199219, train recall=0.624868282402529\n",
            "train precision=0.31375661375661373, specificity=0.9950343613009437\n",
            "train F-measure=0.4177527298344487, IoU=0.26402493321460374\n",
            "test mean loss=5.068608224391937\n",
            "test accuracy=0.9459167041366906, recall=0.0795863309352518\n",
            "test precision=0.5372940156114484, specificity=0.9960349901896665\n",
            "test F-measure=0.13863712655253443, IoU=0.0744815148782687\n",
            "[epoch 35]\n",
            "train mean loss=0.8685929775238037\n",
            "train accuracy=0.99652099609375, train recall=0.5825049701789264\n",
            "train precision=0.29447236180904524, specificity=0.9973169342725338\n",
            "train F-measure=0.39118825100133514, IoU=0.24315352697095435\n",
            "test mean loss=4.984155595302582\n",
            "test accuracy=0.9462135369829137, recall=0.0775950668036999\n",
            "test precision=0.5593887473952304, specificity=0.9964641922825376\n",
            "test F-measure=0.13628543223804823, IoU=0.0731256999303853\n",
            "[epoch 36]\n",
            "train mean loss=0.8537407517433167\n",
            "train accuracy=0.9917716979980469, train recall=0.4887254901960784\n",
            "train precision=0.47228801515869256, specificity=0.9957170977762742\n",
            "train F-measure=0.4803661768248615, IoU=0.31610653138871275\n",
            "test mean loss=4.907354354858398\n",
            "test accuracy=0.9464717288669064, recall=0.07679213771839671\n",
            "test precision=0.5800582241630277, specificity=0.9967837713300434\n",
            "test F-measure=0.13562879346531284, IoU=0.07274774089512277\n",
            "[epoch 37]\n",
            "train mean loss=0.8401899337768555\n",
            "train accuracy=0.9927406311035156, train recall=0.4937641723356009\n",
            "train precision=0.4630515683147262, specificity=0.9961210538443813\n",
            "train F-measure=0.47791495198902606, IoU=0.31398702235039655\n",
            "test mean loss=4.825272083282471\n",
            "test accuracy=0.9467123566771583, recall=0.0763103802672148\n",
            "test precision=0.6007585335018963, specificity=0.9970661900231881\n",
            "test F-measure=0.1354193382918697, IoU=0.07262723521320495\n",
            "[epoch 38]\n",
            "train mean loss=0.8246252536773682\n",
            "train accuracy=0.9942131042480469, train recall=0.597027972027972\n",
            "train precision=0.3927544565842438, specificity=0.9959540229885058\n",
            "train F-measure=0.4738120013874436, IoU=0.3104545454545454\n",
            "test mean loss=4.711134493350983\n",
            "test accuracy=0.9468686769334532, recall=0.07672790339157246\n",
            "test precision=0.6138232271325796, specificity=0.9972073993697604\n",
            "test F-measure=0.1364051615850177, IoU=0.0731946444437636\n",
            "[epoch 39]\n",
            "train mean loss=0.8017483353614807\n",
            "train accuracy=0.9930038452148438, train recall=0.49645390070921985\n",
            "train precision=0.499725425590335, specificity=0.9965003399779494\n",
            "train F-measure=0.49808429118773945, IoU=0.33163265306122447\n",
            "test mean loss=4.6161826848983765\n",
            "test accuracy=0.946970548561151, recall=0.07717754367934224\n",
            "test precision=0.622216468151217, specificity=0.9972891521493549\n",
            "test F-measure=0.13732213269329677, IoU=0.07372296364473079\n",
            "[epoch 40]\n",
            "train mean loss=0.8097593784332275\n",
            "train accuracy=0.9962730407714844, train recall=0.7591836734693878\n",
            "train precision=0.16847826086956522, specificity=0.9964948319772126\n",
            "train F-measure=0.27575982209043737, IoU=0.15993121238177127\n",
            "test mean loss=4.527248024940491\n",
            "test accuracy=0.946916099932554, recall=0.07955421377183967\n",
            "test precision=0.6129670873546152, specificity=0.9970940602889589\n",
            "test F-measure=0.1408306563948034, IoU=0.07574923547400611\n",
            "[epoch 41]\n",
            "train mean loss=0.7732614278793335\n",
            "train accuracy=0.9969291687011719, train recall=0.6487985212569316\n",
            "train precision=0.36335403726708076, specificity=0.9976491095285605\n",
            "train F-measure=0.46582614465826144, IoU=0.3036332179930796\n",
            "test mean loss=4.451549053192139\n",
            "test accuracy=0.9469301511915468, recall=0.08106372045220966\n",
            "test precision=0.6115822631451417, specificity=0.9970215975979547\n",
            "test F-measure=0.143152879789014, IoU=0.07709459665841963\n",
            "[epoch 42]\n",
            "train mean loss=0.7585588693618774\n",
            "train accuracy=0.9969940185546875, train recall=0.7731343283582089\n",
            "train precision=0.266735324407827, specificity=0.997280460182805\n",
            "train F-measure=0.3966309341500766, IoU=0.24737344794651384\n",
            "test mean loss=4.37645560503006\n",
            "test accuracy=0.946933664006295, recall=0.08177029804727647\n",
            "test precision=0.6106980091148957, specificity=0.9969844372435935\n",
            "test F-measure=0.14422886276731342, IoU=0.07771910009463048\n",
            "[epoch 43]\n",
            "train mean loss=0.7416827082633972\n",
            "train accuracy=0.9929695129394531, train recall=0.5416243654822335\n",
            "train precision=0.5316392625809666, specificity=0.9963870332931039\n",
            "train F-measure=0.5365853658536586, IoU=0.36666666666666664\n",
            "test mean loss=4.29667466878891\n",
            "test accuracy=0.9469231255620504, recall=0.08218782117163412\n",
            "test precision=0.609140680790288, specificity=0.9969491349069505\n",
            "test F-measure=0.144834026657611, IoU=0.07807065714808713\n",
            "[epoch 44]\n",
            "train mean loss=0.7162925004959106\n",
            "train accuracy=0.9971733093261719, train recall=0.6588983050847458\n",
            "train precision=0.3490460157126824, specificity=0.9977834846678284\n",
            "train F-measure=0.4563462949376376, IoU=0.2956273764258555\n",
            "test mean loss=4.221085071563721\n",
            "test accuracy=0.9469670357464028, recall=0.08138489208633093\n",
            "test precision=0.6141541444498303, specificity=0.9970420357928533\n",
            "test F-measure=0.14372412228461232, IoU=0.07742605719872891\n",
            "[epoch 45]\n",
            "train mean loss=0.7129648923873901\n",
            "train accuracy=0.9941978454589844, train recall=0.6441176470588236\n",
            "train precision=0.45791949817041294, specificity=0.996023529050862\n",
            "train F-measure=0.5352887259395049, IoU=0.36545682102628285\n",
            "test mean loss=4.146504461765289\n",
            "test accuracy=0.94708998426259, recall=0.07977903391572456\n",
            "test precision=0.627906976744186, specificity=0.9972649979190201\n",
            "test F-measure=0.1415707283711387, IoU=0.07617762512266929\n",
            "[epoch 46]\n",
            "train mean loss=0.7091842889785767\n",
            "train accuracy=0.9945564270019531, train recall=0.6398601398601399\n",
            "train precision=0.41900400686891814, specificity=0.9961111111111111\n",
            "train F-measure=0.5063991698374265, IoU=0.33904585456229736\n",
            "test mean loss=4.073534786701202\n",
            "test accuracy=0.9472129327787769, recall=0.07669578622816033\n",
            "test precision=0.6464537087168382, specificity=0.9975734288602176\n",
            "test F-measure=0.13712316968130922, IoU=0.07360828555576104\n",
            "[epoch 47]\n",
            "train mean loss=0.6797472238540649\n",
            "train accuracy=0.9938125610351562, train recall=0.5270049099836334\n",
            "train precision=0.5613015688553167, specificity=0.9970996231430865\n",
            "train F-measure=0.5436128306133933, IoU=0.3732612055641422\n",
            "test mean loss=3.996080994606018\n",
            "test accuracy=0.9473885735161871, recall=0.07637461459403906\n",
            "test precision=0.6653609401231113, specificity=0.9977778108092039\n",
            "test F-measure=0.13702103140305388, IoU=0.0735494247185451\n",
            "[epoch 48]\n",
            "train mean loss=0.7467014789581299\n",
            "train accuracy=0.9948005676269531, train recall=0.9607843137254902\n",
            "train precision=0.06726149622512011, specificity=0.9948138084734508\n",
            "train F-measure=0.12572161642078256, IoU=0.06707734428473648\n",
            "test mean loss=3.932291626930237\n",
            "test accuracy=0.9474799066996403, recall=0.07714542651593011\n",
            "test precision=0.6728291316526611, specificity=0.9978298353053094\n",
            "test F-measure=0.13841986976315332, IoU=0.07435611688954928\n",
            "[epoch 49]\n",
            "train mean loss=0.653327465057373\n",
            "train accuracy=0.9948348999023438, train recall=0.6061415220293725\n",
            "train precision=0.5430622009569378, specificity=0.9970688213131987\n",
            "train F-measure=0.5728706624605678, IoU=0.40141467727674623\n",
            "test mean loss=3.8585346341133118\n",
            "test accuracy=0.9475238168839928, recall=0.07955421377183967\n",
            "test precision=0.6703653585926929, specificity=0.9977369344194066\n",
            "test F-measure=0.14222962303694986, IoU=0.07655931260431477\n",
            "[epoch 50]\n",
            "train mean loss=0.6492340564727783\n",
            "train accuracy=0.9952392578125, train recall=0.6596417281348789\n",
            "train precision=0.40361057382333976, specificity=0.9964585845824001\n",
            "train F-measure=0.5007999999999999, IoU=0.33404482390608325\n",
            "test mean loss=3.7815062403678894\n",
            "test accuracy=0.947560701438849, recall=0.08173818088386434\n",
            "test precision=0.6679790026246719, specificity=0.997649607586658\n",
            "test F-measure=0.14565329365306473, IoU=0.07854695842720903\n",
            "[epoch 51]\n",
            "train mean loss=0.6765936613082886\n",
            "train accuracy=0.9926071166992188, train recall=0.9662162162162162\n",
            "train precision=0.06888246628131021, specificity=0.9926220247637368\n",
            "train F-measure=0.12859712230215828, IoU=0.06871696299855838\n",
            "test mean loss=3.7195432782173157\n",
            "test accuracy=0.9476432525854317, recall=0.07974691675231244\n",
            "test precision=0.6823303105248695, specificity=0.9978521315179262\n",
            "test F-measure=0.14280373831775703, IoU=0.07689210950080515\n",
            "[epoch 52]\n",
            "train mean loss=0.6259951591491699\n",
            "train accuracy=0.9966964721679688, train recall=0.7922437673130194\n",
            "train precision=0.265552460538533, specificity=0.9969784134187476\n",
            "train F-measure=0.39777468706536856, IoU=0.2482638888888889\n",
            "test mean loss=3.663086712360382\n",
            "test accuracy=0.9476362269559353, recall=0.07727389516957862\n",
            "test precision=0.6895958727429062, specificity=0.9979877668113443\n",
            "test F-measure=0.13897472924187726, IoU=0.07467643316055743\n",
            "[epoch 53]\n",
            "train mean loss=0.6138256192207336\n",
            "train accuracy=0.9939842224121094, train recall=0.5328798185941043\n",
            "train precision=0.5552274069698759, specificity=0.9971080728166526\n",
            "train F-measure=0.5438241249638414, IoU=0.3734604688120779\n",
            "test mean loss=3.5997084379196167\n",
            "test accuracy=0.9476590602517986, recall=0.0763103802672148\n",
            "test precision=0.6955503512880562, specificity=0.9980676615732208\n",
            "test F-measure=0.13753183607316508, IoU=0.073843858776728\n",
            "[epoch 54]\n",
            "train mean loss=0.6333606243133545\n",
            "train accuracy=0.9919281005859375, train recall=0.7926829268292683\n",
            "train precision=0.05877034358047016, specificity=0.9920528284601878\n",
            "train F-measure=0.10942760942760942, IoU=0.0578806767586821\n",
            "test mean loss=3.5477269887924194\n",
            "test accuracy=0.9476450089928058, recall=0.0724563206577595\n",
            "test precision=0.7085427135678392, specificity=0.9982757595576431\n",
            "test F-measure=0.13146853146853146, IoU=0.07035928143712575\n",
            "[epoch 55]\n",
            "train mean loss=0.6269670128822327\n",
            "train accuracy=0.9929580688476562, train recall=0.9117647058823529\n",
            "train precision=0.04818652849740933, specificity=0.9929896734111325\n",
            "train F-measure=0.09153543307086615, IoU=0.0479628674574523\n",
            "test mean loss=3.5145288705825806\n",
            "test accuracy=0.9474623426258992, recall=0.06330292908530319\n",
            "test precision=0.7251655629139073, specificity=0.9986120607646114\n",
            "test F-measure=0.11644118863354404, IoU=0.06181977856537967\n",
            "[epoch 56]\n",
            "train mean loss=0.5705994963645935\n",
            "train accuracy=0.9958992004394531, train recall=0.5701468189233279\n",
            "train precision=0.5605453087409783, specificity=0.9978997232847101\n",
            "train F-measure=0.5653052972098667, IoU=0.39402480270574974\n",
            "test mean loss=3.478483259677887\n",
            "test accuracy=0.9474008683678058, recall=0.05771454265159301\n",
            "test precision=0.7471933471933472, specificity=0.9988703252274214\n",
            "test F-measure=0.10715244029695001, IoU=0.05660912298387097\n",
            "[epoch 57]\n",
            "train mean loss=0.554999589920044\n",
            "train accuracy=0.9978561401367188, train recall=0.5526802218114603\n",
            "train precision=0.48303715670436187, specificity=0.998776772437625\n",
            "train F-measure=0.5155172413793103, IoU=0.34727061556329847\n",
            "test mean loss=3.424177587032318\n",
            "test accuracy=0.9474307272931655, recall=0.05626927029804728\n",
            "test precision=0.762402088772846, specificity=0.998985522325941\n",
            "test F-measure=0.10480349344978167, IoU=0.055299539170506916\n",
            "[epoch 58]\n",
            "train mean loss=0.5421555042266846\n",
            "train accuracy=0.9976043701171875, train recall=0.5252525252525253\n",
            "train precision=0.5490196078431373, specificity=0.998856382266658\n",
            "train F-measure=0.5368731563421829, IoU=0.36693548387096775\n",
            "test mean loss=3.3549473881721497\n",
            "test accuracy=0.94757826551259, recall=0.05826053442959918\n",
            "test precision=0.7758768177929855, specificity=0.9990263987157382\n",
            "test F-measure=0.10838262532114476, IoU=0.0572962728995578\n",
            "[epoch 59]\n",
            "train mean loss=0.55909264087677\n",
            "train accuracy=0.9957427978515625, train recall=0.8269230769230769\n",
            "train precision=0.10591133004926108, specificity=0.9958433210681406\n",
            "train F-measure=0.1877729257641921, IoU=0.10361445783132531\n",
            "test mean loss=3.298242926597595\n",
            "test accuracy=0.9476502782149281, recall=0.05874229188078109\n",
            "test precision=0.785990545767082, specificity=0.9990747071764077\n",
            "test F-measure=0.10931476556196396, IoU=0.057817538091926406\n",
            "[epoch 60]\n",
            "train mean loss=0.5174505710601807\n",
            "train accuracy=0.9970664978027344, train recall=0.5690200210748156\n",
            "train precision=0.6, specificity=0.9986217194050422\n",
            "train F-measure=0.5840995132504057, IoU=0.4125286478227655\n",
            "test mean loss=3.2314446568489075\n",
            "test accuracy=0.947824162544964, recall=0.06092625899280576\n",
            "test precision=0.8024534686971235, specificity=0.9991323057256674\n",
            "test F-measure=0.11325373134328358, IoU=0.06002594690377496\n",
            "[epoch 61]\n",
            "train mean loss=0.5014302730560303\n",
            "train accuracy=0.998321533203125, train recall=0.5526838966202783\n",
            "train precision=0.563894523326572, specificity=0.9991782633455766\n",
            "train F-measure=0.5582329317269076, IoU=0.3871866295264624\n",
            "test mean loss=3.1716131567955017\n",
            "test accuracy=0.9479383290242805, recall=0.06237153134635149\n",
            "test precision=0.8128924236082042, specificity=0.9991694660800285\n",
            "test F-measure=0.1158538404175988, IoU=0.06148877560713042\n",
            "[epoch 62]\n",
            "train mean loss=0.5529524087905884\n",
            "train accuracy=0.9960060119628906, train recall=0.8257575757575758\n",
            "train precision=0.09620476610767872, specificity=0.9960917820557837\n",
            "train F-measure=0.17233201581027668, IoU=0.09429065743944637\n",
            "test mean loss=3.110443592071533\n",
            "test accuracy=0.9480366878372302, recall=0.06452338129496403\n",
            "test precision=0.814349412241589, specificity=0.9991490278851299\n",
            "test F-measure=0.11957265720322591, IoU=0.06358802304234981\n",
            "[epoch 63]\n",
            "train mean loss=0.48449742794036865\n",
            "train accuracy=0.9978713989257812, train recall=0.6055172413793104\n",
            "train precision=0.6174402250351617, specificity=0.9989595247476274\n",
            "train F-measure=0.6114206128133706, IoU=0.440320962888666\n",
            "test mean loss=3.052587568759918\n",
            "test accuracy=0.948103431317446, recall=0.06619347379239465\n",
            "test precision=0.8136596920647453, specificity=0.9991230156370771\n",
            "test F-measure=0.12242715851376636, IoU=0.06520501138952164\n",
            "[epoch 64]\n",
            "train mean loss=0.49051207304000854\n",
            "train accuracy=0.9978828430175781, train recall=0.7232142857142857\n",
            "train precision=0.24732824427480915, specificity=0.9981177458766035\n",
            "train F-measure=0.36860068259385664, IoU=0.22594142259414227\n",
            "test mean loss=3.0068031549453735\n",
            "test accuracy=0.948033175022482, recall=0.06420220966084275\n",
            "test precision=0.8162515312372397, specificity=0.9991638920268744\n",
            "test F-measure=0.11904123864820604, IoU=0.06328753245108593\n",
            "[epoch 65]\n",
            "train mean loss=0.4915386438369751\n",
            "train accuracy=0.9963912963867188, train recall=0.9901960784313726\n",
            "train precision=0.0965583173996176, specificity=0.9963937078788896\n",
            "train F-measure=0.1759581881533101, IoU=0.0964660936007641\n",
            "test mean loss=2.9700652360916138\n",
            "test accuracy=0.9479207649505396, recall=0.06073355601233299\n",
            "test precision=0.8232477144101001, specificity=0.9992456448064688\n",
            "test F-measure=0.1131217659198995, IoU=0.059951810284699765\n",
            "[epoch 66]\n",
            "train mean loss=0.46378305554389954\n",
            "train accuracy=0.9964103698730469, train recall=0.5091185410334347\n",
            "train precision=0.694300518134715, specificity=0.9988689864585091\n",
            "train F-measure=0.5874616396317405, IoU=0.4158907510862818\n",
            "test mean loss=2.9271387457847595\n",
            "test accuracy=0.9479154957284173, recall=0.05951310380267215\n",
            "test precision=0.83318345323741, specificity=0.9993106754266009\n",
            "test F-measure=0.11109112709832134, IoU=0.058812327419303645\n",
            "[epoch 67]\n",
            "train mean loss=0.4498993754386902\n",
            "train accuracy=0.9960098266601562, train recall=0.4854557207498384\n",
            "train precision=0.7502497502497503, specificity=0.9990406643207711\n",
            "train F-measure=0.5894819466248037, IoU=0.41791875347801893\n",
            "test mean loss=2.876639246940613\n",
            "test accuracy=0.9479435982464028, recall=0.0605087358684481\n",
            "test precision=0.8299559471365638, specificity=0.99928280516083\n",
            "test F-measure=0.11279410884272287, IoU=0.05976778123215532\n",
            "[epoch 68]\n",
            "train mean loss=0.44214147329330444\n",
            "train accuracy=0.9949264526367188, train recall=0.4401015228426396\n",
            "train precision=0.7925045703839122, specificity=0.9991275069761006\n",
            "train F-measure=0.5659268929503917, IoU=0.3946290395994538\n",
            "test mean loss=2.816678047180176\n",
            "test accuracy=0.9480226365782374, recall=0.0630781089414183\n",
            "test precision=0.8234800838574423, specificity=0.999217774540698\n",
            "test F-measure=0.11718027505146028, IoU=0.062236587761827805\n",
            "[epoch 69]\n",
            "train mean loss=0.4269775152206421\n",
            "train accuracy=0.9975051879882812, train recall=0.5981416957026713\n",
            "train precision=0.6257594167679222, specificity=0.998821201532438\n",
            "train F-measure=0.6116389548693587, IoU=0.4405474764756202\n",
            "test mean loss=2.764590799808502\n",
            "test accuracy=0.9480173673561151, recall=0.06266058581706063\n",
            "test precision=0.8259949195596952, specificity=0.9992363547178786\n",
            "test F-measure=0.11648456624276075, IoU=0.06184423241512663\n",
            "[epoch 70]\n",
            "train mean loss=0.42674940824508667\n",
            "train accuracy=0.9976997375488281, train recall=0.6517857142857143\n",
            "train precision=0.21758569299552907, specificity=0.9979955711667685\n",
            "train F-measure=0.32625698324022345, IoU=0.19492656875834447\n",
            "test mean loss=2.727782130241394\n",
            "test accuracy=0.9479365726169064, recall=0.06018756423432682\n",
            "test precision=0.831410825199645, specificity=0.9992939532671383\n",
            "test F-measure=0.11224917640011979, IoU=0.05946186064221348\n",
            "[epoch 71]\n",
            "train mean loss=0.4064052104949951\n",
            "train accuracy=0.9983253479003906, train recall=0.5637707948243993\n",
            "train precision=0.6003937007874016, specificity=0.9992240150151184\n",
            "train F-measure=0.5815061963775024, IoU=0.4099462365591398\n",
            "test mean loss=2.692911207675934\n",
            "test accuracy=0.9478803675809353, recall=0.05874229188078109\n",
            "test precision=0.8328779599271403, specificity=0.9993181074974731\n",
            "test F-measure=0.10974438977559103, IoU=0.0580579627337079\n",
            "[epoch 72]\n",
            "train mean loss=0.4280189573764801\n",
            "train accuracy=0.9973373413085938, train recall=1.0\n",
            "train precision=0.07427055702917772, specificity=0.9973367723817954\n",
            "train F-measure=0.1382716049382716, IoU=0.07427055702917772\n",
            "test mean loss=2.6605851650238037\n",
            "test accuracy=0.9477591754721223, recall=0.05585174717368962\n",
            "test precision=0.834052757793765, specificity=0.9993571258695523\n",
            "test F-measure=0.10469281478582823, IoU=0.05523791372847977\n",
            "[epoch 73]\n",
            "train mean loss=0.4123207926750183\n",
            "train accuracy=0.9975090026855469, train recall=0.9803921568627451\n",
            "train precision=0.13315579227696406, specificity=0.9975156654276796\n",
            "train F-measure=0.23446658851113716, IoU=0.13280212483399734\n",
            "test mean loss=2.6370257139205933\n",
            "test accuracy=0.9475694834757195, recall=0.051130524152106886\n",
            "test precision=0.8383359662980516, specificity=0.9994295885605565\n",
            "test F-measure=0.09638262448917816, IoU=0.05063130108450212\n",
            "[epoch 74]\n",
            "train mean loss=0.3883141577243805\n",
            "train accuracy=0.9987411499023438, train recall=0.6026785714285714\n",
            "train precision=0.35904255319148937, specificity=0.9990798717165547\n",
            "train F-measure=0.45, IoU=0.2903225806451613\n",
            "test mean loss=2.6195304095745087\n",
            "test accuracy=0.9473411505170863, recall=0.045766957862281604\n",
            "test precision=0.8407079646017699, specificity=0.9994983352161246\n",
            "test F-measure=0.08680819956748195, IoU=0.045373495510412025\n",
            "[epoch 75]\n",
            "train mean loss=0.39129406213760376\n",
            "train accuracy=0.9979476928710938, train recall=0.5333333333333333\n",
            "train precision=0.08823529411764706, specificity=0.998107260335656\n",
            "train F-measure=0.15141955835962145, IoU=0.08191126279863481\n",
            "test mean loss=2.601798564195633\n",
            "test accuracy=0.9471584841501799, recall=0.041206320657759504\n",
            "test precision=0.8468646864686469, specificity=0.9995689398894108\n",
            "test F-measure=0.07858871091237635, IoU=0.040901555725580206\n",
            "[epoch 76]\n",
            "train mean loss=0.37326809763908386\n",
            "train accuracy=0.9986610412597656, train recall=0.5588235294117647\n",
            "train precision=0.304, specificity=0.9990035886080781\n",
            "train F-measure=0.3937823834196891, IoU=0.24516129032258063\n",
            "test mean loss=2.575688987970352\n",
            "test accuracy=0.9470741765962231, recall=0.03834789311408016\n",
            "test precision=0.8620938628158845, specificity=0.9996451186158511\n",
            "test F-measure=0.07342947633836598, IoU=0.03811408688990328\n",
            "[epoch 77]\n",
            "train mean loss=0.3713206350803375\n",
            "train accuracy=0.998931884765625, train recall=0.7727272727272727\n",
            "train precision=0.1118421052631579, specificity=0.9989698588325067\n",
            "train F-measure=0.19540229885057472, IoU=0.10828025477707007\n",
            "test mean loss=2.552637815475464\n",
            "test accuracy=0.9469933818570144, recall=0.03545734840698869\n",
            "test precision=0.882494004796163, specificity=0.9997268713954456\n",
            "test F-measure=0.06817550251644178, IoU=0.03529073298596682\n",
            "[epoch 78]\n",
            "train mean loss=0.369425892829895\n",
            "train accuracy=0.9989852905273438, train recall=0.4642857142857143\n",
            "train precision=0.09923664122137404, specificity=0.9990995390861085\n",
            "train F-measure=0.16352201257861637, IoU=0.08904109589041095\n",
            "test mean loss=2.527096152305603\n",
            "test accuracy=0.9469986510791367, recall=0.03459018499486125\n",
            "test precision=0.9020100502512562, specificity=0.9997826119269874\n",
            "test F-measure=0.06662542530157749, IoU=0.034460691773589736\n",
            "[epoch 79]\n",
            "train mean loss=0.3463135063648224\n",
            "train accuracy=0.9988822937011719, train recall=0.3519163763066202\n",
            "train precision=0.4855769230769231, specificity=0.9995913800280306\n",
            "train F-measure=0.40808080808080804, IoU=0.2563451776649746\n",
            "test mean loss=2.492443174123764\n",
            "test accuracy=0.9470583689298561, recall=0.03491135662898253\n",
            "test precision=0.9211864406779661, specificity=0.9998272043522207\n",
            "test F-measure=0.06727317737343731, IoU=0.034807390566460664\n",
            "[epoch 80]\n",
            "train mean loss=0.33890607953071594\n",
            "train accuracy=0.9988517761230469, train recall=0.34328358208955223\n",
            "train precision=0.5867346938775511, specificity=0.9996906141500101\n",
            "train F-measure=0.4331450094161959, IoU=0.2764423076923077\n",
            "test mean loss=2.4493636190891266\n",
            "test accuracy=0.9472322532598921, recall=0.03789825282631038\n",
            "test precision=0.9313338595106551, specificity=0.999838352458529\n",
            "test F-measure=0.07283276239854335, IoU=0.037792652852064186\n",
            "[epoch 81]\n",
            "train mean loss=0.33887821435928345\n",
            "train accuracy=0.9990043640136719, train recall=0.8392857142857143\n",
            "train precision=0.15719063545150502, specificity=0.9990384908885566\n",
            "train F-measure=0.26478873239436623, IoU=0.1525974025974026\n",
            "test mean loss=2.414357900619507\n",
            "test accuracy=0.9472550865557554, recall=0.03799460431654676\n",
            "test precision=0.9388888888888889, specificity=0.9998569326357096\n",
            "test F-measure=0.07303370786516854, IoU=0.037900874635568516\n",
            "[epoch 82]\n",
            "train mean loss=0.33188140392303467\n",
            "train accuracy=0.9986038208007812, train recall=0.7549019607843137\n",
            "train precision=0.18421052631578946, specificity=0.9986986818906893\n",
            "train F-measure=0.2961538461538461, IoU=0.17381489841986456\n",
            "test mean loss=2.3960091173648834\n",
            "test accuracy=0.9471128175584532, recall=0.03455806783144913\n",
            "test precision=0.9547471162377995, specificity=0.9999052410963791\n",
            "test F-measure=0.06670179462542232, IoU=0.03450155513515247\n",
            "[epoch 83]\n",
            "train mean loss=0.3213253617286682\n",
            "train accuracy=0.9992485046386719, train recall=0.8225806451612904\n",
            "train precision=0.21518987341772153, specificity=0.9992902984562083\n",
            "train F-measure=0.34113712374581945, IoU=0.2056451612903226\n",
            "test mean loss=2.3893428444862366\n",
            "test accuracy=0.9468792153776978, recall=0.02974049331963001\n",
            "test precision=0.9645833333333333, specificity=0.999936827397586\n",
            "test F-measure=0.057701894317048856, IoU=0.029708052614693614\n",
            "[epoch 84]\n",
            "train mean loss=0.31194260716438293\n",
            "train accuracy=0.9980201721191406, train recall=0.3751651254953765\n",
            "train precision=0.8606060606060606, specificity=0.9998240157314633\n",
            "train F-measure=0.5225390984360626, IoU=0.35367372353673726\n",
            "test mean loss=2.358008235692978\n",
            "test accuracy=0.9468950230440647, recall=0.02986896197327852\n",
            "test precision=0.9697601668404588, specificity=0.9999461174861763\n",
            "test F-measure=0.057952952173235706, IoU=0.02984116797689716\n",
            "[epoch 85]\n",
            "train mean loss=0.30538076162338257\n",
            "train accuracy=0.9991683959960938, train recall=0.3641304347826087\n",
            "train precision=0.39880952380952384, specificity=0.999614444953428\n",
            "train F-measure=0.3806818181818181, IoU=0.23508771929824562\n",
            "test mean loss=2.324303299188614\n",
            "test accuracy=0.9469477152652878, recall=0.030800359712230215\n",
            "test precision=0.9716312056737588, specificity=0.9999479755038944\n",
            "test F-measure=0.05970799738505121, IoU=0.03077268643306379\n",
            "[epoch 86]\n",
            "train mean loss=0.30094850063323975\n",
            "train accuracy=0.9990653991699219, train recall=0.5784313725490197\n",
            "train precision=0.2260536398467433, specificity=0.9992291312079743\n",
            "train F-measure=0.32506887052341593, IoU=0.19407894736842105\n",
            "test mean loss=2.2940312325954437\n",
            "test accuracy=0.9469459588579137, recall=0.030575539568345324\n",
            "test precision=0.9774127310061602, specificity=0.9999591236102028\n",
            "test F-measure=0.0592961694176269, IoU=0.030553950831247192\n",
            "[epoch 87]\n",
            "train mean loss=0.29111725091934204\n",
            "train accuracy=0.9983100891113281, train recall=0.3549432739059968\n",
            "train precision=0.8295454545454546, specificity=0.9998279336359153\n",
            "train F-measure=0.4971623155505108, IoU=0.33081570996978854\n",
            "test mean loss=2.2479782700538635\n",
            "test accuracy=0.9471005227068345, recall=0.03336973278520041\n",
            "test precision=0.980188679245283, specificity=0.9999609816279208\n",
            "test F-measure=0.06454217915268977, IoU=0.033347241390377765\n",
            "[epoch 88]\n",
            "train mean loss=0.284714937210083\n",
            "train accuracy=0.9990081787109375, train recall=0.3588850174216028\n",
            "train precision=0.5754189944134078, specificity=0.9997097652535544\n",
            "train F-measure=0.4420600858369099, IoU=0.2837465564738292\n",
            "test mean loss=2.205964148044586\n",
            "test accuracy=0.9472094199640287, recall=0.03532887975334018\n",
            "test precision=0.9821428571428571, specificity=0.9999628396456388\n",
            "test F-measure=0.06820436507936509, IoU=0.035306201052766724\n",
            "[epoch 89]\n",
            "train mean loss=0.2937914729118347\n",
            "train accuracy=0.9952583312988281, train recall=0.2903981264637002\n",
            "train precision=0.9411764705882353, specificity=0.9998809688368735\n",
            "train F-measure=0.4438478747203579, IoU=0.2852213916043703\n",
            "test mean loss=2.1469673812389374\n",
            "test accuracy=0.9474799066996403, recall=0.040724563206577594\n",
            "test precision=0.9738863287250384, specificity=0.999936827397586\n",
            "test F-measure=0.07817991244836302, IoU=0.04068014116137311\n",
            "[epoch 90]\n",
            "train mean loss=0.27351492643356323\n",
            "train accuracy=0.9989280700683594, train recall=0.3693379790940767\n",
            "train precision=0.5145631067961165, specificity=0.9996181121757295\n",
            "train F-measure=0.4300202839756592, IoU=0.2739018087855297\n",
            "test mean loss=2.1045852303504944\n",
            "test accuracy=0.9476063680305755, recall=0.04329393627954779\n",
            "test precision=0.9697841726618706, specificity=0.9999219632558416\n",
            "test F-measure=0.0828875361249462, IoU=0.043235614856629676\n",
            "[epoch 91]\n",
            "train mean loss=0.27156397700309753\n",
            "train accuracy=0.9991302490234375, train recall=0.8636363636363636\n",
            "train precision=0.14615384615384616, specificity=0.999152995040061\n",
            "train F-measure=0.25, IoU=0.14285714285714285\n",
            "test mean loss=2.0830571353435516\n",
            "test accuracy=0.9474974707733813, recall=0.040949383350462484\n",
            "test precision=0.9762633996937213, specificity=0.9999424014507402\n",
            "test F-measure=0.07860181246532273, IoU=0.04090865338338628\n",
            "[epoch 92]\n",
            "train mean loss=0.26447585225105286\n",
            "train accuracy=0.9991798400878906, train recall=0.7340425531914894\n",
            "train precision=0.26640926640926643, specificity=0.9992749475290975\n",
            "train F-measure=0.39093484419263463, IoU=0.24295774647887325\n",
            "test mean loss=2.078073740005493\n",
            "test accuracy=0.9472533301483813, recall=0.03606757451181911\n",
            "test precision=0.9842243645924628, specificity=0.999966555681075\n",
            "test F-measure=0.06958515351488676, IoU=0.03604673557167619\n",
            "[epoch 93]\n",
            "train mean loss=0.2583872675895691\n",
            "train accuracy=0.9983024597167969, train recall=0.4317241379310345\n",
            "train precision=0.9046242774566474, specificity=0.9998737658701166\n",
            "train F-measure=0.5845004668534081, IoU=0.4129287598944591\n",
            "test mean loss=2.0588555335998535\n",
            "test accuracy=0.9471532149280576, recall=0.03401207605344296\n",
            "test precision=0.9897196261682243, specificity=0.9999795618051014\n",
            "test F-measure=0.06576414332733031, IoU=0.03400006421164157\n",
            "[epoch 94]\n",
            "train mean loss=0.25456464290618896\n",
            "train accuracy=0.999359130859375, train recall=0.20408163265306123\n",
            "train precision=0.18181818181818182, specificity=0.9996565488502018\n",
            "train F-measure=0.19230769230769232, IoU=0.10638297872340426\n",
            "test mean loss=2.037426680326462\n",
            "test accuracy=0.9470530997077338, recall=0.03208504624871531\n",
            "test precision=0.9920556107249255, specificity=0.9999851358582555\n",
            "test F-measure=0.062159723734561176, IoU=0.03207680452093501\n",
            "[epoch 95]\n",
            "train mean loss=0.2496359646320343\n",
            "train accuracy=0.9996185302734375, train recall=0.3870967741935484\n",
            "train precision=0.27906976744186046, specificity=0.9997634328187361\n",
            "train F-measure=0.32432432432432434, IoU=0.1935483870967742\n",
            "test mean loss=2.012847810983658\n",
            "test accuracy=0.9469916254496403, recall=0.030960945529290854\n",
            "test precision=0.9917695473251029, specificity=0.9999851358582555\n",
            "test F-measure=0.06004734022673477, IoU=0.030952992550732084\n",
            "[epoch 96]\n",
            "train mean loss=0.2440563142299652\n",
            "train accuracy=0.9996566772460938, train recall=0.3108108108108108\n",
            "train precision=0.3709677419354839, specificity=0.9998511847979548\n",
            "train F-measure=0.338235294117647, IoU=0.20353982300884957\n",
            "test mean loss=1.98598712682724\n",
            "test accuracy=0.9469986510791367, recall=0.030960945529290854\n",
            "test precision=0.9958677685950413, specificity=0.9999925679291277\n",
            "test F-measure=0.060054821829055575, IoU=0.030956968529222865\n",
            "[epoch 97]\n",
            "train mean loss=0.23835065960884094\n",
            "train accuracy=0.9996681213378906, train recall=0.5392156862745098\n",
            "train precision=0.5789473684210527, specificity=0.9998473527144504\n",
            "train F-measure=0.5583756345177665, IoU=0.3873239436619718\n",
            "test mean loss=1.9577100574970245\n",
            "test accuracy=0.947039048448741, recall=0.03166752312435766\n",
            "test precision=0.9969666329625885, specificity=0.9999944259468458\n",
            "test F-measure=0.06138521400778211, IoU=0.03166447220527313\n",
            "[epoch 98]\n",
            "train mean loss=0.2341425120830536\n",
            "train accuracy=0.9997520446777344, train recall=0.18181818181818182\n",
            "train precision=0.21621621621621623, specificity=0.9998893552079359\n",
            "train F-measure=0.19753086419753088, IoU=0.1095890410958904\n",
            "test mean loss=1.929665893316269\n",
            "test accuracy=0.9471461892985612, recall=0.03375513874614594\n",
            "test precision=0.9933837429111532, specificity=0.9999869938759736\n",
            "test F-measure=0.06529166925514071, IoU=0.033747551616735705\n",
            "[epoch 99]\n",
            "train mean loss=0.23032227158546448\n",
            "train accuracy=0.9982948303222656, train recall=0.3986206896551724\n",
            "train precision=0.9633333333333334, specificity=0.9999579219567055\n",
            "train F-measure=0.5639024390243903, IoU=0.39266304347826086\n",
