hubmap-vanilla-unet-w-b-pytorch-2-3-train.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b1687980",
   "metadata": {
    "papermill": {
     "duration": 0.008388,
     "end_time": "2022-07-24T09:02:00.238524",
     "exception": false,
     "start_time": "2022-07-24T09:02:00.230136",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### Changelog\n",
    "\n",
    "|| Version | Comments | LB |\n",
    "|---|  --- | --- | --- |\n",
    "|**Best**|8| Vanilla Unet w/ 256x256, 40 epoch, no aug, no lr sch, no pre/post processing, L2 norm | `0.04` |"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "3ea36fe9",
   "metadata": {
    "papermill": {
     "duration": 11.427272,
     "end_time": "2022-07-24T09:02:11.674696",
     "exception": false,
     "start_time": "2022-07-24T09:02:00.247424",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33mYou are using pip version 10.0.1, however version 22.2 is available.\n",
      "You should consider upgrading via the 'pip install --upgrade pip' command.\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "# !pip3 install -q tifffile"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c9682b07",
   "metadata": {
    "papermill": {
     "duration": 0.008572,
     "end_time": "2022-07-24T09:02:11.692668",
     "exception": false,
     "start_time": "2022-07-24T09:02:11.684096",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Import, SEED, Config:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "87e5422d",
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:11.712277Z",
     "iopub.status.busy": "2022-07-24T09:02:11.711957Z",
     "iopub.status.idle": "2022-07-24T09:02:16.230412Z",
     "shell.execute_reply": "2022-07-24T09:02:16.229382Z"
    },
    "papermill": {
     "duration": 4.531182,
     "end_time": "2022-07-24T09:02:16.233067",
     "exception": false,
     "start_time": "2022-07-24T09:02:11.701885",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/lakshita/somusan/hubmap_kaggle/.venv/lib/python3.7/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "import numpy as np # linear algebra\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "import os\n",
    "import warnings\n",
    "import cv2\n",
    "import matplotlib.pyplot as plt\n",
    "import json\n",
    "import gc\n",
    "import time\n",
    "from tqdm import tqdm\n",
    "import random\n",
    "from collections import defaultdict\n",
    "from IPython.display import display\n",
    "\n",
    "#Pytorch Imports\n",
    "import torch\n",
    "import torchvision # torch package for vision related things\n",
    "import torch.nn.functional as F  # Parameterless functions, like (some) activation functions\n",
    "import torchvision.datasets as datasets  # Standard datasets\n",
    "import torchvision.transforms as transforms  # Transformations we can perform on our dataset for augmentation\n",
    "from torch import optim  # For optimizers like SGD, Adam, etc.\n",
    "from torch import nn  # All neural network modules\n",
    "from torch.utils.data import Dataset, DataLoader  # Gives easier dataset managment by creating mini batches etc.\n",
    "from tqdm import tqdm  # For nice progress bar!\n",
    "from torchvision.transforms import Resize\n",
    "from torchaudio.transforms import MelSpectrogram, AmplitudeToDB\n",
    "from torch.optim import lr_scheduler\n",
    "from tifffile import imread\n",
    "\n",
    "from albumentations.pytorch.transforms import ToTensorV2\n",
    "import albumentations as A\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# For colored terminal text\n",
    "from colorama import Fore, Back, Style\n",
    "c_ = Fore.CYAN\n",
    "sr_ = Style.RESET_ALL\n",
    "b_ = Fore.BLUE\n",
    "\n",
    "# For descriptive error messages\n",
    "os.environ['CUDA_LAUNCH_BLOCKING'] = \"1\"\n",
    "\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "# pd.set_option('display.max_colwidth', 100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "8c4e2555",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:16.252670Z",
     "iopub.status.busy": "2022-07-24T09:02:16.252207Z",
     "iopub.status.idle": "2022-07-24T09:02:16.319412Z",
     "shell.execute_reply": "2022-07-24T09:02:16.318497Z"
    },
    "papermill": {
     "duration": 0.078987,
     "end_time": "2022-07-24T09:02:16.321287",
     "exception": false,
     "start_time": "2022-07-24T09:02:16.242300",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Hyperparameters etc.\n",
    "# Hyperparameters\n",
    "CONFIG = {\n",
    "    \"in_channels\" :3,\n",
    "    \"num_classes\": 1,\n",
    "    \"BATCH_SIZE\" : 8,\n",
    "    \"NUM_EPOCHS\" : 40,\n",
    "    \"n_accumulate\": 1,\n",
    "    \"competition\": \"HuBMAP-Kaggle\", # HuBMAP-Kaggle\n",
    "    \"model_name\": \"Vanilla_Unet\",\n",
    "    \"LEARNING_RATE\": 1e-4,\n",
    "    \"DEVICE\": \"cuda\" if torch.cuda.is_available() else \"cpu\", \n",
    "    \"AUG\": \"No\",\n",
    "    \"SEED\": 42,\n",
    "    \"opt\": 'Adam',\n",
    "    \"Normalization\": \"L2\",\n",
    "    \"img_size\": 256\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "ff310a68",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:16.340606Z",
     "iopub.status.busy": "2022-07-24T09:02:16.339772Z",
     "iopub.status.idle": "2022-07-24T09:02:16.352855Z",
     "shell.execute_reply": "2022-07-24T09:02:16.351624Z"
    },
    "papermill": {
     "duration": 0.02559,
     "end_time": "2022-07-24T09:02:16.355653",
     "exception": false,
     "start_time": "2022-07-24T09:02:16.330063",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> SEEDING DONE\n"
     ]
    }
   ],
   "source": [
    "def set_seed(seed = 42):\n",
    "    '''Sets the seed of the entire notebook so results are the same every time we run.\n",
    "    This is for REPRODUCIBILITY.'''\n",
    "    np.random.seed(seed)\n",
    "    random.seed(seed)\n",
    "    torch.manual_seed(seed)\n",
    "    torch.cuda.manual_seed(seed)\n",
    "    # When running on the CuDNN backend, two further options must be set\n",
    "    torch.backends.cudnn.deterministic = True\n",
    "    torch.backends.cudnn.benchmark = False\n",
    "    # Set a fixed value for the hash seed\n",
    "    os.environ['PYTHONHASHSEED'] = str(seed)\n",
    "    print('> SEEDING DONE')\n",
    "    \n",
    "set_seed(CONFIG[\"SEED\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1f2817dd",
   "metadata": {
    "papermill": {
     "duration": 0.008728,
     "end_time": "2022-07-24T09:02:16.373521",
     "exception": false,
     "start_time": "2022-07-24T09:02:16.364793",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# WandB:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "b9a35021",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:16.393456Z",
     "iopub.status.busy": "2022-07-24T09:02:16.393189Z",
     "iopub.status.idle": "2022-07-24T09:02:23.115190Z",
     "shell.execute_reply": "2022-07-24T09:02:23.114117Z"
    },
    "papermill": {
     "duration": 6.735666,
     "end_time": "2022-07-24T09:02:23.117801",
     "exception": false,
     "start_time": "2022-07-24T09:02:16.382135",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Failed to detect the name of this notebook, you can set it manually with the WANDB_NOTEBOOK_NAME environment variable to enable code saving.\n",
      "\u001b[34m\u001b[1mwandb\u001b[0m: Currently logged in as: \u001b[33msomusan\u001b[0m. Use \u001b[1m`wandb login --relogin`\u001b[0m to force relogin\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "Tracking run with wandb version 0.13.0"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "Run data is saved locally in <code>/home/lakshita/somusan/hubmap_kaggle/wandb/run-20220805_113053-l4spdzkj</code>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "Syncing run <strong><a href=\"https://wandb.ai/somusan/HuBMAP-Kaggle/runs/l4spdzkj\" target=\"_blank\">Vanilla_Unet_1</a></strong> to <a href=\"https://wandb.ai/somusan/HuBMAP-Kaggle\" target=\"_blank\">Weights & Biases</a> (<a href=\"https://wandb.me/run\" target=\"_blank\">docs</a>)<br/>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import wandb\n",
    "\n",
    "wandb.login()\n",
    "\n",
    "run = wandb.init(project=CONFIG['competition'], \n",
    "                 config=CONFIG,\n",
    "                 job_type='Train',\n",
    "                 tags=['semantic segmentation', CONFIG['model_name']],\n",
    "                 anonymous='must',\n",
    "                 name = \"Vanilla_Unet_1\",\n",
    "                 notes = \"\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "27ecdb30",
   "metadata": {
    "papermill": {
     "duration": 0.009224,
     "end_time": "2022-07-24T09:02:23.139750",
     "exception": false,
     "start_time": "2022-07-24T09:02:23.130526",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Tabular:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "96bf7d0c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:23.160040Z",