            "test mean loss=1.8946163952350616\n",
            "test accuracy=0.9473727658498201, recall=0.03809095580678314\n",
            "test precision=0.9891576313594662, specificity=0.9999758457696653\n",
            "test F-measure=0.07335704345136848, IoU=0.03807505858936081\n",
            "[epoch 100]\n",
            "train mean loss=0.2247833013534546\n",
            "train accuracy=0.9992637634277344, train recall=0.16666666666666666\n",
            "train precision=0.136, specificity=0.9995878523290159\n",
            "train F-measure=0.1497797356828194, IoU=0.08095238095238096\n",
            "test mean loss=1.8702476024627686\n",
            "test accuracy=0.9474886887365108, recall=0.04043550873586845\n",
            "test precision=0.9843627834245504, specificity=0.9999628396456388\n",
            "test F-measure=0.07768008637976247, IoU=0.040409551932212094\n",
            "[epoch 101]\n",
            "train mean loss=0.22602994740009308\n",
            "train accuracy=0.9974441528320312, train recall=0.3921765295887663\n",
            "train precision=0.8593406593406593, specificity=0.9997549273014815\n",
            "train F-measure=0.5385674931129476, IoU=0.3685202639019793\n",
            "test mean loss=1.846656709909439\n",
            "test accuracy=0.9476344705485612, recall=0.043582990750256935\n",
            "test precision=0.9748563218390804, specificity=0.999934969379868\n",
            "test F-measure=0.08343580914904082, IoU=0.043534054088736326\n",
            "[epoch 102]\n",
            "train mean loss=0.22857986390590668\n",
            "train accuracy=0.9959487915039062, train recall=0.39951573849878935\n",
            "train precision=0.9041095890410958, specificity=0.9997312777359766\n",
            "train F-measure=0.5541561712846348, IoU=0.3832752613240418\n",
            "test mean loss=1.8154070377349854\n",
            "test accuracy=0.9477943036196043, recall=0.047147995889003085\n",
            "test precision=0.963887065003283, specificity=0.9998978090255068\n",
            "test F-measure=0.0898986496830889, IoU=0.04706485845275881\n",
            "[epoch 103]\n",
            "train mean loss=0.2133113145828247\n",
            "train accuracy=0.999176025390625, train recall=0.8297872340425532\n",
            "train precision=0.2805755395683453, specificity=0.9992367868727342\n",
            "train F-measure=0.4193548387096774, IoU=0.2653061224489796\n",
            "test mean loss=1.796722263097763\n",
            "test accuracy=0.9476906755845323, recall=0.044867677286742035\n",
            "test precision=0.9701388888888889, specificity=0.9999201052381236\n",
            "test F-measure=0.08576866404715128, IoU=0.044805798774816386\n",
            "[epoch 104]\n",
            "train mean loss=0.21878710389137268\n",
            "train accuracy=0.9964218139648438, train recall=0.4032042723631509\n",
            "train precision=0.9320987654320988, specificity=0.9998311886620167\n",
            "train F-measure=0.5629077353215284, IoU=0.39169909208819714\n",
            "test mean loss=1.7767610549926758\n",
            "test accuracy=0.9476537910296763, recall=0.04393627954779034\n",
            "test precision=0.9750534568781183, specificity=0.999934969379868\n",
            "test F-measure=0.08408371492670336, IoU=0.04388694619999358\n",
            "[epoch 105]\n",
            "train mean loss=0.21599653363227844\n",
            "train accuracy=0.9960174560546875, train recall=0.3898305084745763\n",
            "train precision=0.9470588235294117, specificity=0.9998617999785022\n",
            "train F-measure=0.5523156089193825, IoU=0.3815165876777251\n",
            "test mean loss=1.7528617978096008\n",
            "test accuracy=0.9476994576214028, recall=0.044835560123329905\n",
            "test precision=0.9741800418702024, specificity=0.9999312533444319\n",
            "test F-measure=0.08572569007338265, IoU=0.04478234369486415\n",
            "[epoch 106]\n",
            "train mean loss=0.2011755108833313\n",
            "train accuracy=0.9995079040527344, train recall=0.71875\n",
            "train precision=0.16083916083916083, specificity=0.9995421804419484\n",
            "train F-measure=0.26285714285714284, IoU=0.1513157894736842\n",
            "test mean loss=1.7440265715122223\n",
            "test accuracy=0.947560701438849, recall=0.04200924974306269\n",
            "test precision=0.9790419161676647, specificity=0.9999479755038944\n",
            "test F-measure=0.08056171470805618, IoU=0.041971505583365425\n",
            "[epoch 107]\n",
            "train mean loss=0.19967663288116455\n",
            "train accuracy=0.9982681274414062, train recall=0.4317241379310345\n",
            "train precision=0.8816901408450705, specificity=0.9998393383801484\n",
            "train F-measure=0.5796296296296296, IoU=0.40808344198174706\n",
            "test mean loss=1.7355806827545166\n",
            "test accuracy=0.9474711246627698, recall=0.0401464542651593\n",
            "test precision=0.983477576711251, specificity=0.9999609816279208\n",
            "test F-measure=0.07714382695096739, IoU=0.04011939532047373\n",
            "[epoch 108]\n",
            "train mean loss=0.2075425088405609\n",
            "train accuracy=0.995208740234375, train recall=0.35331230283911674\n",
            "train precision=0.9627507163323782, specificity=0.9999000929903705\n",
            "train F-measure=0.5169230769230769, IoU=0.34854771784232363\n",
            "test mean loss=1.7041726410388947\n",
            "test accuracy=0.9476309577338129, recall=0.043101233299075024\n",
            "test precision=0.9838709677419355, specificity=0.9999591236102028\n",
            "test F-measure=0.08258461538461538, IoU=0.043070800436485014\n",
            "[epoch 109]\n",
            "train mean loss=0.20076259970664978\n",
            "train accuracy=0.9963531494140625, train recall=0.4169360051712993\n",
            "train precision=0.9227467811158798, specificity=0.9997927834932866\n",
            "train F-measure=0.5743544078361532, IoU=0.4028732042473454\n",
            "test mean loss=1.665425419807434\n",
            "test accuracy=0.9478891496178058, recall=0.048272096608427545\n",
            "test precision=0.9766081871345029, specificity=0.9999331113621499\n",
            "test F-measure=0.09199693955623565, IoU=0.04821634800461953\n",
            "[epoch 110]\n",
            "train mean loss=0.1992955058813095\n",
            "train accuracy=0.9958419799804688, train recall=0.4257369614512472\n",
            "train precision=0.9070048309178744, specificity=0.9997042783623934\n",
            "train F-measure=0.5794753086419753, IoU=0.40793047256925585\n",
            "test mean loss=1.6243572235107422\n",
            "test accuracy=0.9481807132419064, recall=0.054053186022610486\n",
            "test precision=0.9711482977495672, specificity=0.9999070991140971\n",
            "test F-measure=0.1024065228634884, IoU=0.05396652344000513\n",
            "[epoch 111]\n",
            "train mean loss=0.1850634664297104\n",
            "train accuracy=0.9992561340332031, train recall=0.5931372549019608\n",
            "train precision=0.51931330472103, specificity=0.9995724211651523\n",
            "train F-measure=0.5537757437070938, IoU=0.3829113924050633\n",
            "test mean loss=1.598819375038147\n",
            "test accuracy=0.948312443794964, recall=0.0565262076053443\n",
            "test precision=0.9713024282560706, specificity=0.999903383078661\n",
            "test F-measure=0.10683501274735949, IoU=0.05643196101064512\n",
            "[epoch 112]\n",
            "train mean loss=0.18272364139556885\n",
            "train accuracy=0.9992141723632812, train recall=0.8048780487804879\n",
            "train precision=0.43137254901960786, specificity=0.9993358271623788\n",
            "train F-measure=0.5617021276595745, IoU=0.3905325443786982\n",
            "test mean loss=1.5872582495212555\n",
            "test accuracy=0.9482439439073741, recall=0.05495246659815005\n",
            "test precision=0.9760410724472333, specificity=0.9999219632558416\n",
            "test F-measure=0.1040469457873453, IoU=0.05487843992558856\n",
            "[epoch 113]\n",
            "train mean loss=0.18504966795444489\n",
            "train accuracy=0.9972114562988281, train recall=0.5349544072948328\n",
            "train precision=0.8554070473876063, specificity=0.9995437606391951\n",
            "train F-measure=0.6582515194015894, IoU=0.4905923344947735\n",
            "test mean loss=1.5803521871566772\n",
            "test accuracy=0.9481824696492805, recall=0.053667780061664956\n",
            "test precision=0.9783372365339579, specificity=0.9999312533444319\n",
            "test F-measure=0.10175374497625135, IoU=0.05360408045423925\n",
            "[epoch 114]\n",
            "train mean loss=0.17567282915115356\n",
            "train accuracy=0.9987106323242188, train recall=0.5268022181146026\n",
            "train precision=0.776566757493188, specificity=0.9996865479371414\n",
            "train F-measure=0.6277533039647577, IoU=0.45746388443017655\n",
            "test mean loss=1.5784664154052734\n",
            "test accuracy=0.9480560083183454, recall=0.051130524152106886\n",
            "test precision=0.9815043156596794, specificity=0.9999442594684583\n",
            "test F-measure=0.09719763111301057, IoU=0.05108130655201181\n",
            "[epoch 115]\n",
            "train mean loss=0.17384353280067444\n",
            "train accuracy=0.9995689392089844, train recall=0.8627450980392157\n",
            "train precision=0.47058823529411764, specificity=0.9996221979682647\n",
            "train F-measure=0.6089965397923875, IoU=0.43781094527363185\n",
            "test mean loss=1.574093908071518\n",
            "test accuracy=0.9478329445818345, recall=0.046602004110996915\n",
            "test precision=0.989093387866394, specificity=0.9999702717165111\n",
            "test F-measure=0.0890102137840076, IoU=0.04657806882383154\n",
            "[epoch 116]\n",
            "train mean loss=0.17052863538265228\n",
            "train accuracy=0.9986457824707031, train recall=0.5206106870229008\n",
            "train precision=0.8926701570680629, specificity=0.9998432056415375\n",
            "train F-measure=0.6576663452266153, IoU=0.4899425287356322\n",
            "test mean loss=1.5668562650680542\n",
            "test accuracy=0.9476906755845323, recall=0.043872045220966084\n",
            "test precision=0.9912917271407837, specificity=0.9999777037873833\n",
            "test F-measure=0.08402534292919973, IoU=0.043855143187363554\n",
            "[epoch 117]\n",
            "train mean loss=0.16775023937225342\n",
            "train accuracy=0.9997100830078125, train recall=0.425531914893617\n",
            "train precision=0.6451612903225806, specificity=0.9999160465560007\n",
            "train F-measure=0.5128205128205128, IoU=0.3448275862068966\n",
            "test mean loss=1.5567482113838196\n",
            "test accuracy=0.947666085881295, recall=0.043454522096608425\n",
            "test precision=0.9904831625183016, specificity=0.9999758457696653\n",
            "test F-measure=0.08325641498984679, IoU=0.043436386400847536\n",
            "[epoch 118]\n",
            "train mean loss=0.1658782660961151\n",
            "train accuracy=0.9985160827636719, train recall=0.503448275862069\n",
            "train precision=0.9263959390862944, specificity=0.999889066976769\n",
            "train F-measure=0.6523681858802503, IoU=0.48408488063660476\n",
            "test mean loss=1.5370271503925323\n",
            "test accuracy=0.9478153805080936, recall=0.046377183967112025\n",
            "test precision=0.987012987012987, specificity=0.9999646976633569\n",
            "test F-measure=0.08859167459124513, IoU=0.04634890065800032\n",
            "[epoch 119]\n",
            "train mean loss=0.16141143441200256\n",
            "train accuracy=0.999237060546875, train recall=0.29017857142857145\n",
            "train precision=0.6132075471698113, specificity=0.9998434636530238\n",
            "train F-measure=0.393939393939394, IoU=0.24528301886792453\n",
            "test mean loss=1.5169580280780792\n",
            "test accuracy=0.9479699443570144, recall=0.049621017471736896\n",
            "test precision=0.979708306911858, specificity=0.9999405434330222\n",
            "test F-measure=0.09445786078928867, IoU=0.049570071868583164\n",
            "[epoch 120]\n",
            "train mean loss=0.1703624725341797\n",
            "train accuracy=0.9959030151367188, train recall=0.46424815983175605\n",
            "train precision=0.94136460554371, specificity=0.9997886582488607\n",
            "train F-measure=0.6218309859154929, IoU=0.4512008175779254\n",
            "test mean loss=1.4893549680709839\n",
            "test accuracy=0.9482579951663669, recall=0.05553057553956835\n",
            "test precision=0.9708029197080292, specificity=0.999903383078661\n",
            "test F-measure=0.10505210073822038, IoU=0.05543798897011671\n",
            "[epoch 121]\n",
            "train mean loss=0.15669509768486023\n",
            "train accuracy=0.9993019104003906, train recall=0.8627450980392157\n",
            "train precision=0.3424124513618677, specificity=0.9993550652185528\n",
            "train F-measure=0.49025069637883, IoU=0.3247232472324723\n",
            "test mean loss=1.4756155014038086\n",
            "test accuracy=0.9482825848696043, recall=0.05610868448098664\n",
            "test precision=0.968940654464781, specificity=0.9998959510077888\n",
            "test F-measure=0.10607486566076688, IoU=0.05600795075660426\n",
            "[epoch 122]\n",
            "train mean loss=0.16256654262542725\n",
            "train accuracy=0.9971275329589844, train recall=0.5580774365821095\n",
            "train precision=0.9018338727076591, specificity=0.9996508674600799\n",
            "train F-measure=0.6894845360824743, IoU=0.526117054751416\n",
            "test mean loss=1.4602850079536438\n",
            "test accuracy=0.948312443794964, recall=0.05630138746145941\n",
            "test precision=0.974972191323693, specificity=0.9999163892026874\n",
            "test F-measure=0.10645533491224875, IoU=0.05622013405599564\n",
            "[epoch 123]\n",
            "train mean loss=0.15310820937156677\n",
            "train accuracy=0.9987030029296875, train recall=0.6248275862068966\n",
            "train precision=0.8694817658349329, specificity=0.9997398811869068\n",
            "train F-measure=0.7271268057784912, IoU=0.5712484237074401\n",
            "test mean loss=1.44755220413208\n",
            "test accuracy=0.9483001489433454, recall=0.05598021582733813\n",
            "test precision=0.9764705882352941, specificity=0.9999219632558416\n",
            "test F-measure=0.10588985753774187, IoU=0.05590480466995959\n",
            "[epoch 124]\n",
            "train mean loss=0.15927308797836304\n",
            "train accuracy=0.9966163635253906, train recall=0.5304449648711944\n",
            "train precision=0.9142280524722503, specificity=0.9996736242301372\n",
            "train F-measure=0.6713597628751389, IoU=0.5052983825989961\n",
            "test mean loss=1.4290848076343536\n",
            "test accuracy=0.948259751573741, recall=0.05511305241521069\n",
            "test precision=0.9783352337514253, specificity=0.9999293953267139\n",
            "test F-measure=0.10434782608695653, IoU=0.05504587155963303\n",
            "[epoch 125]\n",
            "train mean loss=0.14751853048801422\n",
            "train accuracy=0.999664306640625, train recall=0.6470588235294118\n",
            "train precision=0.559322033898305, specificity=0.9998015585287855\n",
            "train F-measure=0.6, IoU=0.42857142857142855\n",
            "test mean loss=1.4168866872787476\n",
            "test accuracy=0.9481403158723022, recall=0.05247944501541624\n",
            "test precision=0.985524728588661, specificity=0.9999554075747666\n",
            "test F-measure=0.0996523754345307, IoU=0.0524390243902439\n",
            "[epoch 126]\n",
            "train mean loss=0.14558696746826172\n",
            "train accuracy=0.9993095397949219, train recall=0.24390243902439024\n",
            "train precision=0.41237113402061853, specificity=0.9997824261393999\n",
            "train F-measure=0.3065134099616858, IoU=0.18099547511312217\n",
            "test mean loss=1.4090387225151062\n",
            "test accuracy=0.9480033160971223, recall=0.04981372045220966\n",
            "test precision=0.9878980891719745, specificity=0.9999646976633569\n",
            "test F-measure=0.09484498257200513, IoU=0.04978334135772749\n",
            "[epoch 127]\n",
            "train mean loss=0.1416952908039093\n",
            "train accuracy=0.9997596740722656, train recall=0.5531914893617021\n",
            "train precision=0.7123287671232876, specificity=0.9999198626216371\n",
            "train F-measure=0.6227544910179641, IoU=0.45217391304347826\n",
            "test mean loss=1.40123251080513\n",
            "test accuracy=0.9479717007643885, recall=0.049171377183967115\n",
            "test precision=0.9890180878552972, specificity=0.9999684136987931\n",
            "test F-measure=0.09368498347815446, IoU=0.049144544666645265\n",
            "[epoch 128]\n",
            "train mean loss=0.14088481664657593\n",
            "train accuracy=0.999603271484375, train recall=0.6418918918918919\n",
            "train precision=0.6506849315068494, specificity=0.9998053405395502\n",
            "train F-measure=0.6462585034013606, IoU=0.47738693467336685\n",
            "test mean loss=1.3931682854890823\n",
            "test accuracy=0.947980482801259, recall=0.049139260020554985\n",
            "test precision=0.9928617780661908, specificity=0.9999795618051014\n",
            "test F-measure=0.09364384735440831, IoU=0.049121905801521815\n",
            "[epoch 129]\n",
            "train mean loss=0.14492858946323395\n",
            "train accuracy=0.9973793029785156, train recall=0.5151975683890577\n",
            "train precision=0.9325997248968363, specificity=0.9998121367337862\n",
            "train F-measure=0.6637298091042584, IoU=0.4967032967032967\n",
            "test mean loss=1.3719123303890228\n",
            "test accuracy=0.9482105721672662, recall=0.05376413155190134\n",
            "test precision=0.9858657243816255, specificity=0.9999554075747666\n",
            "test F-measure=0.10196747274167021, IoU=0.05372272143774069\n",
            "[epoch 130]\n",
            "train mean loss=0.13650956749916077\n",
            "train accuracy=0.9995384216308594, train recall=0.75\n",
            "train precision=0.647887323943662, specificity=0.999713696747595\n",
            "train F-measure=0.6952141057934509, IoU=0.5328185328185329\n",
            "test mean loss=1.3560483306646347\n",
            "test accuracy=0.9483282514613309, recall=0.05598021582733813\n",
            "test precision=0.9853024307518372, specificity=0.9999516915393305\n",
            "test F-measure=0.10594134629995441, IoU=0.05593350876067005\n",
            "[epoch 131]\n",
            "train mean loss=0.134026899933815\n",
            "train accuracy=0.9993362426757812, train recall=0.8070175438596491\n",
            "train precision=0.3770491803278688, specificity=0.9994199137503339\n",
            "train F-measure=0.5139664804469273, IoU=0.3458646616541353\n",
            "test mean loss=1.3435446470975876\n",
            "test accuracy=0.9481859824640287, recall=0.05299331963001028\n",
            "test precision=0.9915865384615384, specificity=0.9999739877519472\n",
            "test F-measure=0.10060975609756098, IoU=0.052969502407704656\n",
            "[epoch 133]\n",
            "train mean loss=0.1313900500535965\n",
            "train accuracy=0.999359130859375, train recall=0.5267857142857143\n",
            "train precision=0.6555555555555556, specificity=0.9997632864996946\n",
            "train F-measure=0.5841584158415841, IoU=0.4125874125874126\n",
            "test mean loss=1.3510169237852097\n",
            "test accuracy=0.9478856368030576, recall=0.047180113052415214\n",
            "test precision=0.9972844534962662, specificity=0.9999925679291277\n",
            "test F-measure=0.0900978257536263, IoU=0.047174052665382146\n",
            "[epoch 134]\n",
            "train mean loss=0.12873835861682892\n",
            "train accuracy=0.9995574951171875, train recall=0.358974358974359\n",
            "train precision=0.7777777777777778, specificity=0.999938928500542\n",
            "train F-measure=0.49122807017543857, IoU=0.32558139534883723\n",
            "test mean loss=1.3478369563817978\n",
            "test accuracy=0.9478382138039568, recall=0.046280832476875644\n",
            "test precision=0.9979224376731302, specificity=0.9999944259468458\n",
            "test F-measure=0.08845917740945364, IoU=0.0462763736793089\n",
            "[epoch 135]\n",
            "train mean loss=0.13867704570293427\n",
            "train accuracy=0.9961013793945312, train recall=0.4238875878220141\n",
            "train precision=0.9501312335958005, specificity=0.9998540908322966\n",
            "train F-measure=0.5862348178137652, IoU=0.4146620847651776\n",
            "test mean loss=1.3266592174768448\n",
            "test accuracy=0.9480577647257195, recall=0.050552415210688595\n",
            "test precision=0.9930599369085173, specificity=0.9999795618051014\n",
            "test F-measure=0.09620732862687571, IoU=0.05053456191607539\n",
            "[epoch 136]\n",
            "train mean loss=0.12643368542194366\n",
            "train accuracy=0.9985427856445312, train recall=0.5640686922060766\n",
            "train precision=0.8914405010438413, specificity=0.9998010612616541\n",
            "train F-measure=0.6909385113268607, IoU=0.5278121137206427\n",
            "test mean loss=1.305964320898056\n",
            "test accuracy=0.9482615079811151, recall=0.05482399794450154\n",
            "test precision=0.9838616714697407, specificity=0.9999479755038944\n",
            "test F-measure=0.1038605457698275, IoU=0.05477474008471313\n",
            "[epoch 137]\n",
            "train mean loss=0.12646853923797607\n",
            "train accuracy=0.998260498046875, train recall=0.5420219244823387\n",
            "train precision=0.8476190476190476, specificity=0.9996938654462102\n",
            "train F-measure=0.661218424962853, IoU=0.4938956714761376\n",
            "test mean loss=1.2905200719833374\n",
            "test accuracy=0.9484459307553957, recall=0.05867805755395684\n",
            "test precision=0.9770053475935829, specificity=0.9999201052381236\n",
            "test F-measure=0.11070714415560806, IoU=0.05859713268546137\n",
            "[epoch 138]\n",
            "train mean loss=0.1224067360162735\n",
            "train accuracy=0.9989356994628906, train recall=0.5193798449612403\n",
            "train precision=0.6836734693877551, specificity=0.9996447086419847\n",
            "train F-measure=0.5903083700440529, IoU=0.41875\n",
            "test mean loss=1.2843577861785889\n",
            "test accuracy=0.9484547127922662, recall=0.05941675231243577\n",
            "test precision=0.968079539508111, specificity=0.9998866609191985\n",
            "test F-measure=0.11196175144491179, IoU=0.05930057377311921\n",
            "[epoch 139]\n",
            "train mean loss=0.12056989222764969\n",
            "train accuracy=0.9988861083984375, train recall=0.538135593220339\n",
            "train precision=0.774390243902439, specificity=0.9997172032162401\n",
            "train F-measure=0.6349999999999999, IoU=0.4652014652014652\n",
            "test mean loss=1.2856017649173737\n",
            "test accuracy=0.9483528411645683, recall=0.05752183967112025\n",
            "test precision=0.9675850891410048, specificity=0.9998885189369165\n",
            "test F-measure=0.1085882317276503, IoU=0.05741120656494422\n",
            "[epoch 140]\n",
            "train mean loss=0.11914822459220886\n",
            "train accuracy=0.9988212585449219, train recall=0.5508474576271186\n",
            "train precision=0.7282913165266106, specificity=0.9996293069185851\n",
            "train F-measure=0.6272617611580217, IoU=0.45694200351493847\n",
            "test mean loss=1.289062574505806\n",
            "test accuracy=0.9482000337230215, recall=0.05421377183967112\n",
            "test precision=0.9745958429561201, specificity=0.9999182472204055\n",
            "test F-measure=0.10271388584641596, IoU=0.0541372674791533\n",
            "[epoch 141]\n",
            "train mean loss=0.11701714247465134\n",
            "train accuracy=0.9996414184570312, train recall=0.9591836734693877\n",
            "train precision=0.5108695652173914, specificity=0.9996565488502018\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.2965262234210968\n",
            "test accuracy=0.9478505086555755, recall=0.046923175745118194\n",
            "test precision=0.989167230873392, specificity=0.9999702717165111\n",
            "test F-measure=0.08959617330512373, IoU=0.04689907550077042\n",
            "[epoch 142]\n",
            "train mean loss=0.11470021307468414\n",
            "train accuracy=0.9994010925292969, train recall=0.43137254901960786\n",
            "train precision=0.6821705426356589, specificity=0.9998434756051005\n",
            "train F-measure=0.5285285285285285, IoU=0.35918367346938773\n",
            "test mean loss=1.2999188005924225\n",
            "test accuracy=0.9476397397706835, recall=0.042844295991778004\n",
            "test precision=0.9932985852568875, specificity=0.9999832778405375\n",
            "test F-measure=0.08214538624957665, IoU=0.04283191523519024\n",
            "[epoch 143]\n",
            "train mean loss=0.11340097337961197\n",
            "train accuracy=0.9998626708984375, train recall=0.4852941176470588\n",
            "train precision=0.9705882352941176, specificity=0.9999961843129473\n",
            "train F-measure=0.6470588235294118, IoU=0.4782608695652174\n",
            "test mean loss=1.295261800289154\n",
            "test accuracy=0.9476783807329137, recall=0.043486639260020554\n",
            "test precision=0.9948567229977957, specificity=0.9999869938759736\n",
            "test F-measure=0.083330768994061, IoU=0.043476864785023925\n",
            "[epoch 144]\n",
            "train mean loss=0.114649198949337\n",
            "train accuracy=0.9985198974609375, train recall=0.4924137931034483\n",
            "train precision=0.946949602122016, specificity=0.9999234944667373\n",
            "train F-measure=0.6479128856624319, IoU=0.4791946308724832\n",
            "test mean loss=1.2764806747436523\n",
            "test accuracy=0.9479190085431655, recall=0.048175745118191164\n",
            "test precision=0.9894459102902374, specificity=0.9999702717165111\n",
            "test F-measure=0.09187798603454612, IoU=0.04815100154083205\n",
            "[epoch 145]\n",
            "train mean loss=0.11779612302780151\n",
            "train accuracy=0.9973297119140625, train recall=0.4693396226415094\n",
            "train precision=0.9598070739549839, specificity=0.9999041675611028\n",
            "train F-measure=0.6304118268215417, IoU=0.4602929838087895\n",
            "test mean loss=1.247177615761757\n",
            "test accuracy=0.9483545975719424, recall=0.05761819116135663\n",
            "test precision=0.9665948275862069, specificity=0.9998848029014805\n",
            "test F-measure=0.10875363724539282, IoU=0.05750368613372652\n",
            "[epoch 146]\n",
            "train mean loss=0.12215487658977509\n",
            "train accuracy=0.9959259033203125, train recall=0.4985294117647059\n",
            "train precision=0.9576271186440678, specificity=0.999826992280011\n",
            "train F-measure=0.655705996131528, IoU=0.48776978417266187\n",
            "test mean loss=1.219673678278923\n",
            "test accuracy=0.9486584560476619, recall=0.06558324768756424\n",
            "test precision=0.9371271225332721, specificity=0.9997454515726262\n",
            "test F-measure=0.12258742308269548, IoU=0.0652959421865507\n",
            "[epoch 147]\n",
            "train mean loss=0.11002276837825775\n",
            "train accuracy=0.9989051818847656, train recall=0.9787234042553191\n",
            "train precision=0.2440318302387268, specificity=0.9989124212936462\n",
            "train F-measure=0.3906581740976645, IoU=0.24274406332453827\n",
            "test mean loss=1.206914871931076\n",
            "test accuracy=0.9486830457508992, recall=0.06551901336073998\n",
            "test precision=0.944007403979639, specificity=0.9997751798561151\n",
            "test F-measure=0.12253356158212451, IoU=0.06526538055475574\n",
            "[epoch 148]\n",
            "train mean loss=0.11180861294269562\n",
            "train accuracy=0.9976119995117188, train recall=0.641112047177759\n",
            "train precision=0.7918834547346514, specificity=0.9992335902083485\n",
            "train F-measure=0.7085661080074488, IoU=0.5486661860129777\n",
            "test mean loss=1.1917036473751068\n",
            "test accuracy=0.9486215714928058, recall=0.06259635149023639\n",
            "test precision=0.9677259185700099, specificity=0.9998792288483263\n",
            "test F-measure=0.11758672699849171, IoU=0.062465946604275506\n",
            "[epoch 149]\n",
            "train mean loss=0.11621452867984772\n",
            "train accuracy=0.9963607788085938, train recall=0.5553299492385787\n",
            "train precision=0.9334470989761092, specificity=0.9997002006349597\n",
            "train F-measure=0.6963717377466582, IoU=0.5341796875\n",
            "test mean loss=1.1818197071552277\n",
            "test accuracy=0.9484547127922662, recall=0.05861382322713258\n",
            "test precision=0.9806555615260613, specificity=0.9999331113621499\n",
            "test F-measure=0.11061611661666214, IoU=0.05854613114333376\n",
            "[epoch 150]\n",
            "train mean loss=0.10369895398616791\n",
            "train accuracy=0.9992713928222656, train recall=0.6278195488721805\n",
            "train precision=0.6447876447876448, specificity=0.9996486913753733\n",
            "train F-measure=0.6361904761904762, IoU=0.4664804469273743\n",
            "test mean loss=1.1842694133520126\n",
            "test accuracy=0.9482053029451439, recall=0.053410842754367936\n",
            "test precision=0.9904705181655747, specificity=0.9999702717165111\n",
            "test F-measure=0.10135608715526437, IoU=0.0533834103749358\n",
            "[epoch 151]\n",
            "train mean loss=0.10314909368753433\n",
            "train accuracy=0.9991416931152344, train recall=0.4728682170542636\n",
            "train precision=0.8970588235294118, specificity=0.9999197729191578\n",
            "train F-measure=0.6192893401015228, IoU=0.4485294117647059\n",
            "test mean loss=1.1890926361083984\n",
            "test accuracy=0.9479998032823741, recall=0.049524665981500515\n",
            "test precision=0.9922779922779923, specificity=0.9999777037873833\n",
            "test F-measure=0.09434077699602325, IoU=0.049505586233466035\n",
            "[epoch 152]\n",
            "train mean loss=0.1013035774230957\n",
            "train accuracy=0.9989891052246094, train recall=0.5407554671968191\n",
            "train precision=0.8888888888888888, specificity=0.9998700509476726\n",
            "train F-measure=0.6724351050679852, IoU=0.5065176908752328\n",
            "test mean loss=1.1873491406440735\n",
            "test accuracy=0.9480243929856115, recall=0.04994218910585817\n",
            "test precision=0.9929757343550447, specificity=0.9999795618051014\n",
            "test F-measure=0.09510121705094488, IoU=0.049924551321154524\n",
            "[epoch 153]\n",
            "train mean loss=0.10024435818195343\n",
            "train accuracy=0.998992919921875, train recall=0.42857142857142855\n",
            "train precision=0.9402985074626866, specificity=0.9999541464943085\n",
            "train F-measure=0.5887850467289719, IoU=0.41721854304635764\n",
            "test mean loss=1.1757422238588333\n",
            "test accuracy=0.9482369182778777, recall=0.05402106885919836\n",
            "test precision=0.9899941141848146, specificity=0.9999684136987931\n",
            "test F-measure=0.10245165220039593, IoU=0.053991589895034185\n",
            "[epoch 154]\n",
            "train mean loss=0.10222800076007843\n",
            "train accuracy=0.9994392395019531, train recall=0.06818181818181818\n",
            "train precision=0.2727272727272727, specificity=0.9999084011419325\n",
            "train F-measure=0.10909090909090909, IoU=0.057692307692307696\n",
            "test mean loss=1.149950385093689\n",
            "test accuracy=0.9485425331609713, recall=0.06041238437821172\n",
            "test precision=0.9781591263650546, specificity=0.9999219632558416\n",
            "test F-measure=0.11379654557004146, IoU=0.06033100263005966\n",
            "[epoch 155]\n",
            "train mean loss=0.10332703590393066\n",
            "train accuracy=0.997528076171875, train recall=0.5880503144654088\n",
            "train precision=0.8577981651376146, specificity=0.9995246711030698\n",
            "train F-measure=0.6977611940298508, IoU=0.5358166189111748\n",
            "test mean loss=1.129502296447754\n",
            "test accuracy=0.9487497892311151, recall=0.06535842754367935\n",
            "test precision=0.9630856601987695, specificity=0.9998550746179916\n",
            "test F-measure=0.12240969653222657, IoU=0.06519510476068431\n",
            "[epoch 156]\n",
            "train mean loss=0.0968777984380722\n",
            "train accuracy=0.9993629455566406, train recall=0.9629629629629629\n",
            "train precision=0.3895131086142322, specificity=0.9993779480682043\n",
            "train F-measure=0.5546666666666666, IoU=0.3837638376383764\n",
            "test mean loss=1.1251875758171082\n",
            "test accuracy=0.9487076354541367, recall=0.06455549845837616\n",
            "test precision=0.9631049353138477, specificity=0.9998569326357096\n",
            "test F-measure=0.12100051169370617, IoU=0.06439624515426265\n",
            "[epoch 157]\n",
            "train mean loss=0.0959644466638565\n",
            "train accuracy=0.9994354248046875, train recall=0.7380952380952381\n",
            "train precision=0.44711538461538464, specificity=0.9995610988558038\n",
            "train F-measure=0.5568862275449102, IoU=0.38589211618257263\n",
            "test mean loss=1.1318428963422775\n",
            "test accuracy=0.9485021357913669, recall=0.05983427543679342\n",
            "test precision=0.9753926701570681, specificity=0.9999126731672513\n",
            "test F-measure=0.11275192156388066, IoU=0.059744091331815415\n",
            "[epoch 158]\n",
            "train mean loss=0.09434890747070312\n",
            "train accuracy=0.9989814758300781, train recall=0.5826271186440678\n",
            "train precision=0.7971014492753623, specificity=0.9997324895288758\n",
            "train F-measure=0.6731946144430844, IoU=0.507380073800738\n",
            "test mean loss=1.14690263569355\n",
            "test accuracy=0.9482509695368705, recall=0.05434224049331963\n",
            "test precision=0.988895382817066, specificity=0.9999646976633569\n",
            "test F-measure=0.1030231071330715, IoU=0.054309099662975445\n",
            "[epoch 159]\n",
            "train mean loss=0.10277460515499115\n",
            "train accuracy=0.9964027404785156, train recall=0.5177304964539007\n",
            "train precision=0.941468253968254, specificity=0.9997733480336981\n",
            "train F-measure=0.6680746216121084, IoU=0.5015856236786469\n",
            "test mean loss=1.1433572620153427\n",
            "test accuracy=0.9481526107239209, recall=0.05222250770811922\n",
            "test precision=0.9944954128440368, specificity=0.9999832778405375\n",
            "test F-measure=0.09923407891123248, IoU=0.05220741692085407\n",
            "[epoch 160]\n",
            "train mean loss=0.09835612028837204\n",
            "train accuracy=0.9972419738769531, train recall=0.5373831775700935\n",
            "train precision=0.9640718562874252, specificity=0.9998849013604659\n",
            "train F-measure=0.690098585512216, IoU=0.5268324607329843\n",
            "test mean loss=1.1169197857379913\n",
            "test accuracy=0.9483844564973022, recall=0.05668679342240494\n",
            "test precision=0.9910162829870859, specificity=0.9999702717165111\n",
            "test F-measure=0.10723942036030017, IoU=0.05665767847971238\n",
            "[epoch 161]\n",
            "train mean loss=0.09221141040325165\n",
            "train accuracy=0.9990234375, train recall=0.6080508474576272\n",
            "train precision=0.8016759776536313, specificity=0.9997286679507169\n",
            "train F-measure=0.691566265060241, IoU=0.5285451197053407\n",
            "test mean loss=1.1103178709745407\n",
            "test accuracy=0.9484880845323741, recall=0.05871017471736896\n",
            "test precision=0.9891774891774892, specificity=0.9999628396456388\n",
            "test F-measure=0.11084162017948095, IoU=0.05867248684041597\n",
            "[epoch 162]\n",
            "train mean loss=0.09226982295513153\n",
            "train accuracy=0.9994850158691406, train recall=0.10135135135135136\n",
            "train precision=0.8823529411764706, specificity=0.9999923662956687\n",
            "train F-measure=0.18181818181818185, IoU=0.1\n",
            "test mean loss=1.090431496500969\n",
            "test accuracy=0.9488200455260791, recall=0.06580806783144913\n",
            "test precision=0.9752498810090433, specificity=0.999903383078661\n",
            "test F-measure=0.12329632638324758, IoU=0.06569834551750674\n",
            "[epoch 163]\n",
            "train mean loss=0.09911637008190155\n",
            "train accuracy=0.9966773986816406, train recall=0.5941043083900227\n",
            "train precision=0.8711554447215295, specificity=0.9994047161840387\n",
            "train F-measure=0.7064374789349511, IoU=0.546117769671704\n",
            "test mean loss=1.0758594572544098\n",
            "test accuracy=0.948923673561151, recall=0.06773509763617677\n",
            "test precision=0.9754856614246068, specificity=0.999901525060943\n",
            "test F-measure=0.12667427473121506, IoU=0.06761999422873449\n",
            "[epoch 164]\n",
            "train mean loss=0.0900384783744812\n",
            "train accuracy=0.9993934631347656, train recall=0.8255813953488372\n",
            "train precision=0.3302325581395349, specificity=0.9994505033236917\n",
            "train F-measure=0.4717607973421928, IoU=0.30869565217391304\n",
            "test mean loss=1.078219696879387\n",
            "test accuracy=0.9488235583408273, recall=0.0651978417266187\n",
            "test precision=0.984958757884522, specificity=0.9999424014507402\n",
            "test F-measure=0.12230020785010691, IoU=0.06513299323001893\n",
            "[epoch 165]\n",
            "train mean loss=0.08800311386585236\n",
            "train accuracy=0.9995689392089844, train recall=0.6916666666666667\n",
            "train precision=0.5220125786163522, specificity=0.9997099502335663\n",
            "train F-measure=0.5949820788530467, IoU=0.42346938775510207\n",
            "test mean loss=1.0871794074773788\n",
            "test accuracy=0.948556584419964, recall=0.05977004110996917\n",
            "test precision=0.9925333333333334, specificity=0.9999739877519472\n",