     "iopub.status.busy": "2022-07-24T09:02:23.159693Z",
     "iopub.status.idle": "2022-07-24T09:02:24.356498Z",
     "shell.execute_reply": "2022-07-24T09:02:24.355431Z"
    },
    "papermill": {
     "duration": 1.209941,
     "end_time": "2022-07-24T09:02:24.359280",
     "exception": false,
     "start_time": "2022-07-24T09:02:23.149339",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>organ</th>\n",
       "      <th>data_source</th>\n",
       "      <th>img_height</th>\n",
       "      <th>img_width</th>\n",
       "      <th>pixel_size</th>\n",
       "      <th>tissue_thickness</th>\n",
       "      <th>rle</th>\n",
       "      <th>age</th>\n",
       "      <th>sex</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10044</td>\n",
       "      <td>prostate</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>1459676 77 1462675 82 1465674 87 1468673 92 1471671 97 1474669 101 1477667 105 1480665 109 14836...</td>\n",
       "      <td>37.0</td>\n",
       "      <td>Male</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10274</td>\n",
       "      <td>prostate</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>715707 2 718705 8 721703 11 724701 18 727692 34 730692 35 733679 54 736677 58 739676 65 742675 6...</td>\n",
       "      <td>76.0</td>\n",
       "      <td>Male</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10392</td>\n",
       "      <td>spleen</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>1228631 20 1231629 24 1234624 40 1237623 47 1240619 61 1243618 63 1246618 63 1249610 82 1252609 ...</td>\n",
       "      <td>82.0</td>\n",
       "      <td>Male</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10488</td>\n",
       "      <td>lung</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>3446519 15 3449517 17 3452514 20 3455510 24 3458491 43 3461488 46 3464486 48 3467485 49 3470484 ...</td>\n",
       "      <td>78.0</td>\n",
       "      <td>Male</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10610</td>\n",
       "      <td>spleen</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>478925 68 481909 87 484893 105 487863 154 490854 169 493845 182 496836 194 499827 205 502820 216...</td>\n",
       "      <td>21.0</td>\n",
       "      <td>Female</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id     organ data_source  img_height  img_width  pixel_size  \\\n",
       "0  10044  prostate         HPA        3000       3000         0.4   \n",
       "1  10274  prostate         HPA        3000       3000         0.4   \n",
       "2  10392    spleen         HPA        3000       3000         0.4   \n",
       "3  10488      lung         HPA        3000       3000         0.4   \n",
       "4  10610    spleen         HPA        3000       3000         0.4   \n",
       "\n",
       "   tissue_thickness  \\\n",
       "0                 4   \n",
       "1                 4   \n",
       "2                 4   \n",
       "3                 4   \n",
       "4                 4   \n",
       "\n",
       "                                                                                                   rle  \\\n",
       "0  1459676 77 1462675 82 1465674 87 1468673 92 1471671 97 1474669 101 1477667 105 1480665 109 14836...   \n",
       "1  715707 2 718705 8 721703 11 724701 18 727692 34 730692 35 733679 54 736677 58 739676 65 742675 6...   \n",
       "2  1228631 20 1231629 24 1234624 40 1237623 47 1240619 61 1243618 63 1246618 63 1249610 82 1252609 ...   \n",
       "3  3446519 15 3449517 17 3452514 20 3455510 24 3458491 43 3461488 46 3464486 48 3467485 49 3470484 ...   \n",
       "4  478925 68 481909 87 484893 105 487863 154 490854 169 493845 182 496836 194 499827 205 502820 216...   \n",
       "\n",
       "    age     sex  \n",
       "0  37.0    Male  \n",
       "1  76.0    Male  \n",
       "2  82.0    Male  \n",
       "3  78.0    Male  \n",
       "4  21.0  Female  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DIR = \"/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/\"\n",
    "train_df = pd.read_csv(os.path.join(DIR,\"train.csv\"))\n",
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "186c07d9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:24.383387Z",
     "iopub.status.busy": "2022-07-24T09:02:24.383108Z",
     "iopub.status.idle": "2022-07-24T09:02:25.252918Z",
     "shell.execute_reply": "2022-07-24T09:02:25.252001Z"
    },
    "papermill": {
     "duration": 0.886912,
     "end_time": "2022-07-24T09:02:25.256086",
     "exception": false,
     "start_time": "2022-07-24T09:02:24.369174",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>organ</th>\n",
       "      <th>data_source</th>\n",
       "      <th>img_height</th>\n",
       "      <th>img_width</th>\n",
       "      <th>pixel_size</th>\n",
       "      <th>tissue_thickness</th>\n",
       "      <th>rle</th>\n",
       "      <th>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>img_path</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10044</td>\n",
       "      <td>prostate</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>1459676 77 1462675 82 1465674 87 1468673 92 1471671 97 1474669 101 1477667 105 1480665 109 14836...</td>\n",
       "      <td>37.0</td>\n",
       "      <td>Male</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10044.tiff</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10274</td>\n",
       "      <td>prostate</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>715707 2 718705 8 721703 11 724701 18 727692 34 730692 35 733679 54 736677 58 739676 65 742675 6...</td>\n",
       "      <td>76.0</td>\n",
       "      <td>Male</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10274.tiff</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10392</td>\n",
       "      <td>spleen</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>1228631 20 1231629 24 1234624 40 1237623 47 1240619 61 1243618 63 1246618 63 1249610 82 1252609 ...</td>\n",
       "      <td>82.0</td>\n",
       "      <td>Male</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10392.tiff</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10488</td>\n",
       "      <td>lung</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>3446519 15 3449517 17 3452514 20 3455510 24 3458491 43 3461488 46 3464486 48 3467485 49 3470484 ...</td>\n",
       "      <td>78.0</td>\n",
       "      <td>Male</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10488.tiff</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10610</td>\n",
       "      <td>spleen</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>478925 68 481909 87 484893 105 487863 154 490854 169 493845 182 496836 194 499827 205 502820 216...</td>\n",
       "      <td>21.0</td>\n",
       "      <td>Female</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10610.tiff</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id     organ data_source  img_height  img_width  pixel_size  \\\n",
       "0  10044  prostate         HPA        3000       3000         0.4   \n",
       "1  10274  prostate         HPA        3000       3000         0.4   \n",
       "2  10392    spleen         HPA        3000       3000         0.4   \n",
       "3  10488      lung         HPA        3000       3000         0.4   \n",
       "4  10610    spleen         HPA        3000       3000         0.4   \n",
       "\n",
       "   tissue_thickness  \\\n",
       "0                 4   \n",
       "1                 4   \n",
       "2                 4   \n",
       "3                 4   \n",
       "4                 4   \n",
       "\n",
       "                                                                                                   rle  \\\n",
       "0  1459676 77 1462675 82 1465674 87 1468673 92 1471671 97 1474669 101 1477667 105 1480665 109 14836...   \n",
       "1  715707 2 718705 8 721703 11 724701 18 727692 34 730692 35 733679 54 736677 58 739676 65 742675 6...   \n",
       "2  1228631 20 1231629 24 1234624 40 1237623 47 1240619 61 1243618 63 1246618 63 1249610 82 1252609 ...   \n",
       "3  3446519 15 3449517 17 3452514 20 3455510 24 3458491 43 3461488 46 3464486 48 3467485 49 3470484 ...   \n",
       "4  478925 68 481909 87 484893 105 487863 154 490854 169 493845 182 496836 194 499827 205 502820 216...   \n",
       "\n",
       "    age     sex  \\\n",
       "0  37.0    Male   \n",
       "1  76.0    Male   \n",
       "2  82.0    Male   \n",
       "3  78.0    Male   \n",
       "4  21.0  Female   \n",
       "\n",
       "                                                                                             img_path  \n",
       "0  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10044.tiff  \n",
       "1  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10274.tiff  \n",