            "test F-measure=0.11275029535609342, IoU=0.05974317817014446\n",
            "[epoch 166]\n",
            "train mean loss=0.09199965745210648\n",
            "train accuracy=0.9976387023925781, train recall=0.5671232876712329\n",
            "train precision=0.8107049608355091, specificity=0.999444548724569\n",
            "train F-measure=0.6673831273508867, IoU=0.5008064516129033\n",
            "test mean loss=1.090068742632866\n",
            "test accuracy=0.9483458155350719, recall=0.05591598150051388\n",
            "test precision=0.992022792022792, specificity=0.9999739877519472\n",
            "test F-measure=0.10586482624426136, IoU=0.0558908507223114\n",
            "[epoch 167]\n",
            "train mean loss=0.08580204844474792\n",
            "train accuracy=0.9995079040527344, train recall=0.07575757575757576\n",
            "train precision=0.5882352941176471, specificity=0.9999732836663969\n",
            "train F-measure=0.1342281879194631, IoU=0.07194244604316546\n",
            "test mean loss=1.0789822936058044\n",
            "test accuracy=0.9484547127922662, recall=0.05809994861253854\n",
            "test precision=0.9890650628758885, specificity=0.9999628396456388\n",
            "test F-measure=0.1097527680873654, IoU=0.05806265245859545\n",
            "[epoch 168]\n",
            "train mean loss=0.08934855461120605\n",
            "train accuracy=0.9975318908691406, train recall=0.584045584045584\n",
            "train precision=0.928652321630804, specificity=0.9997583799953977\n",
            "train F-measure=0.7170966331438565, IoU=0.5589638718473074\n",
            "test mean loss=1.0608696788549423\n",
            "test accuracy=0.94873046875, recall=0.06352774922918808\n",
            "test precision=0.9840796019900497, specificity=0.9999405434330222\n",
            "test F-measure=0.11935075122186689, IoU=0.06346252566735112\n",
            "[epoch 169]\n",
            "train mean loss=0.08368302881717682\n",
            "train accuracy=0.999725341796875, train recall=0.5425531914893617\n",
            "train precision=0.6375, specificity=0.9998893340965465\n",
            "train F-measure=0.5862068965517241, IoU=0.4146341463414634\n",
            "test mean loss=1.0476011633872986\n",
            "test accuracy=0.9489078658947842, recall=0.06728545734840699\n",
            "test precision=0.9776014932337844, specificity=0.9999108151495333\n",
            "test F-measure=0.1259052255175937, IoU=0.06718188814776808\n",
            "[epoch 170]\n",
            "train mean loss=0.0827268660068512\n",
            "train accuracy=0.9988555908203125, train recall=0.5434083601286174\n",
            "train precision=0.5168195718654435, specificity=0.9993965619306963\n",
            "train F-measure=0.5297805642633229, IoU=0.3603411513859275\n",
            "test mean loss=1.0437124371528625\n",
            "test accuracy=0.9488147763039568, recall=0.06555113052415211\n",
            "test precision=0.9774904214559387, specificity=0.9999126731672513\n",
            "test F-measure=0.12286299060919817, IoU=0.06545232979508066\n",
            "[epoch 171]\n",
            "train mean loss=0.08109097927808762\n",
            "train accuracy=0.9995193481445312, train recall=0.8968253968253969\n",
            "train precision=0.5, specificity=0.9995687319191812\n",
            "train F-measure=0.6420454545454546, IoU=0.47280334728033474\n",
            "test mean loss=1.0545333474874496\n",
            "test accuracy=0.9484775460881295, recall=0.05867805755395684\n",
            "test precision=0.9865010799136069, specificity=0.9999535495570486\n",
            "test F-measure=0.11076755183703164, IoU=0.058630981033984786\n",
            "[epoch 172]\n",
            "train mean loss=0.07945854961872101\n",
            "train accuracy=0.99951171875, train recall=0.3048780487804878\n",
            "train precision=0.78125, specificity=0.9999465608061684\n",
            "train F-measure=0.43859649122807015, IoU=0.2808988764044944\n",
            "test mean loss=1.0589347630739212\n",
            "test accuracy=0.9483440591276978, recall=0.05607656731757451\n",
            "test precision=0.9886749716874292, specificity=0.9999628396456388\n",
            "test F-measure=0.10613336575284177, IoU=0.05604057003466427\n",
            "[epoch 173]\n",
            "train mean loss=0.08051469922065735\n",
            "train accuracy=0.9998588562011719, train recall=0.21875\n",
            "train precision=0.3684210526315789, specificity=0.9999542180441948\n",
            "train F-measure=0.2745098039215686, IoU=0.1590909090909091\n",
            "test mean loss=1.0600603520870209\n",
            "test accuracy=0.9483545975719424, recall=0.05646197327852004\n",
            "test precision=0.9854260089686099, specificity=0.9999516915393305\n",
            "test F-measure=0.10680437424058323, IoU=0.056414864257749824\n",
            "[epoch 174]\n",
            "train mean loss=0.07867270708084106\n",
            "train accuracy=0.9990768432617188, train recall=0.6598890942698706\n",
            "train precision=0.8602409638554217, specificity=0.9997782900043195\n",
            "train F-measure=0.7468619246861925, IoU=0.5959933222036727\n",
            "test mean loss=1.0658840835094452\n",
            "test accuracy=0.9483704052383094, recall=0.05681526207605344\n",
            "test precision=0.9844184752365053, specificity=0.9999479755038944\n",
            "test F-measure=0.10743023714814927, IoU=0.056764215120010265\n",
            "[epoch 175]\n",
            "train mean loss=0.0769791305065155\n",
            "train accuracy=0.9993324279785156, train recall=0.5671641791044776\n",
            "train precision=0.8636363636363636, specificity=0.9998854126481519\n",
            "train F-measure=0.6846846846846847, IoU=0.5205479452054794\n",
            "test mean loss=1.069017544388771\n",
            "test accuracy=0.9484634948291367, recall=0.05871017471736896\n",
            "test precision=0.9817400644468314, specificity=0.999936827397586\n",
            "test F-measure=0.11079459361173404, IoU=0.05864613410330446\n",
            "[epoch 176]\n",
            "train mean loss=0.07908180356025696\n",
            "train accuracy=0.9983749389648438, train recall=0.5371498172959805\n",
            "train precision=0.9055441478439425, specificity=0.9998239726315709\n",
            "train F-measure=0.6743119266055047, IoU=0.5086505190311419\n",
            "test mean loss=1.0570931732654572\n",
            "test accuracy=0.9486654816771583, recall=0.06288540596094552\n",
            "test precision=0.9755854509217738, specificity=0.9999089571318152\n",
            "test F-measure=0.11815466312645202, IoU=0.06278659611992946\n",
            "[epoch 177]\n",
            "train mean loss=0.08351494371891022\n",
            "train accuracy=0.9972686767578125, train recall=0.5575465196416265\n",
            "train precision=0.9161947904869762, specificity=0.9997161412082411\n",
            "train F-measure=0.6932305055698371, IoU=0.5304918032786885\n",
            "test mean loss=1.0351969450712204\n",
            "test accuracy=0.9488200455260791, recall=0.06673946557040082\n",
            "test precision=0.9624826308476147, specificity=0.9998495005648373\n",
            "test F-measure=0.12482354707914102, IoU=0.06656629400647084\n",
            "[epoch 178]\n",
            "train mean loss=0.08034570515155792\n",
            "train accuracy=0.9975318908691406, train recall=0.6274928774928775\n",
            "train precision=0.8766169154228856, specificity=0.999524430467132\n",
            "train F-measure=0.7314238273142384, IoU=0.5765706806282722\n",
            "test mean loss=1.0149839967489243\n",
            "test accuracy=0.9489025966726619, recall=0.06828108941418294\n",
            "test precision=0.9628623188405797, specificity=0.9998476425471193\n",
            "test F-measure=0.12751919385796545, IoU=0.06810173617784611\n",
            "[epoch 179]\n",
            "train mean loss=0.07400690764188766\n",
            "train accuracy=0.9995994567871094, train recall=0.9705882352941176\n",
            "train precision=0.4925373134328358, specificity=0.9996107494218484\n",
            "train F-measure=0.6534653465346534, IoU=0.4852941176470588\n",
            "test mean loss=1.0119670629501343\n",
            "test accuracy=0.948748032823741, recall=0.0643627954779034\n",
            "test precision=0.9766081871345029, specificity=0.9999108151495333\n",
            "test F-measure=0.12076654212365917, IoU=0.06426372498717291\n",
            "[epoch 180]\n",
            "train mean loss=0.07323379069566727\n",
            "train accuracy=0.9992713928222656, train recall=0.5940298507462687\n",
            "train precision=0.7834645669291339, specificity=0.9997899231882785\n",
            "train F-measure=0.6757215619694397, IoU=0.5102564102564102\n",
            "test mean loss=1.0172520130872726\n",
            "test accuracy=0.9485144306429856, recall=0.05948098663926002\n",
            "test precision=0.9845826687931951, specificity=0.9999461174861763\n",
            "test F-measure=0.11218463215919071, IoU=0.059425637734638215\n",
            "[epoch 181]\n",
            "train mean loss=0.07697667181491852\n",
            "train accuracy=0.9978713989257812, train recall=0.6022012578616353\n",
            "train precision=0.9364303178484108, specificity=0.9998006685270937\n",
            "train F-measure=0.7330143540669858, IoU=0.5785498489425982\n",
            "test mean loss=1.0097828209400177\n",
            "test accuracy=0.9485706356789568, recall=0.06054085303186023\n",
            "test precision=0.9843342036553525, specificity=0.9999442594684583\n",
            "test F-measure=0.11406614020755802, IoU=0.06048257716742604\n",
            "[epoch 182]\n",
            "train mean loss=0.07665171474218369\n",
            "train accuracy=0.9976119995117188, train recall=0.5440251572327044\n",
            "train precision=0.9376693766937669, specificity=0.9998236683124291\n",
            "train F-measure=0.6885572139303483, IoU=0.5250379362670713\n",
            "test mean loss=0.9955172687768936\n",
            "test accuracy=0.9488499044514388, recall=0.06744604316546762\n",
            "test precision=0.9606587374199451, specificity=0.999840210476247\n",
            "test F-measure=0.126042854570554, IoU=0.06726026519761706\n",
            "[epoch 183]\n",
            "train mean loss=0.07216918468475342\n",
            "train accuracy=0.9995765686035156, train recall=0.7045454545454546\n",
            "train precision=0.5636363636363636, specificity=0.9997252034257973\n",
            "train F-measure=0.6262626262626263, IoU=0.45588235294117646\n",
            "test mean loss=0.9822894632816315\n",
            "test accuracy=0.9489623145233813, recall=0.06991906474820143\n",
            "test precision=0.9565026362038664, specificity=0.9998160562459124\n",
            "test F-measure=0.13031246258829163, IoU=0.06969745477829357\n",
            "[epoch 184]\n",
            "train mean loss=0.07544787228107452\n",
            "train accuracy=0.9974861145019531, train recall=0.6353996737357259\n",
            "train precision=0.786074672048436, specificity=0.9991874841904353\n",
            "train F-measure=0.7027514659449706, IoU=0.5417246175243393\n",
            "test mean loss=0.9711053669452667\n",
            "test accuracy=0.9489148915242805, recall=0.06709275436793423\n",
            "test precision=0.9821344616831218, specificity=0.9999293953267139\n",
            "test F-measure=0.12560502660613895, IoU=0.06701097068069545\n",
            "[epoch 185]\n",
            "train mean loss=0.0701448917388916\n",
            "train accuracy=0.9987907409667969, train recall=0.7287705956907478\n",
            "train precision=0.8480825958702065, specificity=0.9996059000210442\n",
            "train F-measure=0.7839127471029311, IoU=0.6446188340807175\n",
            "test mean loss=0.9742099940776825\n",
            "test accuracy=0.9486233279001799, recall=0.06108684480986639\n",
            "test precision=0.9911412193850964, specificity=0.9999684136987931\n",
            "test F-measure=0.11508092572984419, IoU=0.061053510095335925\n",
            "[epoch 186]\n",
            "train mean loss=0.06818334758281708\n",
            "train accuracy=0.9998779296875, train recall=0.54\n",
            "train precision=0.75, specificity=0.9999656611749983\n",
            "train F-measure=0.627906976744186, IoU=0.4576271186440678\n",
            "test mean loss=0.9704914838075638\n",
            "test accuracy=0.9485794177158273, recall=0.06012332990750257\n",
            "test precision=0.9936305732484076, specificity=0.9999777037873833\n",
            "test F-measure=0.11338582677165354, IoU=0.06010016694490818\n",
            "[epoch 187]\n",
            "train mean loss=0.06862186640501022\n",
            "train accuracy=0.9989509582519531, train recall=0.524547803617571\n",
            "train precision=0.6904761904761905, specificity=0.9996523493163506\n",
            "train F-measure=0.5961820851688693, IoU=0.4246861924686193\n",
            "test mean loss=0.9653942137956619\n",
            "test accuracy=0.9486321099370504, recall=0.06118319630010278\n",
            "test precision=0.9921875, specificity=0.9999721297342291\n",
            "test F-measure=0.11525895450145207, IoU=0.06115373503258322\n",
            "[epoch 188]\n",
            "train mean loss=0.06666412204504013\n",
            "train accuracy=0.9993820190429688, train recall=0.5918367346938775\n",
            "train precision=0.7004830917874396, specificity=0.9997632675191581\n",
            "train F-measure=0.6415929203539823, IoU=0.4723127035830619\n",
            "test mean loss=0.9690903276205063\n",
            "test accuracy=0.9485636100494604, recall=0.05989850976361768\n",
            "test precision=0.9925492283129325, specificity=0.9999739877519472\n",
            "test F-measure=0.11297894896259275, IoU=0.05987158908507223\n",
            "[epoch 189]\n",
            "train mean loss=0.0668107196688652\n",
            "train accuracy=0.9989013671875, train recall=0.6564019448946515\n",
            "train precision=0.841995841995842, specificity=0.9997093990295457\n",
            "train F-measure=0.7377049180327869, IoU=0.5844155844155844\n",
            "test mean loss=0.9794354885816574\n",
            "test accuracy=0.9484828153102518, recall=0.05832476875642343\n",
            "test precision=0.9939792008757526, specificity=0.9999795618051014\n",
            "test F-measure=0.11018414586051026, IoU=0.05830417054612001\n",
            "[epoch 190]\n",
            "train mean loss=0.06481131911277771\n",
            "train accuracy=0.9997444152832031, train recall=0.32978723404255317\n",
            "train precision=0.8857142857142857, specificity=0.9999847357374547\n",
            "train F-measure=0.4806201550387597, IoU=0.3163265306122449\n",
            "test mean loss=0.9829018115997314\n",
            "test accuracy=0.9485302383093526, recall=0.05919193216855087\n",
            "test precision=0.994066882416397, specificity=0.9999795618051014\n",
            "test F-measure=0.11173082752349196, IoU=0.05917102770732334\n",
            "[epoch 191]\n",
            "train mean loss=0.06540509313344955\n",
            "train accuracy=0.9990768432617188, train recall=0.4835820895522388\n",
            "train precision=0.7012987012987013, specificity=0.9997364490907493\n",
            "train F-measure=0.5724381625441697, IoU=0.400990099009901\n",
            "test mean loss=0.9843399673700333\n",
            "test accuracy=0.9485390203462231, recall=0.05948098663926002\n",
            "test precision=0.9919657204070702, specificity=0.9999721297342291\n",
            "test F-measure=0.11223222131321395, IoU=0.059452345029052035\n",
            "[epoch 192]\n",
            "train mean loss=0.06961853057146072\n",
            "train accuracy=0.9976997375488281, train recall=0.5823817292006526\n",
            "train precision=0.8869565217391304, specificity=0.9996512314213661\n",
            "train F-measure=0.7031019202363368, IoU=0.5421412300683371\n",
            "test mean loss=0.9618377089500427\n",
            "test accuracy=0.9488077506744604, recall=0.06478031860226105\n",
            "test precision=0.9867906066536204, specificity=0.9999498335216125\n",
            "test F-measure=0.12157926461723931, IoU=0.06472419215094824\n",
            "[epoch 193]\n",
            "train mean loss=0.06372568756341934\n",
            "train accuracy=0.9993934631347656, train recall=0.20175438596491227\n",
            "train precision=0.25274725274725274, specificity=0.9997404877304126\n",
            "train F-measure=0.22439024390243903, IoU=0.12637362637362637\n",
            "test mean loss=0.951337531208992\n",
            "test accuracy=0.9488551736735612, recall=0.0659044193216855\n",
            "test precision=0.9832295160517489, specificity=0.999934969379868\n",
            "test F-measure=0.1235288805947687, IoU=0.06583041929999038\n",
            "[epoch 194]\n",
            "train mean loss=0.069453664124012\n",
            "train accuracy=0.9974822998046875, train recall=0.6330882352941176\n",
            "train precision=0.8424657534246576, specificity=0.9993826308362477\n",
            "train F-measure=0.7229219143576827, IoU=0.5660749506903353\n",
            "test mean loss=0.9507795572280884\n",
            "test accuracy=0.948695340602518, recall=0.06291752312435765\n",
            "test precision=0.983433734939759, specificity=0.9999386854153042\n",
            "test F-measure=0.1182685341704902, IoU=0.06285090955757323\n",
            "[epoch 195]\n",
            "train mean loss=0.06975681334733963\n",
            "train accuracy=0.997283935546875, train recall=0.6174899866488651\n",
            "train precision=0.8693609022556391, specificity=0.9994667096368254\n",
            "train F-measure=0.7220921155347384, IoU=0.5650580329871716\n",
            "test mean loss=0.9568419605493546\n",
            "test accuracy=0.9484968665692446, recall=0.05919193216855087\n",
            "test precision=0.9839829151094501, specificity=0.9999442594684583\n",
            "test F-measure=0.1116665151928262, IoU=0.059134954758390555\n",
            "[epoch 196]\n",
            "train mean loss=0.06752590090036392\n",
            "train accuracy=0.9976387023925781, train recall=0.6139705882352942\n",
            "train precision=0.898815931108719, specificity=0.9996395484385545\n",
            "train F-measure=0.72957623416339, IoU=0.5742778541953233\n",
            "test mean loss=0.960369348526001\n",
            "test accuracy=0.9483914821267986, recall=0.05745760534429599\n",
            "test precision=0.9802739726027397, specificity=0.9999331113621499\n",
            "test F-measure=0.1085525317799824, IoU=0.05739124855639677\n",
            "[epoch 197]\n",
            "train mean loss=0.061503928154706955\n",
            "train accuracy=0.9996147155761719, train recall=0.00980392156862745\n",
            "train precision=1.0, specificity=1.0\n",
            "train F-measure=0.01941747572815534, IoU=0.00980392156862745\n",
            "test mean loss=0.952269896864891\n",
            "test accuracy=0.9485618536420863, recall=0.06131166495375129\n",
            "test precision=0.970020325203252, specificity=0.9998903769546347\n",
            "test F-measure=0.1153334944417593, IoU=0.06119570443981407\n",
            "[epoch 198]\n",
            "train mean loss=0.06508141756057739\n",
            "train accuracy=0.9977836608886719, train recall=0.6458966565349544\n",
            "train precision=0.8808290155440415, specificity=0.9995590964160289\n",
            "train F-measure=0.7452871547566856, IoU=0.593990216631726\n",
            "test mean loss=0.9455903619527817\n",
            "test accuracy=0.9486391355665468, recall=0.06375256937307297\n",
            "test precision=0.9561657032755299, specificity=0.9998309203876569\n",
            "test F-measure=0.11953510779236419, IoU=0.06356678515387325\n",
            "[epoch 199]\n",
            "train mean loss=0.06084158271551132\n",
            "train accuracy=0.9991798400878906, train recall=0.8921568627450981\n",
            "train precision=0.30847457627118646, specificity=0.9992214988436968\n",
            "train F-measure=0.4584382871536524, IoU=0.2973856209150327\n",
            "test mean loss=0.9657478034496307\n",
            "test accuracy=0.9483774308678058, recall=0.05784301130524152\n",
            "test precision=0.9698438341410878, specificity=0.9998959510077888\n",
            "test F-measure=0.10917467341557299, IoU=0.05773916388817645\n",
            "[epoch 200]\n",
            "train mean loss=0.06347004324197769\n",
            "train accuracy=0.9979743957519531, train recall=0.6344339622641509\n",
            "train precision=0.9243986254295533, specificity=0.9997470023613113\n",
            "train F-measure=0.7524475524475525, IoU=0.6031390134529148\n",
            "test mean loss=0.9733565896749496\n",
            "test accuracy=0.9481649055755396, recall=0.05389260020554985\n",
            "test precision=0.9688221709006929, specificity=0.999899667043225\n",
            "test F-measure=0.10210539126201777, IoU=0.05379929464571978\n",
            "[epoch 201]\n",
            "train mean loss=0.061469972133636475\n",
            "train accuracy=0.9982757568359375, train recall=0.6089866156787763\n",
            "train precision=0.9367647058823529, specificity=0.9998353108794399\n",
            "train F-measure=0.7381228273464657, IoU=0.5849403122130394\n",
            "test mean loss=0.9654649496078491\n",
            "test accuracy=0.9482105721672662, recall=0.05501670092497431\n",
            "test precision=0.964527027027027, specificity=0.9998829448837624\n",
            "test F-measure=0.10409577053962081, IoU=0.05490560594890862\n",
            "[epoch 202]\n",
            "train mean loss=0.05915966257452965\n",
            "train accuracy=0.9989471435546875, train recall=0.5254237288135594\n",
            "train precision=0.8266666666666667, specificity=0.9998012779357364\n",
            "train F-measure=0.6424870466321244, IoU=0.4732824427480916\n",
            "test mean loss=0.9513093382120132\n",
            "test accuracy=0.9483598667940647, recall=0.05809994861253854\n",
            "test precision=0.9607010090281466, specificity=0.9998625066888638\n",
            "test F-measure=0.10957327599261031, IoU=0.057962191605254725\n",
            "[epoch 203]\n",
            "train mean loss=0.058012835681438446\n",
            "train accuracy=0.9998435974121094, train recall=0.4772727272727273\n",
            "train precision=0.5384615384615384, specificity=0.9999313239221671\n",
            "train F-measure=0.5060240963855421, IoU=0.3387096774193548\n",
            "test mean loss=0.9365124702453613\n",
            "test accuracy=0.948591712567446, recall=0.06294964028776978\n",
            "test precision=0.9547004383828543, specificity=0.9998272043522207\n",
            "test F-measure=0.11811142245924854, IoU=0.06276217618239457\n",
            "[epoch 204]\n",
            "train mean loss=0.05764571577310562\n",
            "train accuracy=0.9993476867675781, train recall=0.7692307692307693\n",
            "train precision=0.47058823529411764, specificity=0.9994847092233232\n",
            "train F-measure=0.583941605839416, IoU=0.41237113402061853\n",
            "test mean loss=0.9328471124172211\n",
            "test accuracy=0.9486654816771583, recall=0.06429856115107914\n",
            "test precision=0.9556085918854416, specificity=0.9998272043522207\n",
            "test F-measure=0.12048990400529627, IoU=0.0641070799577316\n",
            "[epoch 205]\n",
            "train mean loss=0.05637321621179581\n",
            "train accuracy=0.9997634887695312, train recall=0.6470588235294118\n",
            "train precision=0.5365853658536586, specificity=0.9998550038920008\n",
            "train F-measure=0.5866666666666668, IoU=0.41509433962264153\n",
            "test mean loss=0.9282994419336319\n",
            "test accuracy=0.9487199303057554, recall=0.06545477903391572\n",
            "test precision=0.954119850187266, specificity=0.9998179142636304\n",
            "test F-measure=0.12250540995431593, IoU=0.06524940769674073\n",
            "[epoch 206]\n",
            "train mean loss=0.05555257201194763\n",
            "train accuracy=0.9995956420898438, train recall=0.8642857142857143\n",
            "train precision=0.5817307692307693, specificity=0.9996679440008549\n",
            "train F-measure=0.6954022988505748, IoU=0.5330396475770925\n",
            "test mean loss=0.9331677854061127\n",
            "test accuracy=0.9486075202338129, recall=0.06278905447070915\n",
            "test precision=0.9611602753195674, specificity=0.9998532166002735\n",
            "test F-measure=0.11787760024118181, IoU=0.06263014576325485\n",
            "[epoch 207]\n",
            "train mean loss=0.05469658598303795\n",
            "train accuracy=0.9996871948242188, train recall=0.7621951219512195\n",
            "train precision=0.7440476190476191, specificity=0.9998358653332315\n",
            "train F-measure=0.7530120481927711, IoU=0.6038647342995169\n",
            "test mean loss=0.9393073916435242\n",
            "test accuracy=0.9484986229766187, recall=0.06028391572456321\n",
            "test precision=0.9675257731958763, specificity=0.9998829448837624\n",
            "test F-measure=0.11349619059136534, IoU=0.060162184685406586\n",
            "[epoch 208]\n",
            "train mean loss=0.05367431789636612\n",
            "train accuracy=0.9998245239257812, train recall=0.6511627906976745\n",
            "train precision=0.7777777777777778, specificity=0.9999389448137436\n",
            "train F-measure=0.7088607594936709, IoU=0.5490196078431373\n",
            "test mean loss=0.9418478310108185\n",
            "test accuracy=0.9484230974595323, recall=0.05842112024665982\n",
            "test precision=0.9742903053026245, specificity=0.9999108151495333\n",
            "test F-measure=0.11023240311486834, IoU=0.05833119548486403\n",
            "[epoch 209]\n",
            "train mean loss=0.06274715065956116\n",
            "train accuracy=0.9969558715820312, train recall=0.5667447306791569\n",
            "train precision=0.9434697855750487, specificity=0.9997772965335053\n",
            "train F-measure=0.7081199707388441, IoU=0.5481313703284258\n",
            "test mean loss=0.9223447144031525\n",
            "test accuracy=0.9484951101618705, recall=0.05938463514902364\n",
            "test precision=0.9803817603393425, specificity=0.9999312533444319\n",
            "test F-measure=0.11198594876143178, IoU=0.05931415006576204\n",
            "[epoch 210]\n",
            "train mean loss=0.05308392643928528\n",
            "train accuracy=0.9997520446777344, train recall=0.41836734693877553\n",
            "train precision=0.8367346938775511, specificity=0.9999694710089069\n",
            "train F-measure=0.5578231292517006, IoU=0.3867924528301887\n",
            "test mean loss=0.9038041085004807\n",
            "test accuracy=0.9487058790467626, recall=0.06349563206577595\n",
            "test precision=0.9777448071216617, specificity=0.9999163892026874\n",
            "test F-measure=0.11924724048495083, IoU=0.06340399602321926\n",
            "[epoch 211]\n",
            "train mean loss=0.05340222269296646\n",
            "train accuracy=0.9993209838867188, train recall=0.7174515235457064\n",
            "train precision=0.7731343283582089, specificity=0.9997096832109037\n",
            "train F-measure=0.7442528735632185, IoU=0.5926773455377574\n",
            "test mean loss=0.8977939188480377\n",
            "test accuracy=0.9488042378597122, recall=0.06516572456320657\n",
            "test precision=0.9801932367149758, specificity=0.9999238212735596\n",
            "test F-measure=0.1222068300909474, IoU=0.0650800269429387\n",
            "[epoch 212]\n",
            "train mean loss=0.06094016507267952\n",
            "train accuracy=0.9968605041503906, train recall=0.5499655410062027\n",
            "train precision=0.8243801652892562, specificity=0.9993478919648783\n",
            "train F-measure=0.659776767259198, IoU=0.4922887106724244\n",
            "test mean loss=0.8839047104120255\n",
            "test accuracy=0.948939481227518, recall=0.06815262076053442\n",
            "test precision=0.9738412115649381, specificity=0.9998940929900707\n",
            "test F-measure=0.1273900645354945, IoU=0.06802808322380022\n",
            "[epoch 213]\n",
            "train mean loss=0.05226210877299309\n",
            "train accuracy=0.9994010925292969, train recall=0.5957446808510638\n",
            "train precision=0.32, specificity=0.9995458881892768\n",
            "train F-measure=0.4163568773234201, IoU=0.26291079812206575\n",
            "test mean loss=0.886421337723732\n",
            "test accuracy=0.9488305839703237, recall=0.06580806783144913\n",
            "test precision=0.9780429594272076, specificity=0.9999145311849694\n",
            "test F-measure=0.12331858806536068, IoU=0.06571098710794689\n",
            "[epoch 214]\n",
            "train mean loss=0.0521184578537941\n",
            "train accuracy=0.999267578125, train recall=0.6062717770034843\n",
            "train precision=0.6877470355731226, specificity=0.9996983086188264\n",
            "train F-measure=0.6444444444444445, IoU=0.47540983606557374\n",
            "test mean loss=0.8997538238763809\n",
            "test accuracy=0.9486268407149281, recall=0.06182553956834532\n",
            "test precision=0.9806418746816098, specificity=0.9999293953267139\n",
            "test F-measure=0.1163177135260884, IoU=0.061750176429075515\n",
            "[epoch 215]\n",
            "train mean loss=0.06128924340009689\n",
            "train accuracy=0.9965591430664062, train recall=0.5756345177664974\n",
            "train precision=0.945, specificity=0.9997463236141967\n",
            "train F-measure=0.7154574132492113, IoU=0.5569744597249509\n",
            "test mean loss=0.8964648991823196\n",
            "test accuracy=0.9485478023830936, recall=0.06163283658787256\n",
            "test precision=0.9614228456913828, specificity=0.9998569326357096\n",
            "test F-measure=0.11583967161656405, IoU=0.061480793259218916\n",
            "[epoch 216]\n",
            "train mean loss=0.05705122649669647\n",
            "train accuracy=0.997344970703125, train recall=0.5811965811965812\n",
            "train precision=0.8831168831168831, specificity=0.9995857942778247\n",
            "train F-measure=0.7010309278350516, IoU=0.5396825396825397\n",
            "test mean loss=0.8746925890445709\n",
            "test accuracy=0.9486689944919064, recall=0.06468396711202466\n",
            "test precision=0.9513462446858763, specificity=0.9998086241750401\n",
            "test F-measure=0.12113192794635069, IoU=0.06447069368417684\n",
            "[epoch 217]\n",
            "train mean loss=0.05839623510837555\n",
            "train accuracy=0.9969329833984375, train recall=0.601089588377724\n",
            "train precision=0.8725834797891037, specificity=0.9994433610245228\n",
            "train F-measure=0.7118279569892474, IoU=0.5525876460767947\n",
            "test mean loss=0.8518462181091309\n",
            "test accuracy=0.9487778917491008, recall=0.06696428571428571\n",
            "test precision=0.949021392808375, specificity=0.9997919020155777\n",
            "test F-measure=0.1251012510125101, IoU=0.06672427035330261\n",
            "[epoch 218]\n",
            "train mean loss=0.050964660942554474\n",
            "train accuracy=0.9991683959960938, train recall=0.6897506925207756\n",
            "train precision=0.7014084507042253, specificity=0.9995950844783656\n",
            "train F-measure=0.6955307262569832, IoU=0.5331905781584583\n",
            "test mean loss=0.8423653542995453\n",
            "test accuracy=0.9488218019334532, recall=0.06712487153134636\n",
            "test precision=0.9578368469294225, specificity=0.9998290623699387\n",
            "test F-measure=0.12545771054685156, IoU=0.06692711669014986\n",
            "[epoch 219]\n",
            "train mean loss=0.05092919245362282\n",
            "train accuracy=0.9995765686035156, train recall=0.5\n",
            "train precision=0.6666666666666666, specificity=0.9998587764698698\n",
            "train F-measure=0.5714285714285715, IoU=0.4\n",
            "test mean loss=0.8380738645792007\n",
            "test accuracy=0.9488938146357914, recall=0.0683453237410072\n",
            "test precision=0.9598556608028868, specificity=0.9998346364230929\n",
            "test F-measure=0.1276047132192007, IoU=0.06815052041633307\n",
            "[epoch 220]\n",
            "train mean loss=0.04957057535648346\n",
            "train accuracy=0.9995918273925781, train recall=0.6416666666666667\n",
            "train precision=0.5460992907801419, specificity=0.9997557475651085\n",
            "train F-measure=0.5900383141762452, IoU=0.41847826086956524\n",
            "test mean loss=0.8406306952238083\n",
            "test accuracy=0.9488938146357914, recall=0.06828108941418294\n",
            "test precision=0.9606868504292815, specificity=0.999838352458529\n",
            "test F-measure=0.12750007496476656, IoU=0.06809083047753259\n",
            "[epoch 221]\n",
            "train mean loss=0.052408989518880844\n",
            "train accuracy=0.9980545043945312, train recall=0.663013698630137\n",
            "train precision=0.8373702422145328, specificity=0.9994598715183739\n",
            "train F-measure=0.7400611620795107, IoU=0.587378640776699\n",
            "test mean loss=0.8410905450582504\n",
            "test accuracy=0.9488955710431655, recall=0.06767086330935251\n",
            "test precision=0.96918123275069, specificity=0.9998755128128902\n",
            "test F-measure=0.12650855598919242, IoU=0.0675255584398936\n",
            "[epoch 222]\n",
            "train mean loss=0.053825944662094116\n",
            "train accuracy=0.9975318908691406, train recall=0.6426470588235295\n",
            "train precision=0.8444444444444444, specificity=0.9993826308362477\n",
            "train F-measure=0.7298538622129436, IoU=0.5746219592373438\n",
            "test mean loss=0.8362768441438675\n",
            "test accuracy=0.9489025966726619, recall=0.0671891058581706\n",
            "test precision=0.9775700934579439, specificity=0.9999108151495333\n",
            "test F-measure=0.12573626637817045, IoU=0.06708568496664956\n",
            "[epoch 223]\n",
            "train mean loss=0.04856773838400841\n",
            "train accuracy=0.99981689453125, train recall=0.7837837837837838\n",
            "train precision=0.6444444444444445, specificity=0.9998778952188346\n",
            "train F-measure=0.7073170731707318, IoU=0.5471698113207547\n",
            "test mean loss=0.8355408161878586\n",
            "test accuracy=0.9488147763039568, recall=0.06490878725590955\n",
            "test precision=0.98681640625, specificity=0.9999498335216125\n",
            "test F-measure=0.12180568948891032, IoU=0.06485254949780188\n",
            "[epoch 224]\n",
            "train mean loss=0.047939568758010864\n",
            "train accuracy=0.9994888305664062, train recall=0.30128205128205127\n",
            "train precision=0.6527777777777778, specificity=0.9999045757820969\n",
            "train F-measure=0.412280701754386, IoU=0.2596685082872928\n",
            "test mean loss=0.8302605301141739\n",
            "test accuracy=0.9489148915242805, recall=0.06699640287769784\n",
            "test precision=0.9834983498349835, specificity=0.999934969379868\n",
            "test F-measure=0.12544727425805094, IoU=0.06692117673478554\n",
            "[epoch 225]\n",
            "train mean loss=0.05285973474383354\n",
            "train accuracy=0.9979896545410156, train recall=0.6858974358974359\n",
            "train precision=0.9180171591992373, specificity=0.9996701695175271\n",
            "train F-measure=0.7851610273134936, IoU=0.6463087248322148\n",
            "test mean loss=0.8272551745176315\n",
            "test accuracy=0.9490712117805755, recall=0.06998329907502569\n",
            "test precision=0.9824165915238954, specificity=0.9999275373089958\n",
            "test F-measure=0.13065899142531628, IoU=0.06989574979951885\n",
            "[epoch 226]\n",
            "train mean loss=0.04793776571750641\n",
            "train accuracy=0.9993782043457031, train recall=0.7303921568627451\n",
            "train precision=0.5797665369649806, specificity=0.9995876918378255\n",
            "train F-measure=0.6464208242950108, IoU=0.4775641025641026\n",
            "test mean loss=0.83636873960495\n",
            "test accuracy=0.9490237887814749, recall=0.06895554984583761\n",
            "test precision=0.9844108207244383, specificity=0.999936827397586\n",
            "test F-measure=0.128883152744845, IoU=0.06888033365415464\n",
            "[epoch 227]\n",
            "train mean loss=0.04683707654476166\n",
            "train accuracy=0.9995040893554688, train recall=0.7621951219512195\n",
            "train precision=0.5787037037037037, specificity=0.9996526452400947\n",
            "train F-measure=0.6578947368421053, IoU=0.49019607843137253\n",
            "test mean loss=0.8554139286279678\n",
            "test accuracy=0.9487954558228417, recall=0.06429856115107914\n",
            "test precision=0.9905987135081643, specificity=0.9999646976633569\n",
            "test F-measure=0.1207588141267304, IoU=0.06425934841919435\n",
            "[epoch 228]\n",
            "train mean loss=0.045401763170957565\n",
            "train accuracy=0.9997291564941406, train recall=0.6862745098039216\n",
            "train precision=0.6422018348623854, specificity=0.9998511688965891\n",
            "train F-measure=0.6635071090047394, IoU=0.49645390070921985\n",
            "test mean loss=0.8746442943811417\n",
            "test accuracy=0.9486004946043165, recall=0.06038026721479959\n",
            "test precision=0.9957627118644068, specificity=0.9999851358582555\n",
            "test F-measure=0.11385658914728683, IoU=0.060364757256614436\n",
            "[epoch 229]\n",
            "train mean loss=0.05860583484172821\n",
            "train accuracy=0.9958152770996094, train recall=0.49593908629441624\n",
            "train precision=0.9037927844588344, specificity=0.9996002675132796\n",
            "train F-measure=0.6404457554900033, IoU=0.4710703953712633\n",
            "test mean loss=0.8529713004827499\n",
            "test accuracy=0.9489306991906474, recall=0.06715698869475847\n",
            "test precision=0.9853911404335532, specificity=0.9999424014507402\n",
            "test F-measure=0.1257441818509832, IoU=0.0670901915487535\n",
            "[epoch 230]\n",
            "train mean loss=0.0560632087290287\n",
            "train accuracy=0.9963874816894531, train recall=0.5872765509989485\n",
            "train precision=0.8733385457388585, specificity=0.9993775024784624\n",
            "train F-measure=0.7022948758252123, IoU=0.5411821705426356\n",
            "test mean loss=0.8269084095954895\n",
            "test accuracy=0.9490483784847122, recall=0.07091469681397738\n",
            "test precision=0.9646133682830931, specificity=0.9998495005648373\n",
            "test F-measure=0.13211667913238595, IoU=0.07073069161034053\n",
            "[epoch 231]\n",
            "train mean loss=0.04565848410129547\n",
            "train accuracy=0.9989204406738281, train recall=0.7142857142857143\n",
            "train precision=0.6673728813559322, specificity=0.9994000833005353\n",