       "2  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10392.tiff  \n",
       "3  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10488.tiff  \n",
       "4  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10610.tiff  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "TRAIN_DIR = DIR + \"train_images/\"\n",
    "func = lambda x: TRAIN_DIR + str(x) + \".tiff\"\n",
    "train_df[\"img_path\"] = train_df[\"id\"].apply(func)\n",
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "27681839",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:25.278445Z",
     "iopub.status.busy": "2022-07-24T09:02:25.278143Z",
     "iopub.status.idle": "2022-07-24T09:02:26.287199Z",
     "shell.execute_reply": "2022-07-24T09:02:26.286214Z"
    },
    "papermill": {
     "duration": 1.022336,
     "end_time": "2022-07-24T09:02:26.289557",
     "exception": false,
     "start_time": "2022-07-24T09:02:25.267221",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10892.npy\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>organ</th>\n",
       "      <th>data_source</th>\n",
       "      <th>img_height</th>\n",
       "      <th>img_width</th>\n",
       "      <th>pixel_size</th>\n",
       "      <th>tissue_thickness</th>\n",
       "      <th>rle</th>\n",
       "      <th>age</th>\n",
       "      <th>sex</th>\n",
       "      <th>img_path</th>\n",
       "      <th>mask_path</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10044</td>\n",
       "      <td>prostate</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>1459676 77 1462675 82 1465674 87 1468673 92 1471671 97 1474669 101 1477667 105 1480665 109 14836...</td>\n",
       "      <td>37.0</td>\n",
       "      <td>Male</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10044.tiff</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10044.npy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10274</td>\n",
       "      <td>prostate</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>715707 2 718705 8 721703 11 724701 18 727692 34 730692 35 733679 54 736677 58 739676 65 742675 6...</td>\n",
       "      <td>76.0</td>\n",
       "      <td>Male</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10274.tiff</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10274.npy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>10392</td>\n",
       "      <td>spleen</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>1228631 20 1231629 24 1234624 40 1237623 47 1240619 61 1243618 63 1246618 63 1249610 82 1252609 ...</td>\n",
       "      <td>82.0</td>\n",
       "      <td>Male</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10392.tiff</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10392.npy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10488</td>\n",
       "      <td>lung</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>3446519 15 3449517 17 3452514 20 3455510 24 3458491 43 3461488 46 3464486 48 3467485 49 3470484 ...</td>\n",
       "      <td>78.0</td>\n",
       "      <td>Male</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10488.tiff</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10488.npy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10610</td>\n",
       "      <td>spleen</td>\n",
       "      <td>HPA</td>\n",
       "      <td>3000</td>\n",
       "      <td>3000</td>\n",
       "      <td>0.4</td>\n",
       "      <td>4</td>\n",
       "      <td>478925 68 481909 87 484893 105 487863 154 490854 169 493845 182 496836 194 499827 205 502820 216...</td>\n",
       "      <td>21.0</td>\n",
       "      <td>Female</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10610.tiff</td>\n",
       "      <td>/home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10610.npy</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id     organ data_source  img_height  img_width  pixel_size  \\\n",
       "0  10044  prostate         HPA        3000       3000         0.4   \n",
       "1  10274  prostate         HPA        3000       3000         0.4   \n",
       "2  10392    spleen         HPA        3000       3000         0.4   \n",
       "3  10488      lung         HPA        3000       3000         0.4   \n",
       "4  10610    spleen         HPA        3000       3000         0.4   \n",
       "\n",
       "   tissue_thickness  \\\n",
       "0                 4   \n",
       "1                 4   \n",
       "2                 4   \n",