            "train F-measure=0.6900328587075575, IoU=0.5267558528428093\n",
            "test mean loss=0.819691464304924\n",
            "test accuracy=0.948904353080036, recall=0.0691803699897225\n",
            "test precision=0.9518338488731772, specificity=0.9997974760687318\n",
            "test F-measure=0.1289858977813707, IoU=0.06893903024483918\n",
            "[epoch 232]\n",
            "train mean loss=0.045264266431331635\n",
            "train accuracy=0.9996376037597656, train recall=0.4852941176470588\n",
            "train precision=0.3548387096774194, specificity=0.9997710587768434\n",
            "train F-measure=0.4099378881987578, IoU=0.2578125\n",
            "test mean loss=0.8224275559186935\n",
            "test accuracy=0.9486900713803957, recall=0.06465184994861253\n",
            "test precision=0.9572039942938659, specificity=0.9998327784053749\n",
            "test F-measure=0.12112277746021238, IoU=0.06446550951130468\n",
            "[epoch 233]\n",
            "train mean loss=0.04442116618156433\n",
            "train accuracy=0.99969482421875, train recall=0.7364864864864865\n",
            "train precision=0.7266666666666667, specificity=0.999843509061207\n",
            "train F-measure=0.7315436241610739, IoU=0.5767195767195767\n",
            "test mean loss=0.8277799487113953\n",
            "test accuracy=0.9485196998651079, recall=0.06079779033915725\n",
            "test precision=0.9658163265306122, specificity=0.9998755128128902\n",
            "test F-measure=0.11439448875997099, IoU=0.06066724353427555\n",
            "[epoch 234]\n",
            "train mean loss=0.04651610553264618\n",
            "train accuracy=0.9988555908203125, train recall=0.653160453808752\n",
            "train precision=0.8241308793456033, specificity=0.999671162059749\n",
            "train F-measure=0.7287522603978299, IoU=0.5732574679943101\n",
            "test mean loss=0.8420653641223907\n",
            "test accuracy=0.9485003793839928, recall=0.06070143884892086\n",
            "test precision=0.9618320610687023, specificity=0.9998606486711458\n",
            "test F-measure=0.11419594574182047, IoU=0.060555573355547725\n",
            "[epoch 235]\n",
            "train mean loss=0.04355887323617935\n",
            "train accuracy=0.9997520446777344, train recall=0.5\n",
            "train precision=0.9230769230769231, specificity=0.9999809177785242\n",
            "train F-measure=0.6486486486486487, IoU=0.48\n",
            "test mean loss=0.8512854874134064\n",
            "test accuracy=0.9485653664568345, recall=0.06256423432682426\n",
            "test precision=0.9530332681017613, specificity=0.9998216302990666\n",
            "test F-measure=0.11742013261000604, IoU=0.0623719262295082\n",
            "[epoch 236]\n",
            "train mean loss=0.04321562498807907\n",
            "train accuracy=0.9992866516113281, train recall=0.47038327526132406\n",
            "train precision=0.7941176470588235, specificity=0.9998663392615054\n",
            "train F-measure=0.5908096280087528, IoU=0.4192546583850932\n",
            "test mean loss=0.8556452095508575\n",
            "test accuracy=0.9485600972347122, recall=0.06404162384378212\n",
            "test precision=0.9322113136979897, specificity=0.9997305874308817\n",
            "test F-measure=0.11984973703981969, IoU=0.06374476519292861\n",
            "[epoch 237]\n",
            "train mean loss=0.04372125491499901\n",
            "train accuracy=0.9993438720703125, train recall=0.46794871794871795\n",
            "train precision=0.4506172839506173, specificity=0.9996602897842649\n",
            "train F-measure=0.4591194968553459, IoU=0.2979591836734694\n",
            "test mean loss=0.8634077608585358\n",
            "test accuracy=0.9484810589028777, recall=0.06166495375128469\n",
            "test precision=0.9430255402750491, specificity=0.9997844699447054\n",
            "test F-measure=0.11576027975400939, IoU=0.061436068091642135\n",
            "[epoch 238]\n",
            "train mean loss=0.042843833565711975\n",
            "train accuracy=0.9994697570800781, train recall=0.29545454545454547\n",
            "train precision=0.4588235294117647, specificity=0.9998244355220371\n",
            "train F-measure=0.35944700460829493, IoU=0.21910112359550563\n",
            "test mean loss=0.87381212413311\n",
            "test accuracy=0.9483528411645683, recall=0.0579393627954779\n",
            "test precision=0.9611081513052744, specificity=0.9998643647065818\n",
            "test F-measure=0.10929027958682944, IoU=0.057803838636290815\n",
            "[epoch 239]\n",
            "train mean loss=0.047788795083761215\n",
            "train accuracy=0.997467041015625, train recall=0.5046335299073293\n",
            "train precision=0.8874074074074074, specificity=0.9997087642791724\n",
            "train F-measure=0.6433941997851773, IoU=0.4742676167854315\n",
            "test mean loss=0.8399795293807983\n",
            "test accuracy=0.948574148493705, recall=0.06188977389516958\n",
            "test precision=0.9649474211316975, specificity=0.9998699387597361\n",
            "test F-measure=0.11631907765671687, IoU=0.06175094533102608\n",
            "[epoch 240]\n",
            "train mean loss=0.0413338840007782\n",
            "train accuracy=0.9998626708984375, train recall=0.625\n",
            "train precision=0.8928571428571429, specificity=0.9999771048293546\n",
            "train F-measure=0.7352941176470589, IoU=0.5813953488372093\n",
            "test mean loss=0.8108703643083572\n",
            "test accuracy=0.9488323403776978, recall=0.06686793422404934\n",
            "test precision=0.9638888888888889, specificity=0.9998550746179916\n",
            "test F-measure=0.1250600672753484, IoU=0.06670083936695073\n",
            "[epoch 241]\n",
            "train mean loss=0.041359204798936844\n",
            "train accuracy=0.9997940063476562, train recall=0.8870967741935484\n",
            "train precision=0.5392156862745098, specificity=0.9998206668142032\n",
            "train F-measure=0.6707317073170732, IoU=0.5045871559633027\n",
            "test mean loss=0.7955984026193619\n",
            "test accuracy=0.9489588017086331, recall=0.069051901336074\n",
            "test precision=0.966726618705036, specificity=0.9998625066888638\n",
            "test F-measure=0.12889688249400477, IoU=0.06888817686638898\n",
            "[epoch 242]\n",
            "train mean loss=0.04211699217557907\n",
            "train accuracy=0.9990386962890625, train recall=0.5791044776119403\n",
            "train precision=0.6360655737704918, specificity=0.999576026798162\n",
            "train F-measure=0.60625, IoU=0.4349775784753363\n",
            "test mean loss=0.7960157841444016\n",
            "test accuracy=0.9488973274505396, recall=0.06751027749229188\n",
            "test precision=0.9717984281091078, specificity=0.9998866609191985\n",
            "test F-measure=0.12625003753866484, IoU=0.0673782735519441\n",
            "[epoch 243]\n",
            "train mean loss=0.041616037487983704\n",
            "train accuracy=0.9990921020507812, train recall=0.5706371191135734\n",
            "train precision=0.71280276816609, specificity=0.9996829435066448\n",
            "train F-measure=0.6338461538461538, IoU=0.46396396396396394\n",
            "test mean loss=0.8026527613401413\n",
            "test accuracy=0.9488059942670863, recall=0.06571171634121274\n",
            "test precision=0.9728958630527818, specificity=0.9998940929900707\n",
            "test F-measure=0.12310839676283884, IoU=0.06559163914980926\n",
            "[epoch 244]\n",
            "train mean loss=0.04095685854554176\n",
            "train accuracy=0.9992637634277344, train recall=0.7278911564625851\n",
            "train precision=0.8147208121827412, specificity=0.9997210578403763\n",
            "train F-measure=0.7688622754491019, IoU=0.6245136186770428\n",
            "test mean loss=0.8151881843805313\n",
            "test accuracy=0.9486198150854317, recall=0.06192189105858171\n",
            "test precision=0.9771920932589965, specificity=0.9999163892026874\n",
            "test F-measure=0.11646380138330967, IoU=0.0618325262178891\n",
            "[epoch 245]\n",
            "train mean loss=0.04083559662103653\n",
            "train accuracy=0.9990234375, train recall=0.5192743764172335\n",
            "train precision=0.8388278388278388, specificity=0.9998318704791309\n",
            "train F-measure=0.6414565826330532, IoU=0.47216494845360824\n",
            "test mean loss=0.8235955536365509\n",
            "test accuracy=0.9486408919739209, recall=0.06230729701952724\n",
            "test precision=0.9773299748110831, specificity=0.9999163892026874\n",
            "test F-measure=0.11714622143051236, IoU=0.06221737596613322\n",
            "[epoch 246]\n",
            "train mean loss=0.03971605747938156\n",
            "train accuracy=0.9997940063476562, train recall=0.23529411764705882\n",
            "train precision=0.8888888888888888, specificity=0.9999923686258948\n",
            "train F-measure=0.3720930232558139, IoU=0.22857142857142856\n",
            "test mean loss=0.8253570944070816\n",
            "test accuracy=0.9487497892311151, recall=0.06478031860226105\n",
            "test precision=0.9711121810303323, specificity=0.9998885189369165\n",
            "test F-measure=0.1214584650588625, IoU=0.06465572509296064\n",
            "[epoch 247]\n",
            "train mean loss=0.03937960043549538\n",
            "train accuracy=0.9992828369140625, train recall=0.4564459930313589\n",
            "train precision=0.803680981595092, specificity=0.9998777958962335\n",
            "train F-measure=0.5822222222222222, IoU=0.4106583072100313\n",
            "test mean loss=0.8256886750459671\n",
            "test accuracy=0.9488516608588129, recall=0.06783144912641316\n",
            "test precision=0.9560887279311906, specificity=0.9998197722813484\n",
            "test F-measure=0.12667566351776877, IoU=0.06762078570742484\n",
            "[epoch 248]\n",
            "train mean loss=0.04119833558797836\n",
            "train accuracy=0.99859619140625, train recall=0.6537931034482759\n",
            "train precision=0.8020304568527918, specificity=0.9995524426304132\n",
            "train F-measure=0.7203647416413373, IoU=0.5629453681710214\n",
            "test mean loss=0.8311665058135986\n",
            "test accuracy=0.9488147763039568, recall=0.06744604316546762\n",
            "test precision=0.9519492293744334, specificity=0.999803050121886\n",
            "test F-measure=0.125967248515386, IoU=0.06721720760514692\n",
            "[epoch 249]\n",
            "train mean loss=0.03866273909807205\n",
            "train accuracy=0.9995498657226562, train recall=0.95\n",
            "train precision=0.4, specificity=0.9995649917577386\n",
            "train F-measure=0.5629629629629629, IoU=0.3917525773195876\n",
            "test mean loss=0.855491891503334\n",
            "test accuracy=0.94855482801259, recall=0.0612153134635149\n",
            "test precision=0.9694811800610377, specificity=0.9998885189369165\n",
            "test F-measure=0.11515920488188025, IoU=0.061097576612386205\n",
            "[epoch 250]\n",
            "train mean loss=0.03859959542751312\n",
            "train accuracy=0.9995918273925781, train recall=0.5304878048780488\n",
            "train precision=0.7435897435897436, specificity=0.9998854874417894\n",
            "train F-measure=0.6192170818505338, IoU=0.4484536082474227\n",
            "test mean loss=0.874500647187233\n",
            "test accuracy=0.9483528411645683, recall=0.05700796505652621\n",
            "test precision=0.9758108851017042, specificity=0.9999182472204055\n",
            "test F-measure=0.10772265210135032, IoU=0.05692751763951251\n",
            "[epoch 251]\n",
            "train mean loss=0.039230749011039734\n",
            "train accuracy=0.9989700317382812, train recall=0.4682203389830508\n",
            "train precision=0.9208333333333333, specificity=0.9999273900149805\n",
            "train F-measure=0.6207865168539326, IoU=0.45010183299389\n",
            "test mean loss=0.8734569102525711\n",
            "test accuracy=0.9483827000899281, recall=0.05797147995889003\n",
            "test precision=0.9693877551020408, specificity=0.9998940929900707\n",
            "test F-measure=0.10940056973149886, IoU=0.05786554675728529\n",
            "[epoch 252]\n",
            "train mean loss=0.04041394591331482\n",
            "train accuracy=0.9985389709472656, train recall=0.468975468975469\n",
            "train precision=0.9558823529411765, specificity=0.9999426278729092\n",
            "train F-measure=0.6292352371732818, IoU=0.4590395480225989\n",
            "test mean loss=0.8322763592004776\n",
            "test accuracy=0.9486022510116906, recall=0.06567959917780061\n",
            "test precision=0.9224176815516464, specificity=0.9996804209524942\n",
            "test F-measure=0.12262764968668484, IoU=0.06531876836591287\n",
            "[epoch 253]\n",
            "train mean loss=0.04094669967889786\n",
            "train accuracy=0.9982261657714844, train recall=0.6720841300191205\n",
            "train precision=0.8521212121212122, specificity=0.9995327424951551\n",
            "train F-measure=0.7514698022447889, IoU=0.6018835616438356\n",
            "test mean loss=0.8043488264083862\n",
            "test accuracy=0.9489447504496403, recall=0.07290596094552929\n",
            "test precision=0.9182847896440129, specificity=0.9996246804209525\n",
            "test F-measure=0.13508688407522018, IoU=0.07243602016720914\n",
            "[epoch 254]\n",
            "train mean loss=0.04081875458359718\n",
            "train accuracy=0.9980545043945312, train recall=0.7151454363089268\n",
            "train precision=0.759318423855165, specificity=0.9991345870333567\n",
            "train F-measure=0.7365702479338844, IoU=0.5829926410466068\n",
            "test mean loss=0.7973524183034897\n",
            "test accuracy=0.9491151219649281, recall=0.07235996916752313\n",
            "test precision=0.9624092268261427, specificity=0.999836494440811\n",
            "test F-measure=0.1345998745407295, IoU=0.07215603382013835\n",
            "[epoch 255]\n",
            "train mean loss=0.03700646013021469\n",
            "train accuracy=0.9998130798339844, train recall=0.9215686274509803\n",
            "train precision=0.6962962962962963, specificity=0.9998435365323116\n",
            "train F-measure=0.7932489451476793, IoU=0.6573426573426573\n",
            "test mean loss=0.7998858392238617\n",
            "test accuracy=0.9489535324865108, recall=0.06802415210688592\n",
            "test precision=0.9791955617198336, specificity=0.9999163892026874\n",
            "test F-measure=0.12721102735817896, IoU=0.06792598056508771\n",
            "[epoch 256]\n",
            "train mean loss=0.039288975298404694\n",
            "train accuracy=0.9983482360839844, train recall=0.6512118018967334\n",
            "train precision=0.8583333333333333, specificity=0.999609487164762\n",
            "train F-measure=0.7405632115038946, IoU=0.588011417697431\n",
            "test mean loss=0.8020471334457397\n",
            "test accuracy=0.9487831609712231, recall=0.06429856115107914\n",
            "test precision=0.9871794871794872, specificity=0.9999516915393305\n",
            "test F-measure=0.12073332529248584, IoU=0.06424491367691419\n",
            "[epoch 257]\n",
            "train mean loss=0.03645815700292587\n",
            "train accuracy=0.99981689453125, train recall=0.6632653061224489\n",
            "train precision=0.8125, specificity=0.9999427581417003\n",
            "train F-measure=0.7303370786516853, IoU=0.5752212389380531\n",
            "test mean loss=0.7967098653316498\n",
            "test accuracy=0.9488077506744604, recall=0.06461973278520042\n",
            "test precision=0.9891838741396264, specificity=0.9999591236102028\n",
            "test F-measure=0.12131444075972264, IoU=0.06457410616856024\n",
            "[epoch 258]\n",
            "train mean loss=0.047048792243003845\n",
            "train accuracy=0.9961814880371094, train recall=0.5017006802721088\n",
            "train precision=0.8788480635551142, specificity=0.9995314540287272\n",
            "train F-measure=0.6387585709130278, IoU=0.4692470837751856\n",
            "test mean loss=0.7539434134960175\n",
            "test accuracy=0.9493153524055755, recall=0.07563591983556012\n",
            "test precision=0.9687371452077335, specificity=0.9998587906534276\n",
            "test F-measure=0.14031638216105102, IoU=0.07545174932718185\n",
            "[epoch 259]\n",
            "train mean loss=0.03725925832986832\n",
            "train accuracy=0.9991722106933594, train recall=0.8397212543554007\n",
            "train precision=0.5849514563106796, specificity=0.9993469718204975\n",
            "train F-measure=0.6895565092989985, IoU=0.5262008733624454\n",
            "test mean loss=0.7562941312789917\n",
            "test accuracy=0.9491976731115108, recall=0.0736125385405961\n",
            "test precision=0.9662731871838112, specificity=0.9998513585825555\n",
            "test F-measure=0.13680315148621228, IoU=0.07342388518708355\n",
            "[epoch 260]\n",
            "train mean loss=0.0362086147069931\n",
            "train accuracy=0.9994697570800781, train recall=0.8842443729903537\n",
            "train precision=0.7275132275132276, specificity=0.9996066194864666\n",
            "train F-measure=0.7982583454281568, IoU=0.6642512077294686\n",
            "test mean loss=0.7750379592180252\n",
            "test accuracy=0.9489148915242805, recall=0.06763874614594038\n",
            "test precision=0.974548819990745, specificity=0.9998978090255068\n",
            "test F-measure=0.1264978826921344, IoU=0.06751947677214581\n",
            "[epoch 261]\n",
            "train mean loss=0.03584424406290054\n",
            "train accuracy=0.9996757507324219, train recall=0.09574468085106383\n",
            "train precision=1.0, specificity=1.0\n",
            "train F-measure=0.17475728155339806, IoU=0.09574468085106383\n",
            "test mean loss=0.7782910168170929\n",
            "test accuracy=0.9489342120053957, recall=0.06802415210688592\n",
            "test precision=0.9742410303587856, specificity=0.9998959510077888\n",
            "test F-measure=0.127169018312819, IoU=0.06790202616055399\n",
            "[epoch 262]\n",
            "train mean loss=0.04522358626127243\n",
            "train accuracy=0.9967193603515625, train recall=0.5969543147208122\n",
            "train precision=0.9468599033816425, specificity=0.9997463236141967\n",
            "train F-measure=0.7322540473225405, IoU=0.5776031434184676\n",
            "test mean loss=0.769465446472168\n",
            "test accuracy=0.9490027118929856, recall=0.07001541623843782\n",
            "test precision=0.965028773793714, specificity=0.9998532166002735\n",
            "test F-measure=0.13055846683635272, IoU=0.06983821880506166\n",
            "[epoch 263]\n",
            "train mean loss=0.035105250775814056\n",
            "train accuracy=0.9995002746582031, train recall=0.7218045112781954\n",
            "train precision=0.7710843373493976, specificity=0.9997823413956117\n",
            "train F-measure=0.7456310679611651, IoU=0.5944272445820433\n",
            "test mean loss=0.7665492743253708\n",
            "test accuracy=0.949009737522482, recall=0.0704650565262076\n",
            "test precision=0.9610162067455102, specificity=0.9998346364230929\n",
            "test F-measure=0.13130255244022862, IoU=0.07026421136909528\n",
            "[epoch 264]\n",
            "train mean loss=0.03434177488088608\n",
            "train accuracy=0.9998283386230469, train recall=0.68\n",
            "train precision=0.5396825396825397, specificity=0.9998893526749945\n",
            "train F-measure=0.6017699115044247, IoU=0.43037974683544306\n",
            "test mean loss=0.7671831995248795\n",
            "test accuracy=0.9489781221897482, recall=0.07027235354573484\n",
            "test precision=0.9558759283529926, specificity=0.9998123402104763\n",
            "test F-measure=0.13091997008227374, IoU=0.07004513877773154\n",
            "[epoch 265]\n",
            "train mean loss=0.03736394643783569\n",
            "train accuracy=0.9983978271484375, train recall=0.6759581881533101\n",
            "train precision=0.8049792531120332, specificity=0.9994603552469927\n",
            "train F-measure=0.7348484848484849, IoU=0.5808383233532934\n",
            "test mean loss=0.7702878266572952\n",
            "test accuracy=0.9489377248201439, recall=0.06908401849948613\n",
            "test precision=0.9611260053619303, specificity=0.999838352458529\n",
            "test F-measure=0.12890273865883622, IoU=0.06889152227524581\n",
            "[epoch 266]\n",
            "train mean loss=0.04277318716049194\n",
            "train accuracy=0.99676513671875, train recall=0.5995635570103656\n",
            "train precision=0.9060181368507831, specificity=0.9995620623023999\n",
            "train F-measure=0.7216021011162179, IoU=0.5644581407293272\n",
            "test mean loss=0.7678065747022629\n",
            "test accuracy=0.9489289427832733, recall=0.06892343268242548\n",
            "test precision=0.961038961038961, specificity=0.999838352458529\n",
            "test F-measure=0.12862237405975607, IoU=0.06873138391570317\n",
            "[epoch 267]\n",
            "train mean loss=0.03370538726449013\n",
            "train accuracy=0.9998855590820312, train recall=0.8382352941176471\n",
            "train precision=0.75, specificity=0.9999275019460004\n",
            "train F-measure=0.7916666666666666, IoU=0.6551724137931034\n",
            "test mean loss=0.7698510810732841\n",
            "test accuracy=0.9488709813399281, recall=0.06763874614594038\n",
            "test precision=0.9634034766697164, specificity=0.9998513585825555\n",
            "test F-measure=0.12640297701218411, IoU=0.06746540235776524\n",
            "[epoch 268]\n",
            "train mean loss=0.034183330833911896\n",
            "train accuracy=0.99957275390625, train recall=0.6746031746031746\n",
            "train precision=0.5448717948717948, specificity=0.999729026250105\n",
            "train F-measure=0.6028368794326241, IoU=0.43147208121827413\n",
            "test mean loss=0.7769680023193359\n",
            "test accuracy=0.9487673533048561, recall=0.06526207605344296\n",
            "test precision=0.9690033381020505, specificity=0.9998792288483263\n",
            "test F-measure=0.12228808714229833, IoU=0.06512611775263613\n",
            "[epoch 269]\n",
            "train mean loss=0.033903103321790695\n",
            "train accuracy=0.9993743896484375, train recall=0.29411764705882354\n",
            "train precision=0.75, specificity=0.9999236466366344\n",
            "train F-measure=0.42253521126760557, IoU=0.26785714285714285\n",
            "test mean loss=0.7738818004727364\n",
            "test accuracy=0.9487954558228417, recall=0.06619347379239465\n",
            "test precision=0.9635343618513323, specificity=0.9998550746179916\n",
            "test F-measure=0.12387678437265215, IoU=0.06602806433010829\n",
            "[epoch 270]\n",
            "train mean loss=0.03851020336151123\n",
            "train accuracy=0.9977836608886719, train recall=0.6272058823529412\n",
            "train precision=0.9201725997842503, specificity=0.9997162402601386\n",
            "train F-measure=0.7459554000874509, IoU=0.594839609483961\n",
            "test mean loss=0.7522875815629959\n",
            "test accuracy=0.9489956862634892, recall=0.07062564234326824\n",
            "test precision=0.9556714471968709, specificity=0.9998104821927581\n",
            "test F-measure=0.131530938780393, IoU=0.07039503169216979\n",
            "[epoch 271]\n",
            "train mean loss=0.03292739391326904\n",
            "train accuracy=0.9997901916503906, train recall=0.9680851063829787\n",
            "train precision=0.6363636363636364, specificity=0.9998015645869109\n",
            "train F-measure=0.7679324894514769, IoU=0.6232876712328768\n",
            "test mean loss=0.7461428195238113\n",
            "test accuracy=0.9490554041142086, recall=0.07101104830421377\n",
            "test precision=0.9650807507638586, specificity=0.9998513585825555\n",
            "test F-measure=0.13228826996140844, IoU=0.0708290620194772\n",
            "[epoch 272]\n",
            "train mean loss=0.03311675786972046\n",
            "train accuracy=0.999542236328125, train recall=0.7228260869565217\n",
            "train precision=0.6584158415841584, specificity=0.9997366010077875\n",
            "train F-measure=0.689119170984456, IoU=0.525691699604743\n",
            "test mean loss=0.7522031962871552\n",
            "test accuracy=0.9489166479316546, recall=0.06754239465570401\n",
            "test precision=0.9763231197771588, specificity=0.9999052410963791\n",
            "test F-measure=0.1263442475217783, IoU=0.06743194279667811\n",
            "[epoch 273]\n",
            "train mean loss=0.03299182653427124\n",
            "train accuracy=0.9993629455566406, train recall=0.6237942122186495\n",
            "train precision=0.7950819672131147, specificity=0.9998090385856634\n",
            "train F-measure=0.6990990990990992, IoU=0.5373961218836565\n",
            "test mean loss=0.7589664608240128\n",
            "test accuracy=0.9487884301933454, recall=0.06455549845837616\n",
            "test precision=0.984811366976972, specificity=0.9999424014507402\n",
            "test F-measure=0.12116827922958676, IoU=0.06449128886322071\n",
            "[epoch 274]\n",
            "train mean loss=0.0326535701751709\n",
            "train accuracy=0.9996719360351562, train recall=0.20588235294117646\n",
            "train precision=0.8076923076923077, specificity=0.9999809190893063\n",
            "train F-measure=0.328125, IoU=0.19626168224299065\n",
            "test mean loss=0.7480411902070045\n",
            "test accuracy=0.9490518912994604, recall=0.06991906474820143\n",
            "test precision=0.9784269662921349, specificity=0.9999108151495333\n",
            "test F-measure=0.13051167530949312, IoU=0.0698114417650077\n",
            "[epoch 275]\n",
            "train mean loss=0.03765666484832764\n",
            "train accuracy=0.9978561401367188, train recall=0.6606851549755302\n",
            "train precision=0.8472803347280334, specificity=0.9994404372254885\n",
            "train F-measure=0.7424381301558204, IoU=0.5903790087463557\n",
            "test mean loss=0.7513170540332794\n",
            "test accuracy=0.949183621852518, recall=0.07393371017471737\n",
            "test precision=0.9591666666666666, specificity=0.9998179142636304\n",
            "test F-measure=0.13728530534351144, IoU=0.0737017352884677\n",
            "[epoch 276]\n",
            "train mean loss=0.03292475640773773\n",
            "train accuracy=0.9992713928222656, train recall=0.8337950138504155\n",
            "train precision=0.6967592592592593, specificity=0.999499585534584\n",
            "train F-measure=0.759142496847415, IoU=0.6117886178861789\n",
            "test mean loss=0.7833992838859558\n",
            "test accuracy=0.9488218019334532, recall=0.06661099691675232\n",
            "test precision=0.9646511627906976, specificity=0.9998587906534276\n",
            "test F-measure=0.12461695607763026, IoU=0.0664488017429194\n",
            "[epoch 277]\n",
            "train mean loss=0.03225488215684891\n",
            "train accuracy=0.9993896484375, train recall=0.5939849624060151\n",
            "train precision=0.7523809523809524, specificity=0.9998014342556457\n",
            "train F-measure=0.6638655462184874, IoU=0.4968553459119497\n",
            "test mean loss=0.8177411556243896\n",
            "test accuracy=0.9484476871627698, recall=0.05906346351490236\n",
            "test precision=0.971473851030111, specificity=0.999899667043225\n",
            "test F-measure=0.11135668654818491, IoU=0.05896120551458801\n",
            "[epoch 278]\n",
            "train mean loss=0.044804517179727554\n",
            "train accuracy=0.995758056640625, train recall=0.4747634069400631\n",
            "train precision=0.8887795275590551, specificity=0.999565788765841\n",
            "train F-measure=0.6189170664838931, IoU=0.4481389578163772\n",
            "test mean loss=0.7777810022234917\n",
            "test accuracy=0.9488024814523381, recall=0.06600077081192189\n",
            "test precision=0.9679698539802166, specificity=0.9998736547951721\n",
            "test F-measure=0.12357557352896961, IoU=0.06585694141776695\n",
            "[epoch 279]\n",
            "train mean loss=0.03181719407439232\n",
            "train accuracy=0.9998245239257812, train recall=0.8947368421052632\n",
            "train precision=0.4473684210526316, specificity=0.99983975948662\n",
            "train F-measure=0.5964912280701755, IoU=0.425\n",
            "test mean loss=0.7592834681272507\n",
            "test accuracy=0.948974609375, recall=0.06885919835560124\n",
            "test precision=0.9732183386291421, specificity=0.9998903769546347\n",
            "test F-measure=0.12861813491706411, IoU=0.06872896297483572\n",
            "[epoch 280]\n",
            "train mean loss=0.03499497100710869\n",
            "train accuracy=0.998138427734375, train recall=0.6270810210876804\n",
            "train precision=0.7880055788005579, specificity=0.9994181662283774\n",
            "train F-measure=0.6983930778739185, IoU=0.5365622032288699\n",
            "test mean loss=0.7525201663374901\n",
            "test accuracy=0.9490764810026978, recall=0.0706898766700925\n",
            "test precision=0.9743249225320938, specificity=0.9998922349723527\n",
            "test F-measure=0.13181614014073964, IoU=0.0705584407257806\n",
            "[epoch 281]\n",
            "train mean loss=0.03078833967447281\n",
            "train accuracy=0.9998283386230469, train recall=0.8518518518518519\n",
            "train precision=0.7603305785123967, specificity=0.9998893281839136\n",
            "train F-measure=0.8034934497816595, IoU=0.6715328467153284\n",
            "test mean loss=0.7531604766845703\n",
            "test accuracy=0.9490536477068345, recall=0.07007965056526208\n",
            "test precision=0.9767233661593554, specificity=0.999903383078661\n",
            "test F-measure=0.13077614623913694, IoU=0.06996280620751572\n",
            "[epoch 282]\n",
            "train mean loss=0.03078976273536682\n",
            "train accuracy=0.9998321533203125, train recall=0.4117647058823529\n",
            "train precision=0.875, specificity=0.9999847372517896\n",
            "train F-measure=0.56, IoU=0.3888888888888889\n",
            "test mean loss=0.7469059601426125\n",
            "test accuracy=0.9490940450764388, recall=0.07094681397738951\n",
            "test precision=0.9752759381898455, specificity=0.9998959510077888\n",
            "test F-measure=0.1322714888775785, IoU=0.07081944088227751\n",
            "[epoch 283]\n",
            "train mean loss=0.030411837622523308\n",
            "train accuracy=0.9997978210449219, train recall=0.7352941176470589\n",
            "train precision=0.5882352941176471, specificity=0.9998664509531586\n",
            "train F-measure=0.65359477124183, IoU=0.4854368932038835\n",
            "test mean loss=0.744405671954155\n",
            "test accuracy=0.9490852630395683, recall=0.07101104830421377\n",
            "test precision=0.9722955145118733, specificity=0.9998829448837624\n",
            "test F-measure=0.13235558216102963, IoU=0.07086765601461585\n",
            "[epoch 284]\n",
            "train mean loss=0.03126012161374092\n",
            "train accuracy=0.9990463256835938, train recall=0.711645101663586\n",
            "train precision=0.8037578288100209, specificity=0.9996406769035523\n",
            "train F-measure=0.7549019607843136, IoU=0.6062992125984252\n",
            "test mean loss=0.7504163086414337\n",
            "test accuracy=0.9489974426708633, recall=0.06921248715313463\n",
            "test precision=0.9742314647377939, specificity=0.9998940929900707\n",
            "test F-measure=0.12924313302147059, IoU=0.0690860128875068\n",
            "[epoch 285]\n",
            "train mean loss=0.036298029124736786\n",
            "train accuracy=0.99755859375, train recall=0.636716224951519\n",
            "train precision=0.9266227657572906, specificity=0.9997006872680806\n",
            "train F-measure=0.7547892720306513, IoU=0.6061538461538462\n",
            "test mean loss=0.7388380393385887\n",
            "test accuracy=0.9490501348920863, recall=0.06953365878725591\n",
            "test precision=0.983197093551317, specificity=0.9999312533444319\n",
            "test F-measure=0.1298818165456836, IoU=0.06945112757835306\n",
            "[epoch 286]\n",
            "train mean loss=0.02992543950676918\n",
            "train accuracy=0.9996376037597656, train recall=0.5543478260869565\n",
            "train precision=0.8869565217391304, specificity=0.9999503741029164\n",
            "train F-measure=0.6822742474916388, IoU=0.5177664974619289\n",
            "test mean loss=0.7243098393082619\n",
            "test accuracy=0.9492486089253597, recall=0.07319501541623843\n",
            "test precision=0.983599482088908, specificity=0.9999293953267139\n",
            "test F-measure=0.13625085941470122, IoU=0.07310579328927953\n",
            "[epoch 287]\n",
            "train mean loss=0.029496388509869576\n",
            "train accuracy=0.9998435974121094, train recall=0.5967741935483871\n",
            "train precision=0.6981132075471698, specificity=0.9999389504048352\n",
            "train F-measure=0.6434782608695653, IoU=0.47435897435897434\n",
            "test mean loss=0.7146753221750259\n",
            "test accuracy=0.949427762477518, recall=0.07669578622816033\n",
            "test precision=0.9815043156596794, specificity=0.9999163892026874\n",
            "test F-measure=0.14227412195775863, IoU=0.07658509990058048\n",
            "[epoch 288]\n",
            "train mean loss=0.03263577073812485\n",
            "train accuracy=0.998504638671875, train recall=0.7186261558784677\n",
            "train precision=0.7524204702627939, specificity=0.9993151916506942\n",
            "train F-measure=0.7351351351351351, IoU=0.5811965811965812\n",
            "test mean loss=0.725365400314331\n",
            "test accuracy=0.9492521217401079, recall=0.07313078108941418\n",
            "test precision=0.9852877542189529, specificity=0.999936827397586\n",
            "test F-measure=0.13615570903220017, IoU=0.07305101058710298\n",
            "[epoch 289]\n",
            "train mean loss=0.03503408655524254\n",
            "train accuracy=0.9976615905761719, train recall=0.6063926940639269\n",
            "train precision=0.7848699763593381, specificity=0.9993028128818727\n",
            "train F-measure=0.684183410613086, IoU=0.5199686765857479\n",
            "test mean loss=0.7234512120485306\n",
            "test accuracy=0.949167814186151, recall=0.07216726618705036\n",
            "test precision=0.9773814702044367, specificity=0.999903383078661\n",
            "test F-measure=0.13441004934948406, IoU=0.07204694113120431\n",
            "[epoch 290]\n",
            "train mean loss=0.0318121612071991\n",
            "train accuracy=0.9984245300292969, train recall=0.7064555420219245\n",
            "train precision=0.7712765957446809, specificity=0.999341810709352\n",
            "train F-measure=0.7374443738080103, IoU=0.5840886203423967\n",
            "test mean loss=0.7400558665394783\n",
            "test accuracy=0.9487884301933454, recall=0.06522995889003083\n",
            "test precision=0.9750360057609218, specificity=0.999903383078661\n",
            "test F-measure=0.12227941840512957, IoU=0.06512120046171604\n",
            "[epoch 291]\n",
            "train mean loss=0.03225093334913254\n",
            "train accuracy=0.9984893798828125, train recall=0.6981132075471698\n",
            "train precision=0.8353253652058433, specificity=0.9995253461336763\n",
            "train F-measure=0.7605804111245466, IoU=0.6136585365853658\n",
            "test mean loss=0.7556026950478554\n",
            "test accuracy=0.9486531868255396, recall=0.06246788283658787\n",
            "test precision=0.9783702213279678, specificity=0.9999201052381236\n",
            "test F-measure=0.11743750754739767, IoU=0.06238173129349883\n",
            "[epoch 292]\n",
            "train mean loss=0.0360138863325119\n",
            "train accuracy=0.9972915649414062, train recall=0.5654205607476636\n",
            "train precision=0.934878587196468, specificity=0.9997736393422496\n",
            "train F-measure=0.7046589018302828, IoU=0.5439948619139371\n",
            "test mean loss=0.7305942177772522\n",
            "test accuracy=0.9489517760791367, recall=0.06882708119218911\n",
            "test precision=0.9679313459801264, specificity=0.9998680807420179\n",
            "test F-measure=0.12851574212893555, IoU=0.0686704905950588\n",
            "[epoch 293]\n",
            "train mean loss=0.03025026060640812\n",
            "train accuracy=0.9989700317382812, train recall=0.7017892644135189\n",
            "train precision=0.7463002114164905, specificity=0.9995413562859032\n",
            "train F-measure=0.7233606557377049, IoU=0.5666131621187801\n",
            "test mean loss=0.7216313406825066\n",
            "test accuracy=0.9490290580035972, recall=0.07043293936279547\n",
            "test precision=0.966079295154185, specificity=0.9998569326357096\n",
            "test F-measure=0.131293779560558, IoU=0.07025918687726268\n",
            "[epoch 294]\n",
            "train mean loss=0.030062634497880936\n",
            "train accuracy=0.9991722106933594, train recall=0.8022181146025879\n",
            "train precision=0.7977941176470589, specificity=0.9995795155254336\n",
            "train F-measure=0.8000000000000002, IoU=0.6666666666666666\n",
            "test mean loss=0.7282611951231956\n",
            "test accuracy=0.9490114939298561, recall=0.06914825282631037\n",
            "test precision=0.9786363636363636, specificity=0.9999126731672513\n",
            "test F-measure=0.12916966642668584, IoU=0.06904403040118014\n",
            "[epoch 295]\n",
            "train mean loss=0.03290043771266937\n",
            "train accuracy=0.9978103637695312, train recall=0.6267902274641954\n",
            "train precision=0.8502857142857143, specificity=0.9994980015864683\n",
            "train F-measure=0.7216294859359844, IoU=0.5644916540212443\n",
            "test mean loss=0.7245439291000366\n",
            "test accuracy=0.9491045835206835, recall=0.07040082219938334\n",
            "test precision=0.9851685393258427, specificity=0.9999386854153042\n",
            "test F-measure=0.13141092892898895, IoU=0.07032628573261895\n",
            "[epoch 296]\n",
            "train mean loss=0.028848979622125626\n",
            "train accuracy=0.9996261596679688, train recall=0.9901960784313726\n",
            "train precision=0.51010101010101, specificity=0.9996298303325422\n",
            "train F-measure=0.6733333333333333, IoU=0.507537688442211\n",
            "test mean loss=0.7388281598687172\n",
            "test accuracy=0.9487726225269785, recall=0.06384892086330936\n",
            "test precision=0.9910269192422732, specificity=0.999966555681075\n",
            "test F-measure=0.11996861987810033, IoU=0.0638120305578738\n",
            "[epoch 297]\n",
            "train mean loss=0.0364968478679657\n",
            "train accuracy=0.9968223571777344, train recall=0.5447845804988662\n",
            "train precision=0.9697275479313825, specificity=0.9998847837775559\n",
            "train F-measure=0.6976406533575318, IoU=0.5356744704570792\n",
            "test mean loss=0.6927856579422951\n",
            "test accuracy=0.9493539933678058, recall=0.0750256937307297\n",
            "test precision=0.9852382960776044, specificity=0.999934969379868\n",
            "test F-measure=0.13943355119825707, IoU=0.07494145199063232\n",
            "[epoch 298]\n",
            "train mean loss=0.03083927556872368\n",
            "train accuracy=0.9986534118652344, train recall=0.7466539196940727\n",
            "train precision=0.8987341772151899, specificity=0.999662961799784\n",
            "train F-measure=0.8156657963446476, IoU=0.6887125220458554\n",
            "test mean loss=0.6642463505268097\n",
            "test accuracy=0.9497562106564749, recall=0.083922147995889\n",
            "test precision=0.969213649851632, specificity=0.9998457845294013\n",
            "test F-measure=0.15446914164104988, IoU=0.0836990294372017\n",
            "[epoch 299]\n",
            "train mean loss=0.02888021618127823\n",
            "train accuracy=0.9992561340332031, train recall=0.6285714285714286\n",
            "train precision=0.5968992248062015, specificity=0.9996029003547169\n",
            "train F-measure=0.6123260437375745, IoU=0.44126074498567336\n",
            "test mean loss=0.6597832217812538\n",
            "test accuracy=0.9499248257643885, recall=0.08745503597122302\n",
            "test precision=0.9656028368794326, specificity=0.9998197722813484\n",
            "test F-measure=0.16038402638708915, IoU=0.08718342778471488\n",
            "[epoch 300]\n",
            "train mean loss=0.028135180473327637\n",
            "train accuracy=0.9995651245117188, train recall=0.7045454545454546\n",
            "train precision=0.5535714285714286, specificity=0.9997137535685389\n",
            "train F-measure=0.62, IoU=0.4492753623188406\n",
            "test mean loss=0.6658985689282417\n",
            "test accuracy=0.9498932104316546, recall=0.0868448098663926\n",
            "test precision=0.9657142857142857, specificity=0.9998216302990666\n",
            "test F-measure=0.15935879302215938, IoU=0.08657786885245902\n",
            "[epoch 301]\n",
            "train mean loss=0.028640449047088623\n",
            "train accuracy=0.9990119934082031, train recall=0.7256461232604374\n",
            "train precision=0.7510288065843621, specificity=0.9995375342549524\n",
            "train F-measure=0.738119312436805, IoU=0.5849358974358975\n",
            "test mean loss=0.6844240427017212\n",
            "test accuracy=0.9496297493255396, recall=0.08160971223021583\n",
            "test precision=0.9683689024390244, specificity=0.9998457845294013\n",
            "test F-measure=0.15053317535545024, IoU=0.08139274159966688\n",
            "[epoch 302]\n",
            "train mean loss=0.028612032532691956\n",
            "train accuracy=0.9996566772460938, train recall=0.0\n",
            "train precision=0.0, specificity=1.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=0.6845018044114113\n",
            "test accuracy=0.9497702619154677, recall=0.08501413155190134\n",
            "test precision=0.9604499274310595, specificity=0.9997974760687318\n",
            "test F-measure=0.15620205358196623, IoU=0.0847175548087694\n",
            "[epoch 303]\n",
            "train mean loss=0.03204537183046341\n",
            "train accuracy=0.9980087280273438, train recall=0.7303070761014686\n",
            "train precision=0.9026402640264026, specificity=0.9995472786844993\n",
            "train F-measure=0.807380073800738, IoU=0.676980198019802\n",
            "test mean loss=0.6817668899893761\n",
            "test accuracy=0.9497386465827338, recall=0.08453237410071943\n",
            "test precision=0.9591836734693877, specificity=0.9997919020155777\n",
            "test F-measure=0.1553719008264463, IoU=0.08422939068100359\n",
            "[epoch 304]\n",
            "train mean loss=0.030450783669948578\n",
            "train accuracy=0.9984245300292969, train recall=0.7657232704402516\n",
            "train precision=0.8943985307621671, specificity=0.9995591707810727\n",
            "train F-measure=0.8250741211351122, IoU=0.7022350396539293\n",
            "test mean loss=0.6810503453016281\n",
            "test accuracy=0.9496490698066546, recall=0.08225205549845838\n",
            "test precision=0.9653222766679231, specificity=0.9998290623699387\n",
            "test F-measure=0.15158779484447601, IoU=0.08200973485333675\n",
            "[epoch 305]\n",
            "train mean loss=0.02766267955303192\n",
            "train accuracy=0.99932861328125, train recall=0.8525121555915721\n",
            "train precision=0.8608837970540099, specificity=0.9996749857567288\n",
            "train F-measure=0.8566775244299674, IoU=0.7492877492877493\n",
            "test mean loss=0.6932180747389793\n",
            "test accuracy=0.9494207368480215, recall=0.07730601233299075\n",
            "test precision=0.9725252525252526, specificity=0.9998736547951721\n",
            "test F-measure=0.14322691975841242, IoU=0.07713754646840149\n",
            "[epoch 306]\n",
            "train mean loss=0.02955196425318718\n",
            "train accuracy=0.9984283447265625, train recall=0.6983993260320135\n",
            "train precision=0.9388448471121178, specificity=0.9997930693562541\n",
            "train F-measure=0.8009661835748793, IoU=0.6680096696212732\n",
            "test mean loss=0.688424602150917\n",
            "test accuracy=0.9494962623651079, recall=0.07859069886947585\n",
            "test precision=0.9741242038216561, specificity=0.9998792288483263\n",
            "test F-measure=0.14544698050404184, IoU=0.07842697349443928\n",
            "[epoch 307]\n",
            "train mean loss=0.028893254697322845\n",
            "train accuracy=0.9986724853515625, train recall=0.7034220532319392\n",
            "train precision=0.8295964125560538, specificity=0.9995638116737771\n",
            "train F-measure=0.7613168724279835, IoU=0.6146179401993356\n",
            "test mean loss=0.6781572476029396\n",
            "test accuracy=0.9495893519559353, recall=0.08019655704008222\n",
            "test precision=0.9757717858538492, specificity=0.9998848029014805\n",
            "test F-measure=0.1482119008755008, IoU=0.08003718187063273\n",
            "[epoch 308]\n",
            "train mean loss=0.034314073622226715\n",
            "train accuracy=0.9970245361328125, train recall=0.6034712950600801\n",
            "train precision=0.8293577981651377, specificity=0.9992863884348887\n",
            "train F-measure=0.6986089644513138, IoU=0.5368171021377672\n",
            "test mean loss=0.6670787259936333\n",
            "test accuracy=0.94939439073741, recall=0.07749871531346351\n",
            "test precision=0.9644284572342127, specificity=0.9998346364230929\n",
            "test F-measure=0.14346869611748617, IoU=0.07727782225780624\n",
            "[epoch 309]\n",
            "train mean loss=0.02673763409256935\n",
            "train accuracy=0.9996910095214844, train recall=0.3431372549019608\n",
            "train precision=0.7142857142857143, specificity=0.9999465734500577\n",
            "train F-measure=0.4635761589403974, IoU=0.3017241379310345\n",
            "test mean loss=0.6662941873073578\n",
            "test accuracy=0.9492152371852518, recall=0.0764388489208633\n",
            "test precision=0.9377462568951931, specificity=0.999706433200547\n",
            "test F-measure=0.14135534833996552, IoU=0.07605291749217102\n",
            "[epoch 310]\n",
            "train mean loss=0.03345967084169388\n",
            "train accuracy=0.9973182678222656, train recall=0.7414529914529915\n",
            "train precision=0.7538015930485156, specificity=0.9986960190227813\n",
            "train F-measure=0.7475763016157989, IoU=0.5969036697247706\n",
            "test mean loss=0.6737549230456352\n",
            "test accuracy=0.9491924038893885, recall=0.0772417780061665\n",
            "test precision=0.9246443675509419, specificity=0.9996358285272609\n",
            "test F-measure=0.1425734356937487, IoU=0.07675858547172221\n",
            "[epoch 311]\n",
            "train mean loss=0.028415244072675705\n",
            "train accuracy=0.9985237121582031, train recall=0.7527405602923264\n",
            "train precision=0.770573566084788, specificity=0.9992958905262835\n",
            "train F-measure=0.7615526802218114, IoU=0.6149253731343284\n",
            "test mean loss=0.7066591531038284\n",
            "test accuracy=0.948923673561151, recall=0.07043293936279547\n",
            "test precision=0.9412017167381974, specificity=0.9997454515726262\n",
            "test F-measure=0.1310583876172832, IoU=0.07012438845010073\n",
            "[epoch 312]\n",
            "train mean loss=0.026175225153565407\n",
            "train accuracy=0.9996871948242188, train recall=0.9375\n",
            "train precision=0.4934210526315789, specificity=0.9997061786433848\n",
            "train F-measure=0.646551724137931, IoU=0.47770700636942676\n",
            "test mean loss=0.7643281072378159\n",
            "test accuracy=0.9484494435701439, recall=0.05919193216855087\n",
            "test precision=0.97, specificity=0.9998940929900707\n",
            "test F-measure=0.11157525124107034, IoU=0.05908376879428077\n",
            "[epoch 313]\n",
            "train mean loss=0.026319213211536407\n",
            "train accuracy=0.9995918273925781, train recall=0.2878787878787879\n",
            "train precision=0.7450980392156863, specificity=0.9999503839518801\n",
            "train F-measure=0.4153005464480875, IoU=0.2620689655172414\n",
            "test mean loss=0.7963844686746597\n",
            "test accuracy=0.9481859824640287, recall=0.05386048304213772\n",
            "test precision=0.9761350407450524, specificity=0.9999238212735596\n",
            "test F-measure=0.10208802581116455, IoU=0.05378965262854027\n",
            "[epoch 314]\n",
            "train mean loss=0.025920458137989044\n",
            "train accuracy=0.9997444152832031, train recall=0.0945945945945946\n",
            "train precision=1.0, specificity=1.0\n",
            "train F-measure=0.17283950617283952, IoU=0.0945945945945946\n",
            "test mean loss=0.786105215549469\n",
            "test accuracy=0.9482913669064749, recall=0.05649409044193217\n",
            "test precision=0.9654226125137212, specificity=0.9998829448837624\n",
            "test F-measure=0.10674191395108927, IoU=0.05638001217987756\n",
            "[epoch 315]\n",
            "train mean loss=0.030091047286987305\n",
            "train accuracy=0.9979820251464844, train recall=0.5506692160611855\n",
            "train precision=0.9070866141732283, specificity=0.9997740312066734\n",
            "train F-measure=0.6853063652587745, IoU=0.5212669683257919\n",
            "test mean loss=0.7272413969039917\n",
            "test accuracy=0.9487972122302158, recall=0.06725334018499486\n",
            "test precision=0.9500907441016334, specificity=0.9997956180510137\n",
            "test F-measure=0.12561487702459506, IoU=0.06701657812199961\n",
            "[epoch 316]\n",
            "train mean loss=0.027722883969545364\n",
            "train accuracy=0.9987831115722656, train recall=0.7431457431457431\n",
            "train precision=0.7850609756097561, specificity=0.9994607020053471\n",
            "train F-measure=0.7635285396590067, IoU=0.617505995203837\n",
            "test mean loss=0.725010097026825\n",
            "test accuracy=0.948904353080036, recall=0.06911613566289826\n",
            "test precision=0.9526339088092076, specificity=0.999801192104168\n",
            "test F-measure=0.12888156909717025, IoU=0.06887942899209423\n",
            "[epoch 317]\n",
            "train mean loss=0.029661916196346283\n",
            "train accuracy=0.998779296875, train recall=1.0\n",
            "train precision=0.21182266009852216, specificity=0.9987788962748705\n",
            "train F-measure=0.3495934959349593, IoU=0.21182266009852216\n",
            "test mean loss=0.7710655480623245\n",
            "test accuracy=0.9483897257194245, recall=0.05710431654676259\n",
            "test precision=0.9855875831485588, specificity=0.9999516915393305\n",
            "test F-measure=0.1079538554948391, IoU=0.05705667158718952\n",
            "[epoch 318]\n",
            "train mean loss=0.026825454086065292\n",
            "train accuracy=0.9995803833007812, train recall=0.7593984962406015\n",
            "train precision=0.8145161290322581, specificity=0.9998243456876866\n",
            "train F-measure=0.7859922178988327, IoU=0.6474358974358975\n",
            "test mean loss=0.8073693364858627\n",
            "test accuracy=0.9479822392086331, recall=0.049171377183967115\n",
            "test precision=0.9928664072632944, specificity=0.9999795618051014\n",
            "test F-measure=0.09370218495623969, IoU=0.04915401162230713\n",
            "[epoch 319]\n",
            "train mean loss=0.026493092998862267\n",
            "train accuracy=0.9996795654296875, train recall=0.12222222222222222\n",
            "train precision=0.6875, specificity=0.999980919963061\n",
            "train F-measure=0.20754716981132074, IoU=0.11578947368421053\n",
            "test mean loss=0.8396511226892471\n",
            "test accuracy=0.9479190085431655, recall=0.04795092497430627\n",
            "test precision=0.9940079893475366, specificity=0.9999832778405375\n",
            "test F-measure=0.09148844904712299, IoU=0.04793706855032911\n",
            "[epoch 320]\n",
            "train mean loss=0.02702292613685131\n",
            "train accuracy=0.9996376037597656, train recall=0.24074074074074073\n",
            "train precision=0.6666666666666666, specificity=0.9999503884962372\n",
            "train F-measure=0.35374149659863946, IoU=0.21487603305785125\n",
            "test mean loss=0.8480900228023529\n",
            "test accuracy=0.9480173673561151, recall=0.04981372045220966\n",
            "test precision=0.9929577464788732, specificity=0.9999795618051014\n",
            "test F-measure=0.09486818765673742, IoU=0.04979612803801329\n",
            "[epoch 321]\n",
            "train mean loss=0.03499861806631088\n",
            "train accuracy=0.9963645935058594, train recall=0.4297820823244552\n",
            "train precision=0.984743411927878, specificity=0.9999577722156534\n",
            "train F-measure=0.5983986514959966, IoU=0.4269392663860493\n",
            "test mean loss=0.7224194407463074\n",
            "test accuracy=0.9490659425584532, recall=0.07358042137718397\n",
            "test precision=0.9370143149284254, specificity=0.9997138652714193\n",
            "test F-measure=0.1364462046990858, IoU=0.07321828060083094\n",
            "[epoch 322]\n",
            "train mean loss=0.030116543173789978\n",
            "train accuracy=0.9976654052734375, train recall=0.6937321937321937\n",
            "train precision=0.842560553633218, specificity=0.9993019866533712\n",
            "train F-measure=0.7609375, IoU=0.6141235813366961\n",
            "test mean loss=0.6645311042666435\n",
            "test accuracy=0.9496104288444245, recall=0.087198098663926\n",
            "test precision=0.9101575595038551, specificity=0.9995020512515608\n",
            "test F-measure=0.15914886133825729, IoU=0.08645395491020252\n",
            "[epoch 323]\n",
            "train mean loss=0.03253937512636185\n",
            "train accuracy=0.9971046447753906, train recall=0.7643004824259132\n",
            "train precision=0.7267365661861075, specificity=0.9984004173491425\n",
            "train F-measure=0.7450453476654351, IoU=0.5936830835117773\n",
            "test mean loss=0.6526692360639572\n",
            "test accuracy=0.9497614798785972, recall=0.0901207605344296\n",
            "test precision=0.9113348489769406, specificity=0.9994927611629705\n",
            "test F-measure=0.16402162794096156, IoU=0.08933745104906238\n",
            "[epoch 324]\n",
            "train mean loss=0.029504526406526566\n",
            "train accuracy=0.9982719421386719, train recall=0.944547134935305\n",
            "train precision=0.5471092077087795, specificity=0.9983830460659855\n",
            "train F-measure=0.6928813559322033, IoU=0.5300829875518672\n",
            "test mean loss=0.6745604202151299\n",
            "test accuracy=0.949550710993705, recall=0.08160971223021583\n",
            "test precision=0.9520419632821281, specificity=0.9997621737320888\n",
            "test F-measure=0.15033279100724745, IoU=0.08127558853633572\n",
            "[epoch 325]\n",
            "train mean loss=0.03202761709690094\n",
            "train accuracy=0.9973411560058594, train recall=0.6993998908892526\n",
            "train precision=0.8977591036414566, specificity=0.9994391324223717\n",
            "train F-measure=0.786261882858019, IoU=0.6478019201616978\n",
            "test mean loss=0.7128313332796097\n",
            "test accuracy=0.9490220323741008, recall=0.06956577595066804\n",
            "test precision=0.9756756756756757, specificity=0.999899667043225\n",
            "test F-measure=0.1298716872526682, IoU=0.06944533504328311\n",
            "[epoch 326]\n",
            "train mean loss=0.026773106306791306\n",
            "train accuracy=0.9990997314453125, train recall=0.8530612244897959\n",
            "train precision=0.511002444987775, specificity=0.999236346835994\n",
            "train F-measure=0.6391437308868502, IoU=0.4696629213483146\n",
            "test mean loss=0.7728215456008911\n",
            "test accuracy=0.9481772004271583, recall=0.05260791366906475\n",
            "test precision=0.9957446808510638, specificity=0.9999869938759736\n",
            "test F-measure=0.09993593850096091, IoU=0.052596089008766014\n",
            "[epoch 327]\n",
            "train mean loss=0.025962231680750847\n",
            "train accuracy=0.9991073608398438, train recall=0.5296610169491526\n",
            "train precision=0.9541984732824428, specificity=0.999954141062093\n",
            "train F-measure=0.6811989100817439, IoU=0.5165289256198347\n",
            "test mean loss=0.8122042566537857\n",
            "test accuracy=0.9476292013264388, recall=0.042458890030832475\n",
            "test precision=0.9977358490566037, specificity=0.9999944259468458\n",
            "test F-measure=0.08145158805951758, IoU=0.04245479944763801\n",
            "[epoch 328]\n",
            "train mean loss=0.026674197986721992\n",
            "train accuracy=0.99951171875, train recall=0.20270270270270271\n",
            "train precision=0.75, specificity=0.9999618314783432\n",
            "train F-measure=0.3191489361702128, IoU=0.189873417721519\n",
            "test mean loss=0.8246543407440186\n",
            "test accuracy=0.9474711246627698, recall=0.03953622816032888\n",
            "test precision=0.9983779399837794, specificity=0.9999962839645639\n",
            "test F-measure=0.07606042818746331, IoU=0.03953368874044576\n",
            "[epoch 329]\n",
            "train mean loss=0.026846276596188545\n",
            "train accuracy=0.9987602233886719, train recall=0.3199152542372881\n",
            "train precision=0.9741935483870968, specificity=0.9999847136873643\n",
            "train F-measure=0.4816586921850079, IoU=0.3172268907563025\n",
            "test mean loss=0.7820863127708435\n",
            "test accuracy=0.9478083548785972, recall=0.045702723535457346\n",
            "test precision=0.9985964912280701, specificity=0.9999962839645639\n",
            "test F-measure=0.08740517797364945, IoU=0.045699788040336566\n",
            "[epoch 330]\n",
            "train mean loss=0.026528147980570793\n",
            "train accuracy=0.9986000061035156, train recall=0.4906204906204906\n",
            "train precision=0.96045197740113, specificity=0.9999464526813819\n",
            "train F-measure=0.6494746895893029, IoU=0.4809052333804809\n",
            "test mean loss=0.7276299297809601\n",
            "test accuracy=0.9485425331609713, recall=0.05932040082219938\n",
            "test precision=0.9956873315363881, specificity=0.9999851358582555\n",
            "test F-measure=0.11196993119335574, IoU=0.05930516311328025\n",
            "[epoch 331]\n",
            "train mean loss=0.02532617747783661\n",
            "train accuracy=0.9992828369140625, train recall=0.9456521739130435\n",
            "train precision=0.4943181818181818, specificity=0.9993205069476256\n",
            "train F-measure=0.6492537313432836, IoU=0.48066298342541436\n",
            "test mean loss=0.7125498503446579\n",
            "test accuracy=0.9487919430080936, recall=0.06400950668036999\n",
            "test precision=0.9940149625935162, specificity=0.9999777037873833\n",
            "test F-measure=0.12027398086961767, IoU=0.06398484653910363\n",
            "[epoch 332]\n",
            "train mean loss=0.026375092566013336\n",
            "train accuracy=0.999053955078125, train recall=1.0\n",
            "train precision=0.18421052631578946, specificity=0.9990537529379445\n",
            "train F-measure=0.3111111111111111, IoU=0.18421052631578946\n",
            "test mean loss=0.7305942550301552\n",
            "test accuracy=0.9485091614208633, recall=0.05874229188078109\n",
            "test precision=0.9951033732317737, specificity=0.9999832778405375\n",
            "test F-measure=0.11093588888214957, IoU=0.05872531706533954\n",
            "[epoch 333]\n",
            "train mean loss=0.02543501928448677\n",
            "train accuracy=0.9986610412597656, train recall=0.5893129770992367\n",
            "train precision=0.8247863247863247, specificity=0.999686411283075\n",
            "train F-measure=0.6874443455031167, IoU=0.5237449118046132\n",
            "test mean loss=0.7392751574516296\n",
            "test accuracy=0.948347571942446, recall=0.05569116135662898\n",
            "test precision=0.996551724137931, specificity=0.9999888518936917\n",
            "test F-measure=0.10548728555785375, IoU=0.05568043157151114\n",
            "[epoch 334]\n",
            "train mean loss=0.026823893189430237\n",
            "train accuracy=0.99822998046875, train recall=0.58377239199157\n",
            "train precision=0.8892455858747994, specificity=0.9997358295526331\n",
            "train F-measure=0.7048346055979644, IoU=0.5442043222003929\n",
            "test mean loss=0.7389960587024689\n",
            "test accuracy=0.9484125590152878, recall=0.05691161356628983\n",
            "test precision=0.9960652051714446, specificity=0.9999869938759736\n",
            "test F-measure=0.10767127449491114, IoU=0.05689882156503869\n",
            "[epoch 335]\n",
            "train mean loss=0.034712158143520355\n",
            "train accuracy=0.9960594177246094, train recall=0.4679284963196635\n",
            "train precision=0.9769484083424808, specificity=0.9999193058768377\n",
            "train F-measure=0.632776395307501, IoU=0.4628185127405096\n",
            "test mean loss=0.6985509544610977\n",
            "test accuracy=0.948869224932554, recall=0.06600077081192189\n",
            "test precision=0.9856115107913669, specificity=0.9999442594684583\n",
            "test F-measure=0.12371692604075736, IoU=0.06593723929923635\n",
            "[epoch 336]\n",
            "train mean loss=0.026370488107204437\n",
            "train accuracy=0.9982757568359375, train recall=0.6586998087954111\n",
            "train precision=0.8788265306122449, specificity=0.9996361519429486\n",
            "train F-measure=0.7530054644808745, IoU=0.6038562664329535\n",
            "test mean loss=0.6710455790162086\n",
            "test accuracy=0.9492521217401079, recall=0.07492934224049332\n",
            "test precision=0.9628559636813867, specificity=0.9998327784053749\n",
            "test F-measure=0.13903870794719747, IoU=0.07471337987574457\n",
            "[epoch 337]\n",
            "train mean loss=0.0287993922829628\n",
            "train accuracy=0.9975090026855469, train recall=0.6716791979949874\n",
            "train precision=0.8925895087427144, specificity=0.9995048896940295\n",
            "train F-measure=0.7665355738291024, IoU=0.6214492753623189\n",
            "test mean loss=0.6438612341880798\n",
            "test accuracy=0.9495840827338129, recall=0.0845002569373073\n",
            "test precision=0.9296819787985866, specificity=0.9996302544741067\n",
            "test F-measure=0.15491962550786081, IoU=0.08396361895643849\n",
            "[epoch 338]\n",
            "train mean loss=0.02552974969148636\n",
            "train accuracy=0.9995956420898438, train recall=0.9210526315789473\n",
            "train precision=0.2536231884057971, specificity=0.9996070292171869\n",
            "train F-measure=0.3977272727272727, IoU=0.24822695035460993\n",
            "test mean loss=0.6581006795167923\n",
            "test accuracy=0.9494892367356115, recall=0.08270169578622816\n",
            "test precision=0.9289321789321789, specificity=0.9996339705095427\n",
            "test F-measure=0.15188156187330423, IoU=0.08218172533750359\n",
            "[epoch 339]\n",
            "train mean loss=0.024044902995228767\n",
            "train accuracy=0.9993972778320312, train recall=0.823170731707317\n",
            "train precision=0.5113636363636364, specificity=0.9995075959996946\n",
            "train F-measure=0.6308411214953271, IoU=0.46075085324232085\n",
            "test mean loss=0.7009164914488792\n",
            "test accuracy=0.9490782374100719, recall=0.07297019527235354\n",
            "test precision=0.9466666666666667, specificity=0.9997621737320888\n",
            "test F-measure=0.13549618320610687, IoU=0.07267144319344933\n",
            "[epoch 340]\n",
            "train mean loss=0.03101002611219883\n",
            "train accuracy=0.9969482421875, train recall=0.5926150121065376\n",
            "train precision=0.8851717902350814, specificity=0.9995124610352717\n",
            "train F-measure=0.709934735315446, IoU=0.5503091624508151\n",
            "test mean loss=0.6718394234776497\n",
            "test accuracy=0.9495998904001799, recall=0.08405061664953752\n",
            "test precision=0.9369853204439671, specificity=0.9996729888816219\n",
            "test F-measure=0.15426331456865805, IoU=0.08357818088911599\n",
            "[epoch 341]\n",
            "train mean loss=0.024527233093976974\n",
            "train accuracy=0.9990615844726562, train recall=0.5977443609022557\n",
            "train precision=0.5335570469798657, specificity=0.9994692184910531\n",
            "train F-measure=0.5638297872340425, IoU=0.3925925925925926\n",
            "test mean loss=0.6867711991071701\n",
            "test accuracy=0.9494031727742805, recall=0.07849434737923947\n",
            "test precision=0.9550605705353654, specificity=0.9997863279624235\n",
            "test F-measure=0.14506603353613298, IoU=0.07820549742408243\n",
            "[epoch 342]\n",
            "train mean loss=0.027040140703320503\n",
            "train accuracy=0.9981651306152344, train recall=0.7123287671232876\n",
            "train precision=0.8245243128964059, specificity=0.9993641040570928\n",
            "train F-measure=0.7643312101910827, IoU=0.6185567010309279\n",
            "test mean loss=0.7039515525102615\n",
            "test accuracy=0.9491168783723022, recall=0.07300231243576567\n",
            "test precision=0.9550420168067227, specificity=0.999801192104168\n",
            "test F-measure=0.13563671082468076, IoU=0.07275229651441922\n",
            "[epoch 343]\n",
            "train mean loss=0.02315448597073555\n",
            "train accuracy=0.9996871948242188, train recall=0.21568627450980393\n",
            "train precision=0.9166666666666666, specificity=0.9999923676357225\n",
            "train F-measure=0.3492063492063492, IoU=0.21153846153846154\n",
            "test mean loss=0.7140232473611832\n",
            "test accuracy=0.9489833914118705, recall=0.0706898766700925\n",
            "test precision=0.9519896193771626, specificity=0.9997937600332957\n",
            "test F-measure=0.13160727098780195, IoU=0.07043876212116364\n",
            "[epoch 344]\n",
            "train mean loss=0.02338065579533577\n",
            "train accuracy=0.9996109008789062, train recall=0.6666666666666666\n",
            "train precision=0.5833333333333334, specificity=0.9997710080986802\n",
            "train F-measure=0.6222222222222222, IoU=0.45161290322580644\n",
            "test mean loss=0.7321342751383781\n",
            "test accuracy=0.9488938146357914, recall=0.06882708119218911\n",
            "test precision=0.9537160658655985, specificity=0.9998067661573221\n",
            "test F-measure=0.12838870083575474, IoU=0.06859795134443021\n",
            "[epoch 345]\n",
            "train mean loss=0.022926248610019684\n",
            "train accuracy=0.9995956420898438, train recall=0.1746031746031746\n",
            "train precision=0.9166666666666666, specificity=0.9999923669366226\n",
            "train F-measure=0.29333333333333333, IoU=0.171875\n",
            "test mean loss=0.7361963167786598\n",
            "test accuracy=0.9489131351169064, recall=0.06950154162384378\n",
            "test precision=0.9499561018437226, specificity=0.9997881859801415\n",
            "test F-measure=0.12952654575926256, IoU=0.069248\n",
            "[epoch 346]\n",
            "train mean loss=0.023064617067575455\n",
            "train accuracy=0.9993324279785156, train recall=0.27941176470588236\n",
            "train precision=0.6705882352941176, specificity=0.999893105291288\n",
            "train F-measure=0.3944636678200692, IoU=0.24568965517241378\n",
            "test mean loss=0.729594886302948\n",
            "test accuracy=0.9490150067446043, recall=0.07200668036998972\n",
            "test precision=0.9436026936026936, specificity=0.9997510256257803\n",
            "test F-measure=0.13380281690140847, IoU=0.07169811320754717\n",
            "[epoch 347]\n",
            "train mean loss=0.02375590056180954\n",
            "train accuracy=0.9990043640136719, train recall=0.7190388170055453\n",
            "train precision=0.7811244979919679, specificity=0.999583338111566\n",
            "train F-measure=0.7487969201154958, IoU=0.5984615384615385\n",
            "test mean loss=0.7182128429412842\n",
            "test accuracy=0.9491730834082733, recall=0.07521839671120246\n",
            "test precision=0.9420756234915527, specificity=0.9997324454485998\n",
            "test F-measure=0.1393135447028731, IoU=0.07487212276214834\n",
            "[epoch 348]\n",
            "train mean loss=0.022040484473109245\n",
            "train accuracy=0.9998130798339844, train recall=0.7941176470588235\n",
            "train precision=0.6067415730337079, specificity=0.9998664509531586\n",
            "train F-measure=0.6878980891719746, IoU=0.5242718446601942\n",
            "test mean loss=0.7089374586939812\n",
            "test accuracy=0.949253878147482, recall=0.07688848920863309\n",
            "test precision=0.9410377358490566, specificity=0.9997212973422914\n",
            "test F-measure=0.14216152019002376, IoU=0.07651984913379786\n",
            "[epoch 349]\n",
            "train mean loss=0.022171979770064354\n",
            "train accuracy=0.9995841979980469, train recall=0.8444444444444444\n",
            "train precision=0.4444444444444444, specificity=0.9996374792981599\n",
            "train F-measure=0.5823754789272031, IoU=0.41081081081081083\n",
            "test mean loss=0.7142472267150879\n",
            "test accuracy=0.9492082115557554, recall=0.07592497430626927\n",
            "test precision=0.9418326693227091, specificity=0.9997287294131637\n",
            "test F-measure=0.1405219045354574, IoU=0.07557061568953392\n",
            "[epoch 350]\n",
            "train mean loss=0.023610828444361687\n",
            "train accuracy=0.9986419677734375, train recall=0.7048275862068966\n",
            "train precision=0.7825421133231241, specificity=0.9994568107138349\n",
            "train F-measure=0.741654571843251, IoU=0.5893886966551326\n",
            "test mean loss=0.730539433658123\n",
            "test accuracy=0.9490571605215827, recall=0.07271325796505652\n",
            "test precision=0.9449081803005008, specificity=0.9997547416612165\n",
            "test F-measure=0.13503519026601454, IoU=0.07240629397467059\n",
            "[epoch 351]\n",
            "train mean loss=0.021238937973976135\n",
            "train accuracy=0.9998550415039062, train recall=0.8\n",
            "train precision=0.8727272727272727, specificity=0.9999465697798675\n",
            "train F-measure=0.8347826086956521, IoU=0.7164179104477612\n",
            "test mean loss=0.7491835206747055\n",
            "test accuracy=0.9489377248201439, recall=0.07001541623843782\n",
            "test precision=0.9494773519163763, specificity=0.9997844699447054\n",
            "test F-measure=0.13041397463508017, IoU=0.06975553564571868\n",
            "[epoch 352]\n",
            "train mean loss=0.021047648042440414\n",
            "train accuracy=0.9997673034667969, train recall=0.4444444444444444\n",
            "train precision=0.9795918367346939, specificity=0.9999961837304798\n",
            "train F-measure=0.6114649681528662, IoU=0.44036697247706424\n",
            "test mean loss=0.7594586908817291\n",
            "test accuracy=0.9488621993030576, recall=0.06837744090441931\n",
            "test precision=0.951721054984354, specificity=0.9997993340864498\n",
            "test F-measure=0.12758817007760764, IoU=0.06814108308795289\n",
            "[epoch 353]\n",
            "train mean loss=0.027740556746721268\n",
            "train accuracy=0.9974517822265625, train recall=0.5297202797202797\n",
            "train precision=0.8233695652173914, specificity=0.9995019157088123\n",
            "train F-measure=0.6446808510638297, IoU=0.47566718995290425\n",
            "test mean loss=0.7367292121052742\n",
            "test accuracy=0.9490852630395683, recall=0.07284172661870504\n",
            "test precision=0.949748743718593, specificity=0.9997770378738332\n",
            "test F-measure=0.1353060493974466, IoU=0.07256206808292807\n",
            "[epoch 354]\n",
            "train mean loss=0.026905607432127\n",
            "train accuracy=0.9974937438964844, train recall=0.5858662613981763\n",
            "train precision=0.8731596828992072, specificity=0.9995705982486542\n",
            "train F-measure=0.7012278308321965, IoU=0.5399159663865546\n",
            "test mean loss=0.6815685927867889\n",
            "test accuracy=0.9496121852517986, recall=0.08417908530318602\n",
            "test precision=0.9380816034359342, specificity=0.9996785629347762\n",
            "test F-measure=0.15449454759799586, IoU=0.0837139480660513\n",
            "[epoch 355]\n",
            "train mean loss=0.022250570356845856\n",
            "train accuracy=0.9989852905273438, train recall=0.6814404432132964\n",
            "train precision=0.6196473551637279, specificity=0.9994231863795586\n",
            "train F-measure=0.6490765171503957, IoU=0.48046875\n",
            "test mean loss=0.6566277295351028\n",
            "test accuracy=0.9499248257643885, recall=0.08963900308324768\n",
            "test precision=0.9441813261163735, specificity=0.9996934270765206\n",
            "test F-measure=0.16373342719699635, IoU=0.08916648030414363\n",
            "[epoch 356]\n",
            "train mean loss=0.021011166274547577\n",
            "train accuracy=0.9997177124023438, train recall=0.8968253968253969\n",
            "train precision=0.6494252873563219, specificity=0.9997671915669916\n",
            "train F-measure=0.7533333333333334, IoU=0.6042780748663101\n",
            "test mean loss=0.6477869004011154\n",
            "test accuracy=0.9500653383543165, recall=0.09182297019527236\n",
            "test precision=0.9492031872509961, specificity=0.9997157232891373\n",
            "test F-measure=0.16744758111748856, IoU=0.09137396529131643\n",
            "[epoch 357]\n",
            "train mean loss=0.026473578065633774\n",
            "train accuracy=0.9974555969238281, train recall=0.7122641509433962\n",
            "train precision=0.7506213753106876, specificity=0.9988461774356773\n",
            "train F-measure=0.7309398951189996, IoU=0.5759694850603941\n",
            "test mean loss=0.6507590115070343\n",
            "test accuracy=0.9498405182104317, recall=0.08642728674203494\n",
            "test precision=0.9597004279600571, specificity=0.9997900439978595\n",
            "test F-measure=0.15857395403653507, IoU=0.0861147556721815\n",
            "[epoch 358]\n",
            "train mean loss=0.022091232240200043\n",
            "train accuracy=0.9988288879394531, train recall=0.7633711507293355\n",
            "train precision=0.745253164556962, specificity=0.9993843847862744\n",
            "train F-measure=0.7542033626901521, IoU=0.6053984575835476\n",
            "test mean loss=0.6799818500876427\n",
            "test accuracy=0.9493733138489209, recall=0.07589285714285714\n",
            "test precision=0.9788732394366197, specificity=0.9999052410963791\n",
            "test F-measure=0.1408643815201192, IoU=0.07576874979959598\n",
            "[epoch 359]\n",
            "train mean loss=0.022441452369093895\n",
            "train accuracy=0.9988555908203125, train recall=0.7393422655298416\n",
            "train precision=0.8759018759018758, specificity=0.9996709053546761\n",
            "train F-measure=0.8018494055482166, IoU=0.6692392502756339\n",
            "test mean loss=0.7108120173215866\n",
            "test accuracy=0.9489833914118705, recall=0.06812050359712231\n",
            "test precision=0.9855947955390335, specificity=0.9999424014507402\n",
            "test F-measure=0.1274333093006489, IoU=0.06805274809895082\n",
            "[epoch 360]\n",
            "train mean loss=0.025022033601999283\n",
            "train accuracy=0.9979209899902344, train recall=0.4974924774322969\n",
            "train precision=0.9185185185185185, specificity=0.9998315125197685\n",
            "train F-measure=0.6454131424853612, IoU=0.47646493756003844\n",
            "test mean loss=0.7027507573366165\n",
            "test accuracy=0.9490395964478417, recall=0.06882708119218911\n",
            "test precision=0.9902957486136784, specificity=0.9999609816279208\n",
            "test F-measure=0.12870870870870874, IoU=0.06878069133742017\n",
            "[epoch 361]\n",
            "train mean loss=0.021490877494215965\n",
            "train accuracy=0.9998550415039062, train recall=0.0\n",
            "train precision=0.0, specificity=1.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=0.6825401410460472\n",
            "test accuracy=0.9491783526303957, recall=0.07149280575539568\n",
            "test precision=0.9888938249666814, specificity=0.9999535495570486\n",
            "test F-measure=0.13334531404438854, IoU=0.07143544815634928\n",
            "[epoch 362]\n",
            "train mean loss=0.021040406078100204\n",
            "train accuracy=0.9994659423828125, train recall=0.0\n",
            "train precision=0.0, specificity=1.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=0.6515855938196182\n",
            "test accuracy=0.9495630058453237, recall=0.07891187050359712\n",
            "test precision=0.9851643945469126, specificity=0.9999312533444319\n",