       "3                 4   \n",
       "4                 4   \n",
       "\n",
       "                                                                                                   rle  \\\n",
       "0  1459676 77 1462675 82 1465674 87 1468673 92 1471671 97 1474669 101 1477667 105 1480665 109 14836...   \n",
       "1  715707 2 718705 8 721703 11 724701 18 727692 34 730692 35 733679 54 736677 58 739676 65 742675 6...   \n",
       "2  1228631 20 1231629 24 1234624 40 1237623 47 1240619 61 1243618 63 1246618 63 1249610 82 1252609 ...   \n",
       "3  3446519 15 3449517 17 3452514 20 3455510 24 3458491 43 3461488 46 3464486 48 3467485 49 3470484 ...   \n",
       "4  478925 68 481909 87 484893 105 487863 154 490854 169 493845 182 496836 194 499827 205 502820 216...   \n",
       "\n",
       "    age     sex  \\\n",
       "0  37.0    Male   \n",
       "1  76.0    Male   \n",
       "2  82.0    Male   \n",
       "3  78.0    Male   \n",
       "4  21.0  Female   \n",
       "\n",
       "                                                                                             img_path  \\\n",
       "0  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10044.tiff   \n",
       "1  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10274.tiff   \n",
       "2  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10392.tiff   \n",
       "3  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10488.tiff   \n",
       "4  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/hubmap-organ-segmentation/train_images/10610.tiff   \n",
       "\n",
       "                                                                   mask_path  \n",
       "0  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10044.npy  \n",
       "1  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10274.npy  \n",
       "2  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10392.npy  \n",
       "3  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10488.npy  \n",
       "4  /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10610.npy  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "MASK_DIR = \"/home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/\"\n",
    "func1 = lambda x: MASK_DIR + str(x) + \".npy\"\n",
    "train_df[\"mask_path\"] = train_df[\"id\"].apply(func1)\n",
    "print(train_df[\"mask_path\"].iloc[9])\n",
    "train_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "869a3e97",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10044.npy\n",
       "1      /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10274.npy\n",
       "2      /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10392.npy\n",
       "3      /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10488.npy\n",
       "4      /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/10610.npy\n",
       "                                         ...                                    \n",
       "346     /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/9517.npy\n",
       "347     /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/9769.npy\n",
       "348     /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/9777.npy\n",
       "349     /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/9791.npy\n",
       "350     /home/lakshita/somusan/hubmap_kaggle/hubmap_data/train_masks_np/9904.npy\n",
       "Name: mask_path, Length: 351, dtype: object"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df[\"mask_path\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "2d27bd28",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:26.312343Z",
     "iopub.status.busy": "2022-07-24T09:02:26.311794Z",
     "iopub.status.idle": "2022-07-24T09:02:27.173091Z",
     "shell.execute_reply": "2022-07-24T09:02:27.172161Z"
    },
    "papermill": {
     "duration": 0.875129,
     "end_time": "2022-07-24T09:02:27.175524",
     "exception": false,
     "start_time": "2022-07-24T09:02:26.300395",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(224, 56)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df, test_df = train_test_split(train_df, test_size=0.2, random_state=42)\n",
    "\n",
    "train_df.__len__(), len(test_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3c54ede7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:27.198363Z",
     "iopub.status.busy": "2022-07-24T09:02:27.197404Z",
     "iopub.status.idle": "2022-07-24T09:02:28.411504Z",