            "test F-measure=0.1461195361284567, IoU=0.07881820806467135\n",
            "[epoch 363]\n",
            "train mean loss=0.021472789347171783\n",
            "train accuracy=0.9990692138671875, train recall=0.8162230671736375\n",
            "train precision=0.8667563930013459, specificity=0.9996212048745958\n",
            "train F-measure=0.8407310704960834, IoU=0.7252252252252253\n",
            "test mean loss=0.6396390050649643\n",
            "test accuracy=0.9497825567670863, recall=0.0832155704008222\n",
            "test precision=0.9825559347743648, specificity=0.9999145311849694\n",
            "test F-measure=0.15343617682764338, IoU=0.0830928099544609\n",
            "[epoch 364]\n",
            "train mean loss=0.020211374387145042\n",
            "train accuracy=0.9995460510253906, train recall=0.9537037037037037\n",
            "train precision=0.47465437788018433, specificity=0.9995649452746951\n",
            "train F-measure=0.6338461538461538, IoU=0.46396396396396394\n",
            "test mean loss=0.6459704041481018\n",
            "test accuracy=0.949742159397482, recall=0.08238052415210688\n",
            "test precision=0.9835122699386503, specificity=0.9999201052381236\n",
            "test F-measure=0.152027027027027, IoU=0.08226691042047532\n",
            "[epoch 365]\n",
            "train mean loss=0.020054295659065247\n",
            "train accuracy=0.9995689392089844, train recall=1.0\n",
            "train precision=0.4744186046511628, specificity=0.9995687714183222\n",
            "train F-measure=0.6435331230283912, IoU=0.4744186046511628\n",
            "test mean loss=0.6670422032475471\n",
            "test accuracy=0.9494383009217626, recall=0.07637461459403906\n",
            "test precision=0.9879518072289156, specificity=0.9999461174861763\n",
            "test F-measure=0.14178815252064517, IoU=0.0763035456441521\n",
            "[epoch 366]\n",
            "train mean loss=0.021551363170146942\n",
            "train accuracy=0.9987602233886719, train recall=0.6717241379310345\n",
            "train precision=0.8484320557491289, specificity=0.9996672009303073\n",
            "train F-measure=0.7498075442648191, IoU=0.5997536945812808\n",
            "test mean loss=0.6912999898195267\n",
            "test accuracy=0.9491221475944245, recall=0.07043293936279547\n",
            "test precision=0.9891745602165088, specificity=0.9999554075747666\n",
            "test F-measure=0.13150241357599016, IoU=0.07037869062901156\n",
            "[epoch 367]\n",
            "train mean loss=0.020572837442159653\n",
            "train accuracy=0.9993782043457031, train recall=0.770293609671848\n",
            "train precision=0.9369747899159664, specificity=0.9998853057557395\n",
            "train F-measure=0.8454976303317535, IoU=0.7323481116584565\n",
            "test mean loss=0.7084599211812019\n",
            "test accuracy=0.9489605581160072, recall=0.06734969167523125\n",
            "test precision=0.990552668871044, specificity=0.9999628396456388\n",
            "test F-measure=0.12612395874056478, IoU=0.06730645782513801\n",
            "[epoch 368]\n",
            "train mean loss=0.01947689615190029\n",
            "train accuracy=0.9997329711914062, train recall=0.2765957446808511\n",
            "train precision=0.9285714285714286, specificity=0.9999923678687274\n",
            "train F-measure=0.42622950819672134, IoU=0.2708333333333333\n",
            "test mean loss=0.7124043926596642\n",
            "test accuracy=0.9489535324865108, recall=0.0671891058581706\n",
            "test precision=0.9909995262908574, specificity=0.9999646976633569\n",
            "test F-measure=0.12584594098715673, IoU=0.06714813031616113\n",
            "[epoch 369]\n",
            "train mean loss=0.019756093621253967\n",
            "train accuracy=0.9994468688964844, train recall=0.608955223880597\n",
            "train precision=0.9357798165137615, specificity=0.9999465259024709\n",
            "train F-measure=0.7377938517179023, IoU=0.5845272206303725\n",
            "test mean loss=0.6996254473924637\n",
            "test accuracy=0.949201185926259, recall=0.07207091469681398\n",
            "test precision=0.9868073878627969, specificity=0.9999442594684583\n",
            "test F-measure=0.13433103861119425, IoU=0.07200154013989604\n",
            "[epoch 370]\n",
            "train mean loss=0.019580630585551262\n",
            "train accuracy=0.9994735717773438, train recall=0.36764705882352944\n",
            "train precision=0.8928571428571429, specificity=0.9999656409864854\n",
            "train F-measure=0.5208333333333335, IoU=0.352112676056338\n",
            "test mean loss=0.6810975521802902\n",
            "test accuracy=0.9494874803282374, recall=0.07826952723535457\n",
            "test precision=0.9759711653984782, specificity=0.9998885189369165\n",
            "test F-measure=0.1449171944221449, IoU=0.07811898961405309\n",
            "[epoch 371]\n",
            "train mean loss=0.01922452077269554\n",
            "train accuracy=0.9996337890625, train recall=0.782608695652174\n",
            "train precision=0.72, specificity=0.999786226904871\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=0.67417261749506\n",
            "test accuracy=0.9496754159172662, recall=0.0826374614594039\n",
            "test precision=0.9665664913598798, specificity=0.9998346364230929\n",
            "test F-measure=0.15225753003136278, IoU=0.08240192153722978\n",
            "[epoch 372]\n",
            "train mean loss=0.020218905061483383\n",
            "train accuracy=0.9991722106933594, train recall=0.8480725623582767\n",
            "train precision=0.7137404580152672, specificity=0.9994268311788554\n",
            "train F-measure=0.7751295336787565, IoU=0.6328257191201354\n",
            "test mean loss=0.6875366643071175\n",
            "test accuracy=0.9495454417715827, recall=0.08000385405960946\n",
            "test precision=0.96850699844479, specificity=0.9998495005648373\n",
            "test F-measure=0.14779874213836477, IoU=0.07979626485568761\n",
            "[epoch 373]\n",
            "train mean loss=0.026118451729416847\n",
            "train accuracy=0.9974021911621094, train recall=0.6570294784580499\n",
            "train precision=0.9384615384615385, specificity=0.9997081189031416\n",
            "train F-measure=0.772924308102701, IoU=0.6298913043478261\n",
            "test mean loss=0.6801484376192093\n",
            "test accuracy=0.9495893519559353, recall=0.08112795477903391\n",
            "test precision=0.965227359572029, specificity=0.9998309203876569\n",
            "test F-measure=0.14967558439249845, IoU=0.08089153617062157\n",
            "[epoch 374]\n",
            "train mean loss=0.024243874475359917\n",
            "train accuracy=0.9978065490722656, train recall=0.6922374429223744\n",
            "train precision=0.7610441767068273, specificity=0.9990882937686029\n",
            "train F-measure=0.725011956001913, IoU=0.568642160540135\n",
            "test mean loss=0.671358086168766\n",
            "test accuracy=0.949568275067446, recall=0.08019655704008222\n",
            "test precision=0.97121742512641, specificity=0.9998625066888638\n",
            "test F-measure=0.14815913608449283, IoU=0.0800064082024992\n",
            "[epoch 375]\n",
            "train mean loss=0.01896386593580246\n",
            "train accuracy=0.9997329711914062, train recall=0.8981481481481481\n",
            "train precision=0.6217948717948718, specificity=0.9997748400983071\n",
            "train F-measure=0.7348484848484848, IoU=0.5808383233532934\n",
            "test mean loss=0.6716975420713425\n",
            "test accuracy=0.9494347881070144, recall=0.07698484069886948\n",
            "test precision=0.9795668165100122, specificity=0.9999070991140971\n",
            "test F-measure=0.14275079653396064, IoU=0.07686141217212852\n",
            "[epoch 376]\n",
            "train mean loss=0.019058015197515488\n",
            "train accuracy=0.9993705749511719, train recall=0.7984496124031008\n",
            "train precision=0.7803030303030303, specificity=0.9996676306650825\n",
            "train F-measure=0.789272030651341, IoU=0.6518987341772152\n",
            "test mean loss=0.6817167103290558\n",
            "test accuracy=0.9491871346672662, recall=0.0718782117163412\n",
            "test precision=0.9859030837004406, specificity=0.9999405434330222\n",
            "test F-measure=0.1339879063641262, IoU=0.07180441478439425\n",
            "[epoch 377]\n",
            "train mean loss=0.01896735653281212\n",
            "train accuracy=0.9996528625488281, train recall=0.6829268292682927\n",
            "train precision=0.7417218543046358, specificity=0.9998511336743263\n",
            "train F-measure=0.7111111111111112, IoU=0.5517241379310345\n",
            "test mean loss=0.6948760151863098\n",
            "test accuracy=0.9489201607464028, recall=0.06654676258992806\n",
            "test precision=0.9913875598086125, specificity=0.999966555681075\n",
            "test F-measure=0.1247216035634744, IoU=0.0665083135391924\n",
            "[epoch 378]\n",
            "train mean loss=0.01851620338857174\n",
            "train accuracy=0.9997406005859375, train recall=0.5\n",
            "train precision=0.8823529411764706, specificity=0.9999694684456386\n",
            "train F-measure=0.6382978723404256, IoU=0.46875\n",
            "test mean loss=0.7023292183876038\n",
            "test accuracy=0.9488059942670863, recall=0.06423432682425488\n",
            "test precision=0.9945300845350572, specificity=0.9999795618051014\n",
            "test F-measure=0.12067457085105743, IoU=0.06421164157061675\n",
            "[epoch 379]\n",
            "train mean loss=0.01912219077348709\n",
            "train accuracy=0.99957275390625, train recall=0.30405405405405406\n",
            "train precision=0.8333333333333334, specificity=0.9999656483305088\n",
            "train F-measure=0.44554455445544555, IoU=0.28662420382165604\n",
            "test mean loss=0.6962783560156822\n",
            "test accuracy=0.9488323403776978, recall=0.06465184994861253\n",
            "test precision=0.9955489614243324, specificity=0.9999832778405375\n",
            "test F-measure=0.12141866216297725, IoU=0.06463316744260716\n",
            "[epoch 380]\n",
            "train mean loss=0.017746953293681145\n",
            "train accuracy=0.9998855590820312, train recall=0.8061224489795918\n",
            "train precision=0.8777777777777778, specificity=0.9999580226372469\n",
            "train F-measure=0.8404255319148936, IoU=0.7247706422018348\n",
            "test mean loss=0.6910184919834137\n",
            "test accuracy=0.9488621993030576, recall=0.06513360739979444\n",
            "test precision=0.9965601965601966, specificity=0.9999869938759736\n",
            "test F-measure=0.12227548159536944, IoU=0.0651189673441865\n",
            "[epoch 381]\n",
            "train mean loss=0.023320244625210762\n",
            "train accuracy=0.9976539611816406, train recall=0.5448613376835236\n",
            "train precision=0.9213793103448276, specificity=0.999781540560636\n",
            "train F-measure=0.6847770374167093, IoU=0.5206547155105222\n",
            "test mean loss=0.6574362069368362\n",
            "test accuracy=0.9492328012589928, recall=0.07219938335046248\n",
            "test precision=0.9929328621908127, specificity=0.9999702717165111\n",
            "test F-measure=0.13461077844311378, IoU=0.0721623009758603\n",
            "[epoch 382]\n",
            "train mean loss=0.02048567309975624\n",
            "train accuracy=0.9985694885253906, train recall=0.7376185458377239\n",
            "train precision=0.847457627118644, specificity=0.9995176017917647\n",
            "train F-measure=0.7887323943661972, IoU=0.6511627906976745\n",
            "test mean loss=0.6337889283895493\n",
            "test accuracy=0.9495068008093526, recall=0.07749871531346351\n",
            "test precision=0.9897456931911403, specificity=0.9999535495570486\n",
            "test F-measure=0.1437421814499315, IoU=0.07743653926382336\n",
            "[epoch 383]\n",
            "train mean loss=0.022605858743190765\n",
            "train accuracy=0.9979324340820312, train recall=0.7058404558404558\n",
            "train precision=0.8848214285714285, specificity=0.9995052542762906\n",
            "train F-measure=0.7852614896988906, IoU=0.6464448793215917\n",
            "test mean loss=0.6135312765836716\n",
            "test accuracy=0.9496964928057554, recall=0.08119218910585817\n",
            "test precision=0.9875, specificity=0.9999405434330222\n",
            "test F-measure=0.1500474833808167, IoU=0.0811088295687885\n",
            "[epoch 384]\n",
            "train mean loss=0.023457331582903862\n",
            "train accuracy=0.9978065490722656, train recall=0.7261904761904762\n",
            "train precision=0.8936006168080185, specificity=0.9994703471145432\n",
            "train F-measure=0.8012443829934324, IoU=0.668396770472895\n",
            "test mean loss=0.5992010608315468\n",
            "test accuracy=0.9497491850269785, recall=0.08234840698869476\n",
            "test precision=0.9853958493466565, specificity=0.9999293953267139\n",
            "test F-measure=0.15199478333036934, IoU=0.08224802720215564\n",
            "[epoch 385]\n",
            "train mean loss=0.02006659470498562\n",
            "train accuracy=0.9990310668945312, train recall=0.8728813559322034\n",
            "train precision=0.6798679867986799, specificity=0.9992586138371702\n",
            "train F-measure=0.7643784786641931, IoU=0.6186186186186187\n",
            "test mean loss=0.61253472417593\n",
            "test accuracy=0.9494664034397482, recall=0.07698484069886948\n",
            "test precision=0.9868258542610128, specificity=0.9999405434330222\n",
            "test F-measure=0.14282734991806942, IoU=0.07690580082135523\n",
            "[epoch 386]\n",
            "train mean loss=0.019273342564702034\n",
            "train accuracy=0.9989585876464844, train recall=0.7855822550831792\n",
            "train precision=0.7302405498281787, specificity=0.9993998539772098\n",
            "train F-measure=0.7569011576135352, IoU=0.6088825214899714\n",
            "test mean loss=0.6482717394828796\n",
            "test accuracy=0.9491168783723022, recall=0.06998329907502569\n",
            "test precision=0.9940693430656934, specificity=0.9999758457696653\n",
            "test F-measure=0.13076092174747958, IoU=0.06995409162412919\n",
            "[epoch 387]\n",
            "train mean loss=0.018451562151312828\n",
            "train accuracy=0.999755859375, train recall=0.8409090909090909\n",
            "train precision=0.39361702127659576, specificity=0.9997825257535292\n",
            "train F-measure=0.5362318840579711, IoU=0.36633663366336633\n",
            "test mean loss=0.7054602950811386\n",
            "test accuracy=0.9484968665692446, recall=0.05842112024665982\n",
            "test precision=0.9967123287671232, specificity=0.9999888518936917\n",
            "test F-measure=0.11037286490094354, IoU=0.058409864491683255\n",
            "[epoch 388]\n",
            "train mean loss=0.01852618157863617\n",
            "train accuracy=0.9992446899414062, train recall=0.22857142857142856\n",
            "train precision=0.8615384615384616, specificity=0.9999656356076198\n",
            "train F-measure=0.36129032258064514, IoU=0.2204724409448819\n",
            "test mean loss=0.7331101298332214\n",
            "test accuracy=0.9482351618705036, recall=0.0536035457348407\n",
            "test precision=0.9970131421744325, specificity=0.9999907099114097\n",
            "test F-measure=0.10173727522096922, IoU=0.05359493914774734\n",
            "[epoch 389]\n",
            "train mean loss=0.01716643013060093\n",
            "train accuracy=0.9998588562011719, train recall=0.68\n",
            "train precision=0.6181818181818182, specificity=0.999919876074996\n",
            "train F-measure=0.6476190476190475, IoU=0.4788732394366197\n",
            "test mean loss=0.7607966586947441\n",
            "test accuracy=0.9480138545413669, recall=0.049492548818088386\n",
            "test precision=0.9980569948186528, specificity=0.9999944259468458\n",
            "test F-measure=0.09430844553243574, IoU=0.0494877805966794\n",
            "[epoch 390]\n",
            "train mean loss=0.02551862597465515\n",
            "train accuracy=0.9966697692871094, train recall=0.4192256341789052\n",
            "train precision=0.9952456418383518, specificity=0.9999884901360466\n",
            "train F-measure=0.5899483325504932, IoU=0.4183877415056629\n",
            "test mean loss=0.6791231706738472\n",
            "test accuracy=0.9488499044514388, recall=0.06497302158273381\n",
            "test precision=0.9955708661417323, specificity=0.9999832778405375\n",
            "test F-measure=0.12198504582730342, IoU=0.06495424626745866\n",
            "[epoch 391]\n",
            "train mean loss=0.023151464760303497\n",
            "train accuracy=0.9980239868164062, train recall=0.611013986013986\n",
            "train precision=0.905440414507772, specificity=0.9997203065134099\n",
            "train F-measure=0.7296450939457203, IoU=0.5743631881676253\n",
            "test mean loss=0.6134307831525803\n",
            "test accuracy=0.9497474286196043, recall=0.08225205549845838\n",
            "test precision=0.9861378513669619, specificity=0.9999331113621499\n",
            "test F-measure=0.15183944505380487, IoU=0.08215706403182343\n",
            "[epoch 392]\n",
            "train mean loss=0.019531603902578354\n",
            "train accuracy=0.9988555908203125, train recall=0.868421052631579\n",
            "train precision=0.4657258064516129, specificity=0.9989880784181947\n",
            "train F-measure=0.6062992125984252, IoU=0.4350282485875706\n",
            "test mean loss=0.597736619412899\n",
            "test accuracy=0.9500864152428058, recall=0.08899665981500514\n",
            "test precision=0.9812322946175638, specificity=0.999901525060943\n",
            "test F-measure=0.1631919905771496, IoU=0.08884542627208311\n",
            "[epoch 393]\n",
            "train mean loss=0.019865117967128754\n",
            "train accuracy=0.998748779296875, train recall=0.9210526315789473\n",
            "train precision=0.4438405797101449, specificity=0.9988276983939086\n",
            "train F-measure=0.5990220048899755, IoU=0.42757417102966844\n",
            "test mean loss=0.6114848852157593\n",
            "test accuracy=0.9498914540242805, recall=0.08501413155190134\n",
            "test precision=0.9851135094901377, specificity=0.9999256792912777\n",
            "test F-measure=0.156520710758951, IoU=0.08490505517064409\n",
            "[epoch 394]\n",
            "train mean loss=0.018208418041467667\n",
            "train accuracy=0.9993247985839844, train recall=0.9256756756756757\n",
            "train precision=0.4521452145214521, specificity=0.9993664025404968\n",
            "train F-measure=0.6075388026607538, IoU=0.43630573248407645\n",
            "test mean loss=0.6442642882466316\n",
            "test accuracy=0.9493399421088129, recall=0.07412641315519013\n",
            "test precision=0.993542832544124, specificity=0.9999721297342291\n",
            "test F-measure=0.13795989121013777, IoU=0.0740907193990562\n",
            "[epoch 395]\n",
            "train mean loss=0.017114076763391495\n",
            "train accuracy=0.9996986389160156, train recall=0.6078431372549019\n",
            "train precision=0.6138613861386139, specificity=0.9998511688965891\n",
            "train F-measure=0.6108374384236452, IoU=0.4397163120567376\n",
            "test mean loss=0.6847640573978424\n",
            "test accuracy=0.9487866737859713, recall=0.06372045220966084\n",
            "test precision=0.9969849246231156, specificity=0.9999888518936917\n",
            "test F-measure=0.11978506309243493, IoU=0.0637081754543703\n",
            "[epoch 396]\n",
            "train mean loss=0.01757294312119484\n",
            "train accuracy=0.9996109008789062, train recall=0.0\n",
            "train precision=0.0, specificity=1.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=0.7090060040354729\n",
            "test accuracy=0.9485126742356115, recall=0.05864594039054471\n",
            "test precision=0.9978142076502732, specificity=0.9999925679291277\n",
            "test F-measure=0.11078080446520659, IoU=0.05863840719332049\n",
            "[epoch 397]\n",
            "train mean loss=0.01956726238131523\n",
            "train accuracy=0.9987564086914062, train recall=0.5454545454545454\n",
            "train precision=0.9717223650385605, specificity=0.9999579271068001\n",
            "train F-measure=0.6987060998151572, IoU=0.5369318181818182\n",
            "test mean loss=0.6850586459040642\n",
            "test accuracy=0.9488323403776978, recall=0.06468396711202466\n",
            "test precision=0.9950592885375494, specificity=0.9999814198228194\n",
            "test F-measure=0.12147165259348612, IoU=0.06466319912669363\n",
            "[epoch 398]\n",
            "train mean loss=0.022220611572265625\n",
            "train accuracy=0.9977378845214844, train recall=0.47992351816443596\n",
            "train precision=0.911070780399274, specificity=0.9998123310021525\n",
            "train F-measure=0.6286787726988102, IoU=0.45844748858447487\n",
            "test mean loss=0.6281167790293694\n",
            "test accuracy=0.9495700314748201, recall=0.07913669064748201\n",
            "test precision=0.9840255591054313, specificity=0.9999256792912777\n",
            "test F-measure=0.14649227110582638, IoU=0.07903515524762637\n",
            "[epoch 399]\n",
            "train mean loss=0.017980068922042847\n",
            "train accuracy=0.9996109008789062, train recall=0.0\n",
            "train precision=0.0, specificity=1.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=0.5933635756373405\n",
            "test accuracy=0.9501900432778777, recall=0.0930434224049332\n",
            "test precision=0.9602253894597282, specificity=0.9997770378738332\n",
            "test F-measure=0.16964834714373556, IoU=0.09268620424878424\n",
            "[epoch 400]\n",
            "train mean loss=0.016837075352668762\n",
            "train accuracy=0.9998970031738281, train recall=0.9642857142857143\n",
            "train precision=0.6835443037974683, specificity=0.9999046121913251\n",
            "train F-measure=0.7999999999999999, IoU=0.6666666666666666\n",
            "test mean loss=0.5837870463728905\n",
            "test accuracy=0.9505167350494604, recall=0.10129753340184995\n",
            "test precision=0.9428998505231689, specificity=0.9996451186158511\n",
            "test F-measure=0.1829413300078304, IoU=0.10067992466562391\n",
            "========================================\n",
            "Best Epoch : 2\n",
            "Best Accuracy : 0.9505167350494604\n",
            "Best Recall : 0.14558710174717368\n",
            "Best Precision : 0.9985964912280701\n",
            "Best Specificity : 0.9999962839645639\n",
            "Best F-measure : 0.19442419043534206\n",
            "Best IoU : 0.10767988217687721\n",
            "Elapsed time is (sec) 505.53863739967346\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "v9eKXRwxAQxj",
        "outputId": "26ba46d5-90a8-4ae7-8334-688f1e06bbaa"
      },
      "source": [
        "os.chdir(folder_path)\n",
        "!python src/tools/train.py --conf_file confs/CFOS_ndn.cfg --gpu_id 0"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Output Directory not exist! Create...\n",
            "Output Directory: ./results/result_train_nsn_Aug2Mon_2021_135808\n",
            "============================\n",
            "[Dataset]\n",
            "root_path = your_dataset\n",
            "input_format = tif\n",
            "image_dtype = uint16\n",
            "resolution = (1.0, 1.0, 2.18)\n",
            "split_list_train = your_dataset/split_list/train.txt\n",
            "split_list_validation = your_dataset/split_list/validation.txt\n",
            "\n",
            "[Model]\n",
            "model = NDN\n",
            "ndim = 3\n",
            "lossfun = F.softmax_cross_entropy\n",
            "ch_in = 12\n",
            "ch_base = 12\n",
            "ch_out = 12\n",
            "\n",
            "[Runtime]\n",
            "save_dir = results/result_train_nsn\n",
            "batchsize = 1\n",
            "val_batchsize = 1\n",
            "epoch = 400\n",
            "optimizer = Adam\n",
            "init_lr = 0.002\n",
            "weight_decay = 0.0001\n",
            "gpu = 0\n",
            "patch_size = (64, 64, 64)\n",
            "normalization = 0\n",
            "augmentation = 0\n",
            "class_weight = (1, 1)\n",
            "scaling = 1\n",
            "============================\n",
            "\n",
            "Loading datasets...\n",
            "-- train_dataset.size = 1\n",
            "-- validation_dataset.size = 1\n",
            "Initializing models...\n",
            "[epoch 1]\n",
            "/content/drive/MyDrive/QCANet/src/lib/trainer.py:528: UserWarning: The argument 'neighbors' is deprecated and will be removed in scikit-image 0.18, use 'connectivity' instead. For neighbors=4, use connectivity=1\n",
            "  markers_pr = morphology.label(pred, neighbors=4)\n",
            "/content/drive/MyDrive/QCANet/src/lib/trainer.py:538: UserWarning: The argument 'neighbors' is deprecated and will be removed in scikit-image 0.18, use 'connectivity' instead. For neighbors=4, use connectivity=1\n",
            "  markers_gt = morphology.label(y_patch, neighbors=4)\n",
            "/content/drive/MyDrive/QCANet/src/lib/trainer.py:673: UserWarning: The argument 'neighbors' is deprecated and will be removed in scikit-image 0.18, use 'connectivity' instead. For neighbors=4, use connectivity=1\n",
            "  markers_pr = morphology.label(seg_img, neighbors=4)\n",
            "/content/drive/MyDrive/QCANet/src/lib/trainer.py:683: UserWarning: The argument 'neighbors' is deprecated and will be removed in scikit-image 0.18, use 'connectivity' instead. For neighbors=4, use connectivity=1\n",
            "  markers_gt = morphology.label(gt, neighbors=4)\n",
            "train mean loss=2.29664945602417\n",
            "train recall=1.0, presicion=0.003218020917135961\n",
            "train F-measure=0.006415396952686447, IoU=0.003218020917135961\n",
            "test mean loss=9.997631549835205\n",
            "test recall=0.22413793103448276, presicion=0.08333333333333333\n",
            "test F-measure=0.12149532710280374, IoU=0.06467661691542288\n",
            "[epoch 2]\n",
            "train mean loss=2.0016682147979736\n",
            "train recall=1.0, presicion=0.024475524475524476\n",
            "train F-measure=0.04778156996587031, IoU=0.024475524475524476\n",
            "test mean loss=7.902403235435486\n",
            "test recall=0.22413793103448276, presicion=0.4482758620689655\n",
            "test F-measure=0.2988505747126437, IoU=0.17567567567567569\n",
            "[epoch 3]\n",
            "train mean loss=1.5635101795196533\n",
            "train recall=0.6, presicion=0.04054054054054054\n",
            "train F-measure=0.0759493670886076, IoU=0.039473684210526314\n",
            "test mean loss=6.270914196968079\n",
            "test recall=0.22413793103448276, presicion=0.48148148148148145\n",
            "test F-measure=0.30588235294117644, IoU=0.18055555555555555\n",
            "[epoch 4]\n",
            "train mean loss=1.1484001874923706\n",
            "train recall=0.8333333333333334, presicion=0.29411764705882354\n",
            "train F-measure=0.4347826086956522, IoU=0.2777777777777778\n",
            "test mean loss=4.931001007556915\n",
            "test recall=0.20689655172413793, presicion=0.8\n",
            "test F-measure=0.3287671232876712, IoU=0.19672131147540983\n",
            "[epoch 5]\n",
            "train mean loss=0.8847739696502686\n",
            "train recall=0.5, presicion=0.08333333333333333\n",
            "train F-measure=0.14285714285714285, IoU=0.07692307692307693\n",
            "test mean loss=3.6093437671661377\n",
            "test recall=0.22413793103448276, presicion=0.8666666666666667\n",
            "test F-measure=0.35616438356164387, IoU=0.21666666666666667\n",
            "[epoch 6]\n",
            "train mean loss=0.612785816192627\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=2.594940423965454\n",
            "test recall=0.22413793103448276, presicion=0.7647058823529411\n",
            "test F-measure=0.3466666666666667, IoU=0.20967741935483872\n",
            "[epoch 7]\n",
            "train mean loss=0.36307358741760254\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.9067508280277252\n",
            "test recall=0.22413793103448276, presicion=0.8666666666666667\n",
            "test F-measure=0.35616438356164387, IoU=0.21666666666666667\n",
            "[epoch 8]\n",
            "train mean loss=0.23238542675971985\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.519007682800293\n",
            "test recall=0.20689655172413793, presicion=0.8\n",
            "test F-measure=0.3287671232876712, IoU=0.19672131147540983\n",
            "[epoch 9]\n",
            "train mean loss=0.14608857035636902\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.350219801068306\n",
            "test recall=0.05172413793103448, presicion=1.0\n",
            "test F-measure=0.09836065573770492, IoU=0.05172413793103448\n",
            "[epoch 10]\n",
            "train mean loss=0.10589754581451416\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.2768512964248657\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 11]\n",
            "train mean loss=0.06570374220609665\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2675246894359589\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 12]\n",
            "train mean loss=0.03365374356508255\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.306351587176323\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 13]\n",
            "train mean loss=0.04154711589217186\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.3470360338687897\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 14]\n",
            "train mean loss=0.017873793840408325\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.4024404734373093\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 15]\n",
            "train mean loss=0.01271456852555275\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.4654421508312225\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 16]\n",
            "train mean loss=0.047055020928382874\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.4924588799476624\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 17]\n",
            "train mean loss=0.040853049606084824\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.4908551573753357\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 18]\n",
            "train mean loss=0.011048959568142891\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.4915131330490112\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 19]\n",
            "train mean loss=0.0236673466861248\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.4800831377506256\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 20]\n",
            "train mean loss=0.021987605839967728\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.4582753032445908\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 21]\n",
            "train mean loss=0.04715603590011597\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.4026569575071335\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 22]\n",
            "train mean loss=0.015027212910354137\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.3547140806913376\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 23]\n",
            "train mean loss=0.03195958584547043\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.3126249611377716\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 24]\n",
            "train mean loss=0.004184295888990164\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2945543378591537\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 25]\n",
            "train mean loss=0.01238813903182745\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2846331149339676\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 26]\n",
            "train mean loss=0.005710682366043329\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2827596366405487\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 27]\n",
            "train mean loss=0.012050779536366463\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2827033549547195\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 28]\n",
            "train mean loss=0.024927332997322083\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2760194540023804\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 29]\n",
            "train mean loss=0.030941907316446304\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2595890164375305\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 30]\n",
            "train mean loss=0.014165343716740608\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2442693710327148\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 31]\n",
            "train mean loss=0.007122371345758438\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2341424822807312\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 32]\n",
            "train mean loss=0.00476706400513649\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.229973703622818\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 33]\n",
            "train mean loss=0.0051133763045072556\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2316306680440903\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 34]\n",
            "train mean loss=0.022280894219875336\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2260189801454544\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 35]\n",
            "train mean loss=0.006316788494586945\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2246519923210144\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 36]\n",
            "train mean loss=0.00829280074685812\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2248769253492355\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 37]\n",
            "train mean loss=0.0038078331854194403\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2293144762516022\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 38]\n",
            "train mean loss=0.004006927367299795\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.236973524093628\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 39]\n",
            "train mean loss=0.020868415012955666\n",
            "train recall=0.16666666666666666, presicion=1.0\n",
            "train F-measure=0.2857142857142857, IoU=0.16666666666666666\n",
            "test mean loss=1.2354606240987778\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 40]\n",
            "train mean loss=0.026890452951192856\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2220265865325928\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 41]\n",
            "train mean loss=0.00266758119687438\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2125868201255798\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 42]\n",
            "train mean loss=0.01777374930679798\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.196684554219246\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 43]\n",
            "train mean loss=0.005732769146561623\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.1826209276914597\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 44]\n",
            "train mean loss=0.005426845513284206\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.1704302728176117\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 45]\n",
            "train mean loss=0.014071047306060791\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.155484989285469\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 46]\n",
            "train mean loss=0.007923373021185398\n",
            "train recall=0.25, presicion=1.0\n",
            "train F-measure=0.4, IoU=0.25\n",
            "test mean loss=1.142040193080902\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 47]\n",
            "train mean loss=0.004398884251713753\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.1315416395664215\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 48]\n",
            "train mean loss=0.01536762434989214\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1184101551771164\n",
            "test recall=0.0, presicion=0.0\n",
            "test F-measure=0.0, IoU=0.0\n",
            "[epoch 49]\n",
            "train mean loss=0.006663298700004816\n",
            "train recall=0.25, presicion=1.0\n",
            "train F-measure=0.4, IoU=0.25\n",
            "test mean loss=1.107365146279335\n",
            "test recall=0.034482758620689655, presicion=0.6666666666666666\n",
            "test F-measure=0.06557377049180328, IoU=0.03389830508474576\n",
            "[epoch 50]\n",
            "train mean loss=0.0025203940458595753\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.1017924547195435\n",
            "test recall=0.06896551724137931, presicion=0.8\n",
            "test F-measure=0.12698412698412698, IoU=0.06779661016949153\n",
            "[epoch 51]\n",
            "train mean loss=0.010315874591469765\n",
            "train recall=0.4, presicion=1.0\n",
            "train F-measure=0.5714285714285715, IoU=0.4\n",
            "test mean loss=1.096814587712288\n",
            "test recall=0.06896551724137931, presicion=0.8\n",
            "test F-measure=0.12698412698412698, IoU=0.06779661016949153\n",
            "[epoch 52]\n",
            "train mean loss=0.004459199495613575\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.0965912342071533\n",
            "test recall=0.08620689655172414, presicion=0.7142857142857143\n",
            "test F-measure=0.15384615384615388, IoU=0.08333333333333333\n",
            "[epoch 53]\n",
            "train mean loss=0.004454927518963814\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.0982970893383026\n",
            "test recall=0.10344827586206896, presicion=0.8571428571428571\n",
            "test F-measure=0.1846153846153846, IoU=0.1016949152542373\n",
            "[epoch 54]\n",
            "train mean loss=0.00618961313739419\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.102933093905449\n",
            "test recall=0.1206896551724138, presicion=0.875\n",
            "test F-measure=0.21212121212121213, IoU=0.11864406779661017\n",
            "[epoch 55]\n",
            "train mean loss=0.002486360026523471\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.1117455214262009\n",
            "test recall=0.1206896551724138, presicion=0.875\n",
            "test F-measure=0.21212121212121213, IoU=0.11864406779661017\n",
            "[epoch 56]\n",
            "train mean loss=0.0022647783625870943\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.1239071786403656\n",
            "test recall=0.1206896551724138, presicion=0.875\n",
            "test F-measure=0.21212121212121213, IoU=0.11864406779661017\n",
            "[epoch 57]\n",
            "train mean loss=0.005168590694665909\n",
            "train recall=0.25, presicion=1.0\n",
            "train F-measure=0.4, IoU=0.25\n",