     "shell.execute_reply": "2022-07-24T09:02:28.410587Z"
    },
    "papermill": {
     "duration": 1.227746,
     "end_time": "2022-07-24T09:02:28.413817",
     "exception": false,
     "start_time": "2022-07-24T09:02:27.186071",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4.57 ms ± 612 µs per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
     ]
    }
   ],
   "source": [
    "%timeit img_temp = imread(train_df[\"img_path\"].iloc[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8ccb66e2",
   "metadata": {
    "papermill": {
     "duration": 0.010166,
     "end_time": "2022-07-24T09:02:28.434655",
     "exception": false,
     "start_time": "2022-07-24T09:02:28.424489",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Pytorch Data Class:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "f4869912",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:28.460746Z",
     "iopub.status.busy": "2022-07-24T09:02:28.459820Z",
     "iopub.status.idle": "2022-07-24T09:02:29.444762Z",
     "shell.execute_reply": "2022-07-24T09:02:29.443743Z"
    },
    "papermill": {
     "duration": 1.002536,
     "end_time": "2022-07-24T09:02:29.447637",
     "exception": false,
     "start_time": "2022-07-24T09:02:28.445101",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "mean = np.array([0.7720342, 0.74582646, 0.76392896])\n",
    "std = np.array([0.24745085, 0.26182273, 0.25782376])\n",
    "\n",
    "\n",
    "def img2tensor(img,dtype:np.dtype=np.float32):\n",
    "    if img.ndim==2 : img = np.expand_dims(img,2)\n",
    "    img = np.transpose(img,(2,0,1))\n",
    "    return torch.from_numpy(img.astype(dtype, copy=False))\n",
    "\n",
    "\n",
    "\n",
    "class hubmap_data(Dataset):\n",
    "    def __init__(self, df, transform=None):\n",
    "        self.df = df\n",
    "        self.img_paths = df[\"img_path\"].to_numpy()\n",
    "        self.mask_paths = df[\"mask_path\"].to_numpy()\n",
    "        self.transform = transform\n",
    "        \n",
    "        \n",
    "    def __getitem__(self, index):\n",
    "        # load data from the pre-processed npy files\n",
    "        img_file = self.img_paths[index]\n",
    "#         print(img_file)\n",
    "        mask_file = self.mask_paths[index]\n",
    "        img = imread(img_file)#cv2.cvtColor(cv2.imread(img_file), cv2.COLOR_BGR2RGB)\n",
    "#         img = imread(img_file) #np.moveaxis(imread(img_file), -1, 0) #torch.from_numpy(np.moveaxis(imread(img_file), -1, 0))\n",
    "        # semantic = torch.from_numpy(np.load(self.data_path + '/label/{:d}.npy'.format(index)))\n",
    "#         mask = torch.from_numpy(np.moveaxis(np.load(mask_file), -1, 0))\n",
    "        mask = np.load(mask_file)#np.moveaxis(np.load(mask_file), -1, 0) #torch.from_numpy(np.moveaxis(np.load(mask_file), -1, 0))\n",
    "        \n",
    "        \n",
    "#       print(depth.shape)\n",
    "        if self.transform is not None:\n",
    "            transformed = self.transform(image=img, mask=mask)\n",
    "            img = transformed[\"image\"]\n",
    "            mask = transformed[\"mask\"]\n",
    "\n",
    "            if mask.shape[-1] == 1:\n",
    "                mask = mask.permute(2,0,1)\n",
    "                \n",
    "                \n",
    "        return img2tensor((img/255.0 - mean)/std),img2tensor(mask)\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "86c84c6e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:29.476459Z",
     "iopub.status.busy": "2022-07-24T09:02:29.476175Z",
     "iopub.status.idle": "2022-07-24T09:02:30.391566Z",
     "shell.execute_reply": "2022-07-24T09:02:30.389926Z"
    },
    "papermill": {
     "duration": 0.931066,
     "end_time": "2022-07-24T09:02:30.395278",
     "exception": false,
     "start_time": "2022-07-24T09:02:29.464212",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "train_transform = A.Compose(\n",
    "    [\n",
    "        A.Resize(256, 256, p=1),\n",
    "#         A.HorizontalFlip(p=1.0),\n",
    "#         A.RandomCrop(224, 224, p=0.3),\n",
    "#         A.RandomSizedCrop(cfg[\"min_max_height\"], cfg[\"height\"], cfg[\"width\"], cfg[\"w2h_ratio\"], cfg[\"interpolation\"], p=1), \n",
    "#         ToTensorV2(),\n",
    "    ]\n",
    ")\n",
    "\n",
    "test_transform = A.Compose(\n",
    "    [\n",
    "        A.Resize(256, 256, p=1.0),\n",
    "#         ToTensorV2(),\n",