            "test mean loss=1.1376514434814453\n",
            "test recall=0.1206896551724138, presicion=0.7777777777777778\n",
            "test F-measure=0.208955223880597, IoU=0.11666666666666667\n",
            "[epoch 58]\n",
            "train mean loss=0.0018690383294597268\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.1529725790023804\n",
            "test recall=0.1206896551724138, presicion=0.7777777777777778\n",
            "test F-measure=0.208955223880597, IoU=0.11666666666666667\n",
            "[epoch 59]\n",
            "train mean loss=0.005348218139261007\n",
            "train recall=0.25, presicion=1.0\n",
            "train F-measure=0.4, IoU=0.25\n",
            "test mean loss=1.1671425849199295\n",
            "test recall=0.1206896551724138, presicion=0.7777777777777778\n",
            "test F-measure=0.208955223880597, IoU=0.11666666666666667\n",
            "[epoch 60]\n",
            "train mean loss=0.004317183047533035\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.1804697811603546\n",
            "test recall=0.1206896551724138, presicion=0.7777777777777778\n",
            "test F-measure=0.208955223880597, IoU=0.11666666666666667\n",
            "[epoch 61]\n",
            "train mean loss=0.0024725545663386583\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1944726705551147\n",
            "test recall=0.13793103448275862, presicion=0.8\n",
            "test F-measure=0.2352941176470588, IoU=0.13333333333333333\n",
            "[epoch 62]\n",
            "train mean loss=0.017125682905316353\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.192887231707573\n",
            "test recall=0.15517241379310345, presicion=0.8181818181818182\n",
            "test F-measure=0.26086956521739135, IoU=0.15\n",
            "[epoch 63]\n",
            "train mean loss=0.0031076730228960514\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.188688650727272\n",
            "test recall=0.15517241379310345, presicion=0.8181818181818182\n",
            "test F-measure=0.26086956521739135, IoU=0.15\n",
            "[epoch 64]\n",
            "train mean loss=0.001799055258743465\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1888149231672287\n",
            "test recall=0.15517241379310345, presicion=0.8181818181818182\n",
            "test F-measure=0.26086956521739135, IoU=0.15\n",
            "[epoch 65]\n",
            "train mean loss=0.002758628223091364\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.1910987496376038\n",
            "test recall=0.15517241379310345, presicion=0.8181818181818182\n",
            "test F-measure=0.26086956521739135, IoU=0.15\n",
            "[epoch 66]\n",
            "train mean loss=0.005513053387403488\n",
            "train recall=0.25, presicion=1.0\n",
            "train F-measure=0.4, IoU=0.25\n",
            "test mean loss=1.1900886297225952\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 67]\n",
            "train mean loss=0.0029806382954120636\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.1870603412389755\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 68]\n",
            "train mean loss=0.003947364632040262\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.1820226162672043\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 69]\n",
            "train mean loss=0.0017726307269185781\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.1758664399385452\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 70]\n",
            "train mean loss=0.004816599655896425\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1707439720630646\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 71]\n",
            "train mean loss=0.006420533638447523\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1653243899345398\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 72]\n",
            "train mean loss=0.005669577978551388\n",
            "train recall=0.4, presicion=1.0\n",
            "train F-measure=0.5714285714285715, IoU=0.4\n",
            "test mean loss=1.1591236591339111\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 73]\n",
            "train mean loss=0.0014858553186058998\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1591054797172546\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 74]\n",
            "train mean loss=0.0015844095032662153\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.1609420478343964\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 75]\n",
            "train mean loss=0.0010952609591186047\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.1648919731378555\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 76]\n",
            "train mean loss=0.0014016341883689165\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1730908751487732\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 77]\n",
            "train mean loss=0.004278502427041531\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.1823585778474808\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 78]\n",
            "train mean loss=0.0015004535671323538\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1954108476638794\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 79]\n",
            "train mean loss=0.001634523505344987\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.2123860865831375\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 80]\n",
            "train mean loss=0.0019990417640656233\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.229074776172638\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 81]\n",
            "train mean loss=0.002055949531495571\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.2448853105306625\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 82]\n",
            "train mean loss=0.0028157569468021393\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.2586098462343216\n",
            "test recall=0.13793103448275862, presicion=0.8\n",
            "test F-measure=0.2352941176470588, IoU=0.13333333333333333\n",
            "[epoch 83]\n",
            "train mean loss=0.0045347739942371845\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.267171397805214\n",
            "test recall=0.13793103448275862, presicion=0.8\n",
            "test F-measure=0.2352941176470588, IoU=0.13333333333333333\n",
            "[epoch 84]\n",
            "train mean loss=0.01651490479707718\n",
            "train recall=0.42857142857142855, presicion=1.0\n",
            "train F-measure=0.6, IoU=0.42857142857142855\n",
            "test mean loss=1.2543271481990814\n",
            "test recall=0.13793103448275862, presicion=0.8\n",
            "test F-measure=0.2352941176470588, IoU=0.13333333333333333\n",
            "[epoch 85]\n",
            "train mean loss=0.0014913000632077456\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2421029657125473\n",
            "test recall=0.13793103448275862, presicion=0.8\n",
            "test F-measure=0.2352941176470588, IoU=0.13333333333333333\n",
            "[epoch 86]\n",
            "train mean loss=0.0008138060220517218\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.2328092604875565\n",
            "test recall=0.15517241379310345, presicion=0.8181818181818182\n",
            "test F-measure=0.26086956521739135, IoU=0.15\n",
            "[epoch 87]\n",
            "train mean loss=0.000967598520219326\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.2256285548210144\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 88]\n",
            "train mean loss=0.001546079758554697\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.2222113758325577\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 89]\n",
            "train mean loss=0.004006373696029186\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.2177173346281052\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 90]\n",
            "train mean loss=0.008994333446025848\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.208009272813797\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 91]\n",
            "train mean loss=0.001250513014383614\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.2017070949077606\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 92]\n",
            "train mean loss=0.0012695161858573556\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.2003497928380966\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 93]\n",
            "train mean loss=0.0015420138370245695\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.2033606469631195\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 94]\n",
            "train mean loss=0.008102808147668839\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.1991897225379944\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 95]\n",
            "train mean loss=0.006004896946251392\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.1933999359607697\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 96]\n",
            "train mean loss=0.007657341659069061\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.1838005781173706\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 97]\n",
            "train mean loss=0.013829873874783516\n",
            "train recall=0.42857142857142855, presicion=1.0\n",
            "train F-measure=0.6, IoU=0.42857142857142855\n",
            "test mean loss=1.1621676236391068\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 98]\n",
            "train mean loss=0.001559167169034481\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.146970272064209\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 99]\n",
            "train mean loss=0.0017538146348670125\n",
            "train recall=1.0, presicion=0.5\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1384852826595306\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 100]\n",
            "train mean loss=0.01408646535128355\n",
            "train recall=0.42857142857142855, presicion=1.0\n",
            "train F-measure=0.6, IoU=0.42857142857142855\n",
            "test mean loss=1.1219657212495804\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 101]\n",
            "train mean loss=0.01298338733613491\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1002803891897202\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 102]\n",
            "train mean loss=0.001867519924417138\n",
            "train recall=1.0, presicion=0.5\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0875143259763718\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 103]\n",
            "train mean loss=0.004477234557271004\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0793278366327286\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 104]\n",
            "train mean loss=0.0024561474565416574\n",
            "train recall=1.0, presicion=0.5\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0827572345733643\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 105]\n",
            "train mean loss=0.0012490588705986738\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0911069512367249\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 106]\n",
            "train mean loss=0.0021476284600794315\n",
            "train recall=1.0, presicion=0.5\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1062633395195007\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 107]\n",
            "train mean loss=0.008619319647550583\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.118260070681572\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 108]\n",
            "train mean loss=0.0055172024294734\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.1301018297672272\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 109]\n",
            "train mean loss=0.0010356936836615205\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1448400914669037\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 110]\n",
            "train mean loss=0.0033579394221305847\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.1572061777114868\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 111]\n",
            "train mean loss=0.00431815767660737\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1671519577503204\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 112]\n",
            "train mean loss=0.002731675747781992\n",
            "train recall=0.3333333333333333, presicion=1.0\n",
            "train F-measure=0.5, IoU=0.3333333333333333\n",
            "test mean loss=1.1793961077928543\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 113]\n",
            "train mean loss=0.005959495902061462\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.183531329035759\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 114]\n",
            "train mean loss=0.011864280328154564\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1713181287050247\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 115]\n",
            "train mean loss=0.001029796781949699\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1605133563280106\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 116]\n",
            "train mean loss=0.0011684198398143053\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.151684582233429\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 117]\n",
            "train mean loss=0.0033940179273486137\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.144508257508278\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 118]\n",
            "train mean loss=0.0009292225004173815\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1405477225780487\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 119]\n",
            "train mean loss=0.0037626270204782486\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.1363861709833145\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 120]\n",
            "train mean loss=0.0009202292421832681\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.133721798658371\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 121]\n",
            "train mean loss=0.0015738531947135925\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1348405480384827\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 122]\n",
            "train mean loss=0.0066846623085439205\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.1322798132896423\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 123]\n",
            "train mean loss=0.0022792876698076725\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1294672191143036\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 124]\n",
            "train mean loss=0.007897300645709038\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.120257928967476\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 125]\n",
            "train mean loss=0.006975316908210516\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.1086452305316925\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 126]\n",
            "train mean loss=0.0024666665121912956\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.10210382938385\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 127]\n",
            "train mean loss=0.0012162025086581707\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.099432647228241\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 128]\n",
            "train mean loss=0.004096475429832935\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=1.095937818288803\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 129]\n",
            "train mean loss=0.008407440967857838\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0884589701890945\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 130]\n",
            "train mean loss=0.002454982604831457\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0905585289001465\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 131]\n",
            "train mean loss=0.0011792507721111178\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0985721945762634\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 132]\n",
            "train mean loss=0.002405069302767515\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1137144565582275\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 133]\n",
            "train mean loss=0.003413666971027851\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.1280609220266342\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 134]\n",
            "train mean loss=0.007504643872380257\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.1361340135335922\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 135]\n",
            "train mean loss=0.001032613217830658\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1466061770915985\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 136]\n",
            "train mean loss=0.011425670236349106\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1423777490854263\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 137]\n",
            "train mean loss=0.0026094471104443073\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1370792239904404\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 138]\n",
            "train mean loss=0.003944059833884239\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=1.1300880014896393\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 139]\n",
            "train mean loss=0.0009163825307041407\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.125154510140419\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 140]\n",
            "train mean loss=0.0010083341039717197\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.1226067245006561\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 141]\n",
            "train mean loss=0.0023123747669160366\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.1236667186021805\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 142]\n",
            "train mean loss=0.006273170933127403\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.121288999915123\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 143]\n",
            "train mean loss=0.01335662230849266\n",
            "train recall=0.42857142857142855, presicion=1.0\n",
            "train F-measure=0.6, IoU=0.42857142857142855\n",
            "test mean loss=1.1031831055879593\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 144]\n",
            "train mean loss=0.001394588383845985\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0903808921575546\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 145]\n",
            "train mean loss=0.0015230276621878147\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.083672747015953\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 146]\n",
            "train mean loss=0.0010062846122309566\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0823920369148254\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 147]\n",
            "train mean loss=0.005789604503661394\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0813295990228653\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 148]\n",
            "train mean loss=0.005312005989253521\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0788972228765488\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 149]\n",
            "train mean loss=0.0022458028979599476\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0826562494039536\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 150]\n",
            "train mean loss=0.012891032733023167\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=1.069791316986084\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 151]\n",
            "train mean loss=0.0007866398082114756\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0608074516057968\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 152]\n",
            "train mean loss=0.0008782066288404167\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0568843632936478\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 153]\n",
            "train mean loss=0.0009757867082953453\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.055745467543602\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 154]\n",
            "train mean loss=0.006246604025363922\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0536215156316757\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 155]\n",
            "train mean loss=0.002437980379909277\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0581177473068237\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 156]\n",
            "train mean loss=0.005541861988604069\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0637405812740326\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 157]\n",
            "train mean loss=0.007496173493564129\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0612354129552841\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 158]\n",
            "train mean loss=0.006025766488164663\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0537063479423523\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 159]\n",
            "train mean loss=0.0017467668512836099\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0514313131570816\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 160]\n",
            "train mean loss=0.0045318459160625935\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=1.0493679642677307\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 161]\n",
            "train mean loss=0.0064270226284861565\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0454566180706024\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 162]\n",
            "train mean loss=0.0007426243973895907\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0435151308774948\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 163]\n",
            "train mean loss=0.003702957648783922\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0428783595561981\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 164]\n",
            "train mean loss=0.0032483325339853764\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.044461965560913\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 165]\n",
            "train mean loss=0.008954457938671112\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0387661755084991\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 166]\n",
            "train mean loss=0.0014648281503468752\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0385226607322693\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 167]\n",
            "train mean loss=0.0009373424691148102\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.040769711136818\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 168]\n",
            "train mean loss=0.0013617135118693113\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.048146054148674\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 169]\n",
            "train mean loss=0.001640294911339879\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0564925074577332\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 170]\n",
            "train mean loss=0.0010686091845855117\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.068194180727005\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 171]\n",
            "train mean loss=0.0007166644791141152\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0807119607925415\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 172]\n",
            "train mean loss=0.010377981700003147\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0749022960662842\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 173]\n",
            "train mean loss=0.0021341009996831417\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.067552626132965\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 174]\n",
            "train mean loss=0.012492436915636063\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=1.0394582450389862\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 175]\n",
            "train mean loss=0.005177721381187439\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0144056230783463\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 176]\n",
            "train mean loss=0.0007193989586085081\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9952457398176193\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 177]\n",
            "train mean loss=0.0014355364255607128\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9841646254062653\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 178]\n",
            "train mean loss=0.0013227645540609956\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9797055572271347\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 179]\n",
            "train mean loss=0.009536109864711761\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9665186256170273\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 180]\n",
            "train mean loss=0.0021234808955341578\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9626381322741508\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 181]\n",
            "train mean loss=0.003571023466065526\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9622161835432053\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 182]\n",
            "train mean loss=0.0011900842655450106\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9665781855583191\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 183]\n",
            "train mean loss=0.0063368575647473335\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=0.9727059006690979\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 184]\n",
            "train mean loss=0.010898429900407791\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9695713073015213\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 185]\n",
            "train mean loss=0.0028107871767133474\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9713515415787697\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 186]\n",
            "train mean loss=0.00883571058511734\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9686264842748642\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 187]\n",
            "train mean loss=0.004145948216319084\n",
            "train recall=0.8, presicion=1.0\n",
            "train F-measure=0.888888888888889, IoU=0.8\n",
            "test mean loss=0.9681663662195206\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 188]\n",
            "train mean loss=0.0044834609143435955\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.970247209072113\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 189]\n",
            "train mean loss=0.0013502838555723429\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9751090258359909\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 190]\n",
            "train mean loss=0.009443232789635658\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9727454483509064\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 191]\n",
            "train mean loss=0.008905919268727303\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9615319222211838\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 192]\n",
            "train mean loss=0.0078665716573596\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9492979720234871\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 193]\n",
            "train mean loss=0.002071854192763567\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9463027566671371\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 194]\n",
            "train mean loss=0.00600634329020977\n",
            "train recall=0.6, presicion=0.75\n",
            "train F-measure=0.6666666666666665, IoU=0.5\n",
            "test mean loss=0.9464958682656288\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 195]\n",
            "train mean loss=0.005932162515819073\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=0.946940541267395\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 196]\n",
            "train mean loss=0.005952921230345964\n",
            "train recall=0.8, presicion=0.8\n",
            "train F-measure=0.8000000000000002, IoU=0.6666666666666666\n",
            "test mean loss=0.9467226564884186\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 197]\n",
            "train mean loss=0.0013174989726394415\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.952443040907383\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 198]\n",
            "train mean loss=0.002975416136905551\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9702701568603516\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 199]\n",
            "train mean loss=0.0012091654352843761\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9952569901943207\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 200]\n",
            "train mean loss=0.0006605156231671572\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0212361216545105\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 201]\n",
            "train mean loss=0.00699861254543066\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0374241918325424\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 202]\n",
            "train mean loss=0.0030286149121820927\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0509959608316422\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 203]\n",
            "train mean loss=0.0027646725066006184\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.0608931630849838\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 204]\n",
            "train mean loss=0.01106469240039587\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=1.048900619149208\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 205]\n",
            "train mean loss=0.0007463168585672975\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.036308541893959\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 206]\n",
            "train mean loss=0.0009329712484031916\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0256627202033997\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 207]\n",
            "train mean loss=0.0027242624200880527\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.016089379787445\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 208]\n",
            "train mean loss=0.007142673246562481\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0013837218284607\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 209]\n",
            "train mean loss=0.0097439493983984\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9733639508485794\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 210]\n",
            "train mean loss=0.0008904269197955728\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9534683376550674\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 211]\n",
            "train mean loss=0.0014632383827120066\n",
            "train recall=1.0, presicion=0.5\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.942968875169754\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 212]\n",
            "train mean loss=0.0076448312029242516\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9318978041410446\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 213]\n",
            "train mean loss=0.002547853160649538\n",
            "train recall=1.0, presicion=0.6666666666666666\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.93291424959898\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 214]\n",
            "train mean loss=0.005697181914001703\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=0.9346954226493835\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 215]\n",
            "train mean loss=0.0012486276682466269\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9446703940629959\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 216]\n",
            "train mean loss=0.0009344266727566719\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9600011631846428\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 217]\n",
            "train mean loss=0.0007023659418337047\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9782999902963638\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 218]\n",
            "train mean loss=0.0018600983312353492\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9959228783845901\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 219]\n",
            "train mean loss=0.007052431348711252\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0029913932085037\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 220]\n",
            "train mean loss=0.0014573318185284734\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.0080702900886536\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 221]\n",
            "train mean loss=0.0026670177467167377\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.0104946494102478\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 222]\n",
            "train mean loss=0.000641812919639051\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.0119314193725586\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 223]\n",
            "train mean loss=0.0012858149129897356\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0128428637981415\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 224]\n",
            "train mean loss=0.008417075499892235\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0016210079193115\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 225]\n",
            "train mean loss=0.0014070039615035057\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9923304915428162\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 226]\n",
            "train mean loss=0.0010574698681011796\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9839216396212578\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 227]\n",
            "train mean loss=0.00278470222838223\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9768830761313438\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 228]\n",
            "train mean loss=0.002273230580613017\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9737011790275574\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 229]\n",
            "train mean loss=0.0005753865116275847\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9725925326347351\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 230]\n",
            "train mean loss=0.0010859468020498753\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9764585569500923\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 231]\n",
            "train mean loss=0.0007500984938815236\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9808079749345779\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 232]\n",
            "train mean loss=0.0014473269693553448\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9903061985969543\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 233]\n",
            "train mean loss=0.0005170688382349908\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0012989938259125\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 234]\n",
            "train mean loss=0.009336913004517555\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9992505013942719\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 235]\n",
            "train mean loss=0.0039666322991251945\n",
            "train recall=0.8, presicion=1.0\n",
            "train F-measure=0.888888888888889, IoU=0.8\n",
            "test mean loss=0.9975610822439194\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 236]\n",
            "train mean loss=0.006375821307301521\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9872555881738663\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 237]\n",
            "train mean loss=0.0012436225078999996\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9841064065694809\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 238]\n",
            "train mean loss=0.0018147717928513885\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9858760386705399\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 239]\n",
            "train mean loss=0.0007450725533999503\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9904971122741699\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 240]\n",
            "train mean loss=0.0006782825803384185\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9988025724887848\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 241]\n",
            "train mean loss=0.006770273670554161\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9970495104789734\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 242]\n",
            "train mean loss=0.002616693964228034\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9980870932340622\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 243]\n",
            "train mean loss=0.008060645312070847\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9844591468572617\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 244]\n",
            "train mean loss=0.002672945149242878\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.974263459444046\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 245]\n",
            "train mean loss=0.006927789188921452\n",
            "train recall=0.5, presicion=0.75\n",
            "train F-measure=0.6, IoU=0.42857142857142855\n",
            "test mean loss=0.9618247225880623\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 246]\n",
            "train mean loss=0.002908699680119753\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9553612470626831\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 247]\n",
            "train mean loss=0.008319893851876259\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=0.9388910681009293\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 248]\n",
            "train mean loss=0.002761109732091427\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9360301867127419\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 249]\n",
            "train mean loss=0.002757214941084385\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9402367547154427\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 250]\n",
            "train mean loss=0.0017337639583274722\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9534121155738831\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 251]\n",
            "train mean loss=0.0008752629510127008\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9703060984611511\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 252]\n",
            "train mean loss=0.000685415870975703\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9903664886951447\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 253]\n",
            "train mean loss=0.005161764100193977\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0052212327718735\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 254]\n",
            "train mean loss=0.0015599171165376902\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0205747485160828\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 255]\n",
            "train mean loss=0.008418574929237366\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.0137713700532913\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 256]\n",
            "train mean loss=0.0066129183396697044\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9969204813241959\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 257]\n",
            "train mean loss=0.0005857111536897719\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=0.9820810332894325\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 258]\n",
            "train mean loss=0.003184942062944174\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9675410985946655\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 259]\n",
            "train mean loss=0.006527939811348915\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9453773349523544\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 260]\n",
            "train mean loss=0.0022628786973655224\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9334517419338226\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 261]\n",
            "train mean loss=0.0015792460180819035\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9289703965187073\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 262]\n",
            "train mean loss=0.007251346483826637\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9224692434072495\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 263]\n",
            "train mean loss=0.0006799133261665702\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9192255139350891\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 264]\n",
            "train mean loss=0.0049988641403615475\n",
            "train recall=0.8, presicion=1.0\n",
            "train F-measure=0.888888888888889, IoU=0.8\n",
            "test mean loss=0.9198379516601562\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 265]\n",
            "train mean loss=0.005576626863330603\n",
            "train recall=0.8333333333333334, presicion=0.8333333333333334\n",
            "train F-measure=0.8333333333333334, IoU=0.7142857142857143\n",
            "test mean loss=0.9201012179255486\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 266]\n",
            "train mean loss=0.0020567651372402906\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9337686374783516\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 267]\n",
            "train mean loss=0.0029144929721951485\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.951605536043644\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 268]\n",
            "train mean loss=0.002740671392530203\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9690814390778542\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 269]\n",
            "train mean loss=0.0025449348613619804\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9851820766925812\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 270]\n",
            "train mean loss=0.004077230580151081\n",
            "train recall=0.8, presicion=1.0\n",
            "train F-measure=0.888888888888889, IoU=0.8\n",
            "test mean loss=0.9975547343492508\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 271]\n",
            "train mean loss=0.0032539383973926306\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0036275535821915\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 272]\n",
            "train mean loss=0.0007225535809993744\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0100343972444534\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 273]\n",
            "train mean loss=0.0011743884533643723\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0133401453495026\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 274]\n",
            "train mean loss=0.0007110219448804855\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.01756851375103\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 275]\n",
            "train mean loss=0.007152874022722244\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0079810470342636\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 276]\n",
            "train mean loss=0.001961015397682786\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9993364214897156\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 277]\n",
            "train mean loss=0.0010662423446774483\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9948011189699173\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 278]\n",
            "train mean loss=0.0066707804799079895\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9768257886171341\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 279]\n",
            "train mean loss=0.0024095484986901283\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.9631377607584\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 280]\n",
            "train mean loss=0.002999725751578808\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9538037404417992\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 281]\n",
            "train mean loss=0.0009323044214397669\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9514674097299576\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 282]\n",
            "train mean loss=0.0028459811583161354\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.9543365314602852\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 283]\n",
            "train mean loss=0.001040441682562232\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9631113335490227\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 284]\n",
            "train mean loss=0.005874674301594496\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9658379033207893\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 285]\n",
            "train mean loss=0.000992143526673317\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9746033102273941\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 286]\n",
            "train mean loss=0.0026836341712623835\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9841940179467201\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 287]\n",
            "train mean loss=0.0016488747205585241\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9956494122743607\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 288]\n",
            "train mean loss=0.00072243541944772\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.009709283709526\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 289]\n",
            "train mean loss=0.0008085704175755382\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.021797850728035\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 290]\n",
            "train mean loss=0.0007395870052278042\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0356067568063736\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 291]\n",
            "train mean loss=0.010769067332148552\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=1.01007579267025\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 292]\n",
            "train mean loss=0.0056730592623353004\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9746374040842056\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 293]\n",
            "train mean loss=0.0006732675828970969\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9455995932221413\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 294]\n",
            "train mean loss=0.0025183772668242455\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.925122931599617\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 295]\n",
            "train mean loss=0.001268803607672453\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9154627025127411\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 296]\n",
            "train mean loss=0.000633237068541348\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9100513458251953\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 297]\n",
            "train mean loss=0.005452240817248821\n",
            "train recall=0.8333333333333334, presicion=1.0\n",
            "train F-measure=0.9090909090909091, IoU=0.8333333333333334\n",
            "test mean loss=0.9054819419980049\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 298]\n",
            "train mean loss=0.006292279809713364\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.903126448392868\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 299]\n",
            "train mean loss=0.0010280143469572067\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9059095829725266\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 300]\n",
            "train mean loss=0.007050310727208853\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=0.9011748507618904\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 301]\n",
            "train mean loss=0.000673798262141645\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.8998283669352531\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 302]\n",
            "train mean loss=0.0015806915471330285\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9091498702764511\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 303]\n",
            "train mean loss=0.0033078365959227085\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.91916523873806\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 304]\n",
            "train mean loss=0.002752673113718629\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.928516149520874\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 305]\n",
            "train mean loss=0.0008456952637061477\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9411023929715157\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 306]\n",
            "train mean loss=0.0016785322222858667\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9510556906461716\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 307]\n",
            "train mean loss=0.002941761165857315\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9608765989542007\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 308]\n",
            "train mean loss=0.000515485240612179\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9707388281822205\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 309]\n",
            "train mean loss=0.001445614849217236\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9834944903850555\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 310]\n",
            "train mean loss=0.000733816996216774\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9980597347021103\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 311]\n",
            "train mean loss=0.0009792622877284884\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.0050254315137863\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 312]\n",
            "train mean loss=0.0006609039846807718\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.009238213300705\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 313]\n",
            "train mean loss=0.0006791530176997185\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.010576918721199\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 314]\n",
            "train mean loss=0.0010649531614035368\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.0064248889684677\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 315]\n",
            "train mean loss=0.0027969193179160357\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.9961315840482712\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 316]\n",
            "train mean loss=0.008068244904279709\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9696645140647888\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 317]\n",
            "train mean loss=0.004118863493204117\n",
            "train recall=0.8, presicion=1.0\n",
            "train F-measure=0.888888888888889, IoU=0.8\n",
            "test mean loss=0.9414467364549637\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 318]\n",
            "train mean loss=0.0026375995948910713\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.9189859330654144\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 319]\n",
            "train mean loss=0.00476772291585803\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.901863269507885\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 320]\n",
            "train mean loss=0.0022275473456829786\n",
            "train recall=1.0, presicion=0.6666666666666666\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.8995018005371094\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 321]\n",
            "train mean loss=0.0016027153469622135\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9096849411725998\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 322]\n",
            "train mean loss=0.00709114782512188\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=0.909985639154911\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 323]\n",
            "train mean loss=0.003816725220531225\n",
            "train recall=0.8, presicion=1.0\n",
            "train F-measure=0.888888888888889, IoU=0.8\n",
            "test mean loss=0.9156406447291374\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 324]\n",
            "train mean loss=0.0010047964751720428\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9293890297412872\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 325]\n",
            "train mean loss=0.0028869574889540672\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9662570953369141\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 326]\n",
            "train mean loss=0.004898697603493929\n",
            "train recall=0.8333333333333334, presicion=0.8333333333333334\n",
            "train F-measure=0.8333333333333334, IoU=0.7142857142857143\n",
            "test mean loss=0.9960452690720558\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 327]\n",
            "train mean loss=0.0005583651945926249\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0250820368528366\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 328]\n",
            "train mean loss=0.00405763927847147\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0396764129400253\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 329]\n",
            "train mean loss=0.0022293971851468086\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0462572276592255\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 330]\n",
            "train mean loss=0.006857520435005426\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.0241050720214844\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 331]\n",
            "train mean loss=0.004475263878703117\n",
            "train recall=0.8, presicion=1.0\n",
            "train F-measure=0.888888888888889, IoU=0.8\n",
            "test mean loss=0.9904511272907257\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 332]\n",
            "train mean loss=0.0007188916788436472\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9606862813234329\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 333]\n",
            "train mean loss=0.010691497474908829\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=0.9168207868933678\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 334]\n",
            "train mean loss=0.00617631571367383\n",
            "train recall=0.6666666666666666, presicion=0.8\n",
            "train F-measure=0.7272727272727272, IoU=0.5714285714285714\n",
            "test mean loss=0.8823520019650459\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 335]\n",
            "train mean loss=0.008374501019716263\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=0.8525792881846428\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 336]\n",
            "train mean loss=0.005515832919627428\n",
            "train recall=0.8333333333333334, presicion=1.0\n",
            "train F-measure=0.9090909090909091, IoU=0.8333333333333334\n",
            "test mean loss=0.8352104872465134\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 337]\n",
            "train mean loss=0.005624148994684219\n",
            "train recall=1.0, presicion=0.8333333333333334\n",
            "train F-measure=0.9090909090909091, IoU=0.8333333333333334\n",
            "test mean loss=0.8334568589925766\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 338]\n",
            "train mean loss=0.004956278949975967\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.845919169485569\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 339]\n",
            "train mean loss=0.005191142205148935\n",
            "train recall=1.0, presicion=0.8333333333333334\n",
            "train F-measure=0.9090909090909091, IoU=0.8333333333333334\n",
            "test mean loss=0.8681344911456108\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 340]\n",
            "train mean loss=0.0041246977634727955\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9065280184149742\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 341]\n",
            "train mean loss=0.0025789595674723387\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9603176116943359\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 342]\n",
            "train mean loss=0.0015068270731717348\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0234116911888123\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 343]\n",
            "train mean loss=0.0005733905709348619\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0890976637601852\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 344]\n",
            "train mean loss=0.0038343877531588078\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.1379376649856567\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 345]\n",
            "train mean loss=0.0009051338420249522\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=1.1785688698291779\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 346]\n",
            "train mean loss=0.004673607647418976\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.1849038302898407\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 347]\n",
            "train mean loss=0.0005820491351187229\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.189271241426468\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 348]\n",
            "train mean loss=0.0011573298834264278\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.1908035278320312\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 349]\n",
            "train mean loss=0.0010474470909684896\n",
            "train recall=0.0, presicion=0.0\n",
            "train F-measure=0.0, IoU=0.0\n",
            "test mean loss=1.1903486251831055\n",
            "test recall=0.1724137931034483, presicion=0.8333333333333334\n",
            "test F-measure=0.28571428571428575, IoU=0.16666666666666666\n",
            "[epoch 350]\n",
            "train mean loss=0.014784779399633408\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=1.121076062321663\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 351]\n",
            "train mean loss=0.005343493074178696\n",
            "train recall=0.6, presicion=1.0\n",
            "train F-measure=0.7499999999999999, IoU=0.6\n",
            "test mean loss=1.0458160787820816\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 352]\n",
            "train mean loss=0.00610014982521534\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9676605612039566\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 353]\n",
            "train mean loss=0.001137968385592103\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9096899181604385\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 354]\n",
            "train mean loss=0.0015820286935195327\n",
            "train recall=1.0, presicion=0.6666666666666666\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.8669929802417755\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 355]\n",
            "train mean loss=0.006499540060758591\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=0.8273783624172211\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 356]\n",
            "train mean loss=0.003180386498570442\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.8073412850499153\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 357]\n",
            "train mean loss=0.006990418769419193\n",
            "train recall=0.6666666666666666, presicion=0.8\n",
            "train F-measure=0.7272727272727272, IoU=0.5714285714285714\n",
            "test mean loss=0.7961270585656166\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 358]\n",
            "train mean loss=0.0025038314051926136\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.7993507608771324\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 359]\n",
            "train mean loss=0.005313540808856487\n",
            "train recall=1.0, presicion=0.8333333333333334\n",
            "train F-measure=0.9090909090909091, IoU=0.8333333333333334\n",
            "test mean loss=0.8128848895430565\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 360]\n",
            "train mean loss=0.0038520898669958115\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.8364409059286118\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 361]\n",
            "train mean loss=0.007318904157727957\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=0.8557200357317924\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 362]\n",
            "train mean loss=0.0020393216982483864\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.8891962170600891\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 363]\n",
            "train mean loss=0.002866104943677783\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9273771643638611\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 364]\n",
            "train mean loss=0.004796626977622509\n",
            "train recall=0.8333333333333334, presicion=1.0\n",
            "train F-measure=0.9090909090909091, IoU=0.8333333333333334\n",
            "test mean loss=0.9681645408272743\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 365]\n",
            "train mean loss=0.001087583601474762\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.011549949645996\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 366]\n",
            "train mean loss=0.0009822871070355177\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0551563948392868\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 367]\n",
            "train mean loss=0.010582119226455688\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=1.0702523291110992\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 368]\n",
            "train mean loss=0.0008136750548146665\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.087303787469864\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 369]\n",
            "train mean loss=0.011193376034498215\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=1.0719116032123566\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 370]\n",
            "train mean loss=0.0005413991748355329\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.0579441338777542\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 371]\n",
            "train mean loss=0.006162792909890413\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=1.0342314839363098\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 372]\n",
            "train mean loss=0.003146600443869829\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=1.008219912648201\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 373]\n",
            "train mean loss=0.0005206692731007934\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9864338338375092\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 374]\n",
            "train mean loss=0.0024387703742831945\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9694728255271912\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 375]\n",
            "train mean loss=0.0012882242444902658\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9592974781990051\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 376]\n",
            "train mean loss=0.0006824866868555546\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9524018913507462\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 377]\n",
            "train mean loss=0.0029090361203998327\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9464506506919861\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 378]\n",
            "train mean loss=0.001129765179939568\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9425657764077187\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 379]\n",
            "train mean loss=0.003178284503519535\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9408580511808395\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 380]\n",
            "train mean loss=0.0033927811309695244\n",
            "train recall=0.8, presicion=1.0\n",
            "train F-measure=0.888888888888889, IoU=0.8\n",
            "test mean loss=0.9442675858736038\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 381]\n",
            "train mean loss=0.0024470118805766106\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9485713988542557\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 382]\n",
            "train mean loss=0.0006677365745417774\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9563248679041862\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 383]\n",
            "train mean loss=0.003727078903466463\n",
            "train recall=0.8, presicion=1.0\n",
            "train F-measure=0.888888888888889, IoU=0.8\n",
            "test mean loss=0.9653051495552063\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 384]\n",
            "train mean loss=0.008337676525115967\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=0.9638059958815575\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 385]\n",
            "train mean loss=0.0006297879153862596\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9634166955947876\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 386]\n",
            "train mean loss=0.002354986034333706\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.9656476676464081\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 387]\n",
            "train mean loss=0.0005212526302784681\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9685023874044418\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 388]\n",
            "train mean loss=0.003919604234397411\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9727657809853554\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 389]\n",
            "train mean loss=0.0035701636224985123\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9795506969094276\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 390]\n",
            "train mean loss=0.0006326929433271289\n",
            "train recall=0.5, presicion=1.0\n",
            "train F-measure=0.6666666666666666, IoU=0.5\n",
            "test mean loss=0.9853977635502815\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 391]\n",
            "train mean loss=0.005456879734992981\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9845106974244118\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 392]\n",
            "train mean loss=0.0032266322523355484\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.9816609919071198\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 393]\n",
            "train mean loss=0.0038626447785645723\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9815912246704102\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 394]\n",
            "train mean loss=0.005456278566271067\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9791079089045525\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 395]\n",
            "train mean loss=0.0018138124141842127\n",
            "train recall=0.75, presicion=1.0\n",
            "train F-measure=0.8571428571428571, IoU=0.75\n",
            "test mean loss=0.9751293063163757\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 396]\n",
            "train mean loss=0.008641818538308144\n",
            "train recall=0.5714285714285714, presicion=1.0\n",
            "train F-measure=0.7272727272727273, IoU=0.5714285714285714\n",
            "test mean loss=0.9610216319561005\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 397]\n",
            "train mean loss=0.005407499615103006\n",
            "train recall=0.6666666666666666, presicion=1.0\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9445807188749313\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 398]\n",
            "train mean loss=0.0006613598670810461\n",
            "train recall=1.0, presicion=1.0\n",
            "train F-measure=1.0, IoU=1.0\n",
            "test mean loss=0.9326959252357483\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 399]\n",
            "train mean loss=0.004532228223979473\n",
            "train recall=0.8333333333333334, presicion=0.8333333333333334\n",
            "train F-measure=0.8333333333333334, IoU=0.7142857142857143\n",
            "test mean loss=0.9197796285152435\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "[epoch 400]\n",
            "train mean loss=0.0013931330759078264\n",
            "train recall=1.0, presicion=0.6666666666666666\n",
            "train F-measure=0.8, IoU=0.6666666666666666\n",
            "test mean loss=0.9136134907603264\n",
            "test recall=0.1896551724137931, presicion=0.8461538461538461\n",
            "test F-measure=0.30985915492957744, IoU=0.18333333333333332\n",
            "========================================\n",
            "Best Epoch (F-measure) : 7\n",
            "Best Recall : 0.22413793103448276\n",
            "Best Precision : 1.0\n",
            "Best F-measure : 0.35616438356164387\n",
            "Best IoU : 0.21666666666666667\n",
            "Elapsed time is (sec) 1654.3818180561066\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WrnpoX4ofJ90"
      },
      "source": [
        "Plotting the results"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "MaQkw54ellpO",
        "outputId": "79618f8b-d6fa-4ea3-f064-e446e10de47e"
      },
      "source": [
        "#go to the results file in the drive (./QCANet/results) and copy paste the path to the latest results file\n",
        "\n",
        "Train_result = folder_path + \"/results/result_train_nsn_Aug2Mon_2021_134931/TrainResult.csv\" \n",
        "\n",
        "TrainResult = pd.read_csv(Train_result)\n",
        "TrainResult.head()\n",
        "\n",
        "plt.plot( 'Epoch', 'Accuracy', data=TrainResult, marker='', color='blue', linewidth=2)\n",
        "plt.plot( 'Epoch', 'Recall', data=TrainResult, marker='', color='orange', linewidth=2)\n",
        "plt.plot( 'Epoch', 'Precision', data=TrainResult, marker='', color='green', linewidth=2)\n",
        "plt.plot( 'Epoch', 'Specificity', data=TrainResult, marker='', color='red', linewidth=2)\n",
        "plt.plot( 'Epoch', 'F-measure', data=TrainResult, marker='', color='purple', linewidth=2)\n",
        "plt.plot( 'Epoch', 'IoU', data=TrainResult, marker='', color='olive', linewidth=2)\n",
        "plt.legend()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7ffa63942b10>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 60
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "yQf3VYVsBxhQ",
        "outputId": "d45a7c15-4a3b-46ae-c638-f109c30cb640"
      },
      "source": [
        "Train_result = folder_path + \"/results/result_train_nsn_Aug2Mon_2021_135808/TrainResult.csv\" \n",
        "\n",
        "TrainResult = pd.read_csv(Train_result)\n",
        "TrainResult.head()\n",
        "\n",
        "plt.plot( 'Epoch', 'Accuracy', data=TrainResult, marker='', color='blue', linewidth=2)\n",
        "plt.plot( 'Epoch', 'Recall', data=TrainResult, marker='', color='orange', linewidth=2)\n",
        "plt.plot( 'Epoch', 'Precision', data=TrainResult, marker='', color='green', linewidth=2)\n",
        "plt.plot( 'Epoch', 'Specificity', data=TrainResult, marker='', color='red', linewidth=2)\n",
        "plt.plot( 'Epoch', 'F-measure', data=TrainResult, marker='', color='purple', linewidth=2)\n",
        "plt.plot( 'Epoch', 'IoU', data=TrainResult, marker='', color='olive', linewidth=2)\n",
        "plt.legend()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7ffa63988250>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 61
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aFUdHuPSfxk4"
      },
      "source": [
        "Run the trained model on the validation data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W0CUmNHo8tNb"
      },
      "source": [
        "#go to the latest results file in the drive and rename the .npz file to NSN.npz\n",
        "#add the path to the NSN.npz file in the -ms argument below\n",
        "\n",
        "!python src/tools/qca_net.py -i your_dataset/input/ -ms results/result_train_nsn_Aug2Mon_2021_134931/NSN.npz -md results/result_train_nsn_Aug2Mon_2021_135808/NDN.npz --gpu 0\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "M1af9X0fgXzC"
      },
      "source": [
        "Extract the results"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "21ZZxhHV_eES",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "10c4ccc9-3b0e-4af2-e7d4-6dcef566031e"
      },
      "source": [
        "os.chdir(folder_path)\n",
        "\n",
        "#change the path below to the results file created by the previous command\n",
        "\n",
        "!python src/tools/extract.py -i results/result_qcanet_Aug2Mon_2021_130813/WatershedSegmentationImages/"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Output Directory not exist! Create...\n",
            "Output Directory: ./results/result_extract_Aug2Mon_2021_130857\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TTmUfCFUhGjV"
      },
      "source": [
        "Visualisation of the Results"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 749
        },
        "id": "aVfN0vuw_-Db",
        "outputId": "357152ff-1d2a-4b3f-fe68-93b76a7881db"
      },
      "source": [
        "#Plot the predictions\n",
        "# add the path to the results (same as the previous command)\n",
        "\n",
        "results_ = folder_path + \"/results/result_qcanet_Aug2Mon_2021_130813/SegmentationImages/CFOS_noisy2.tif\"\n",
        "results=io.imread(results_)\n",
        "\n",
        "fig = plt.figure(figsize=(15,15))\n",
        "plt.title('Brain_cells', fontsize = 20)\n",
        "ax = [plt.subplot(4,8,i+1) for i in range(32)]#398\n",
        "im_ind = 0\n",
        "for a in ax:\n",
        "    a.imshow(results[im_ind,:,:],cmap = \"rainbow\")\n",
        "    a.set_xticklabels([])\n",
        "    a.set_yticklabels([])\n",
        "    im_ind+=1\n",
        "plt.subplots_adjust(wspace=0, hspace=0)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x1080 with 32 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 218
        },
        "id": "v_y99SgTAOLy",
        "outputId": "e27c96b2-5dfe-4059-ce0a-accc2f153e67"
      },
      "source": [
        "#Plot the predictions along with the groundtruth and raw data (one layer)\n",
        "#add the path to the results (same as the previous command)\n",
        "\n",
        "results_ = folder_path + \"/results/result_qcanet_Aug2Mon_2021_130813/SegmentationImages/CFOS_noisy2.tif\"\n",
        "results=io.imread(results_)\n",
        "\n",
        "\n",
        "data_ = folder_path + \"/your_dataset/input_gt/CFOS_noisy2.tif\"\n",
        "data_gt=io.imread(data_)\n",
        "\n",
        "data_ = folder_path + \"/your_dataset/input/CFOS_noisy2.tif\"\n",
        "data_raw=io.imread(data_)\n",
        "\n",
        "\n",
        "fig, (ax1, ax2, ax3) = plt.subplots(nrows=1, ncols=3, figsize=(8, 3),\n",
        "                                    sharex=True, sharey=True)\n",
        "\n",
        "ax1.imshow(results[20], cmap=\"inferno\")\n",
        "ax1.axis('off')\n",
        "ax1.set_title('Prediction Image 10/32', fontsize=12)\n",
        "\n",
        "ax2.imshow(data_gt[20], cmap=\"inferno\")\n",
        "ax2.axis('off')\n",
        "ax2.set_title(r'Groundtruth Image', fontsize=12)\n",
        "\n",
        "ax3.imshow(data_raw[20], cmap=\"inferno\")\n",
        "ax3.axis('off')\n",
        "ax3.set_title(r'Raw Image', fontsize=12)\n",
        "\n",
        "fig.tight_layout()\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x216 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q2wBmTihCmww"
      },
      "source": [
        "#Plot the predictions along with the groundtruth and raw data (all layers)\n",
        "#add the path to the results (same as the previous command)\n",
        "\n",
        "def show_result(sample_idx):\n",
        "  index = sample_idx\n",
        "\n",
        "  #testing_rawdata_normalized, testing_mask_normalised\n",
        "\n",
        "  fig, (ax1, ax2, ax3) = plt.subplots(nrows=1, ncols=3, figsize=(10, 20),\n",
        "                                      sharex=True, sharey=True)\n",
        "\n",
        "  ax1.imshow(results[index], cmap=\"inferno\")\n",
        "  ax1.axis('off')\n",
        "  ax1.set_title(f'Prediction Image {index}/{results.shape[0]}', fontsize=20)\n",
        "\n",
        "  ax2.imshow(data_gt[index], cmap=\"inferno\")\n",
        "  ax2.axis('off')\n",
        "  ax2.set_title(r'Groundtruth Image', fontsize=20)\n",
        "\n",
        "  ax3.imshow(data_raw[index], cmap=\"inferno\")\n",
        "  ax3.axis('off')\n",
        "  ax3.set_title(r'Raw Image', fontsize=20)\n",
        "\n",
        "  fig.tight_layout()\n",
        "  plt.show()\n",
        "\n",
        "for i in range(results.shape[0]):\n",
        "  show_result(i)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "N4uzO-VahVr2"
      },
      "source": [
        "Compute the IoU and the DICE scores"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RtQ4gUTgAnc0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c1662ec4-7e27-4f10-b62a-bae2a31d343d"
      },
      "source": [
        "results_ = folder_path + \"/results/result_qcanet_Aug2Mon_2021_130813/SegmentationImages/CFOS_noisy2.tif\"\n",
        "results=io.imread(results_)\n",
        "data_ = folder_path + \"/your_dataset/input_gt/CFOS_noisy2.tif\"\n",
        "data_gt=io.imread(data_)\n",
        "data_ = folder_path + \"/your_dataset/input/CFOS_noisy2.tif\"\n",
        "data_raw=io.imread(data_)\n",
        "\n",
        "results[results>0] = 1 # change to binary \n",
        "data_gt[data_gt>0] = 1\n",
        "y_pred = results[:32,:,:].astype(\"uint8\")\n",
        "y_true = data_gt[:32,:,:].astype(\"uint8\")\n",
        "\n",
        "axes = (1,2) # W,H axes of each image\n",
        "intersection = np.sum(np.logical_and(y_pred, y_true), axis=axes)\n",
        "union = np.sum(np.logical_or(y_pred, y_true), axis=axes)\n",
        "mask_sum = np.sum(np.abs(y_true), axis=axes) + np.sum(np.abs(y_pred), axis=axes)\n",
        "\n",
        "smooth = .001\n",
        "iou = (intersection + smooth) / (union + smooth)\n",
        "dice = 2 * (intersection + smooth)/(mask_sum + smooth)\n",
        "\n",
        "iou = np.mean(iou)\n",
        "dice = np.mean(dice)\n",
        "print(\"IoU score is \" + str(iou))\n",
        "print(\"Dice score is \" + str(dice))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "IoU score is 0.47300693323003745\n",
            "Dice score is 0.6353340942249964\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RuJxCQ_piaeD"
      },
      "source": [
        "3D plots"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tuxCBr_aDj-s"
      },
      "source": [
        "z, x, y = results.nonzero()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KBHB7qoYDrXT"
      },
      "source": [
        "fig = plt.figure()\n",
        "ax = plt.axes(projection='3d')\n",
        "ax.scatter3D(x, y, z, c=z, alpha=1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Az97gI8nD-WU"
      },
      "source": [
        "delta = data_gt\n",
        "\n",
        "z, x, y = delta.nonzero()\n",
        "\n",
        "fig = plt.figure()\n",
        "ax = plt.axes(projection='3d')\n",
        "ax.scatter3D(x, y, z, c=z, alpha=1)"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}