    "    ]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "708bdcdf",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:30.418838Z",
     "iopub.status.busy": "2022-07-24T09:02:30.418504Z",
     "iopub.status.idle": "2022-07-24T09:02:31.868643Z",
     "shell.execute_reply": "2022-07-24T09:02:31.867660Z"
    },
    "papermill": {
     "duration": 1.464105,
     "end_time": "2022-07-24T09:02:31.871101",
     "exception": false,
     "start_time": "2022-07-24T09:02:30.406996",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([3, 256, 256]) torch.Size([1, 256, 256])\n"
     ]
    }
   ],
   "source": [
    "train_data = hubmap_data(train_df,train_transform)\n",
    "test_data = hubmap_data(test_df, test_transform)\n",
    "\n",
    "print(train_data[10][0].shape, train_data[10][1].shape)\n",
    "# test_data[10][0].shape, test_data[10][1].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "8648d374",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:31.898832Z",
     "iopub.status.busy": "2022-07-24T09:02:31.898517Z",
     "iopub.status.idle": "2022-07-24T09:02:32.992281Z",
     "shell.execute_reply": "2022-07-24T09:02:32.991257Z"
    },
    "papermill": {
     "duration": 1.109459,
     "end_time": "2022-07-24T09:02:32.994644",
     "exception": false,
     "start_time": "2022-07-24T09:02:31.885185",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f68cf1abe80>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "temp_img = train_data[10][0]\n",
    "temp_mask = train_data[10][1]\n",
    "# plt.imshow(temp_mask.detach().numpy())\n",
    "# plt.imshow(temp_img.permute(1,2,0).detach().numpy())\n",
    "img_trfm = ((temp_img.permute(1,2,0)*std + mean)*255.0).numpy().astype(np.uint8)\n",
    "plt.imshow(img_trfm)\n",
    "# temp_img.permute(1,2,0).detach().numpy()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6c61d7fe",
   "metadata": {
    "papermill": {
     "duration": 0.010547,
     "end_time": "2022-07-24T09:02:33.016193",
     "exception": false,
     "start_time": "2022-07-24T09:02:33.005646",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Unet Model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "e939987b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:33.039007Z",
     "iopub.status.busy": "2022-07-24T09:02:33.038374Z",
     "iopub.status.idle": "2022-07-24T09:02:33.959169Z",
     "shell.execute_reply": "2022-07-24T09:02:33.958130Z"
    },
    "papermill": {
     "duration": 0.934725,
     "end_time": "2022-07-24T09:02:33.961451",
     "exception": false,
     "start_time": "2022-07-24T09:02:33.026726",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def init_model():\n",
    "    model =  smp.Unet(\n",
    "                 encoder_name='efficientnet-b0',\n",
    "                 encoder_weights='imagenet',\n",
    "                 in_channels=3,\n",
    "                 classes=1)\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca60848f",
   "metadata": {
    "papermill": {
     "duration": 0.011053,
     "end_time": "2022-07-24T09:02:33.985064",
     "exception": false,
     "start_time": "2022-07-24T09:02:33.974011",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Model check:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "dc10d00a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:34.007925Z",
     "iopub.status.busy": "2022-07-24T09:02:34.007110Z",
     "iopub.status.idle": "2022-07-24T09:02:43.192846Z",
     "shell.execute_reply": "2022-07-24T09:02:43.191674Z"
    },
    "papermill": {
     "duration": 9.199981,
     "end_time": "2022-07-24T09:02:43.195565",
     "exception": false,
     "start_time": "2022-07-24T09:02:33.995584",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([1, 1, 256, 256])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_on_gpu = torch.cuda.is_available()\n",
    "model = init_model() #UNet(n_channels=CONFIG[\"in_channels\"], n_classes=CONFIG[\"num_classes\"])\n",
    "if train_on_gpu:\n",
    "    model.cuda()\n",
    "\n",
    "model(temp_img.unsqueeze(0).float().cuda()).shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31c9ca21",
   "metadata": {
    "papermill": {
     "duration": 0.011411,
     "end_time": "2022-07-24T09:02:43.223641",
     "exception": false,
     "start_time": "2022-07-24T09:02:43.212230",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# DataLoader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "18912a76",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:43.247444Z",
     "iopub.status.busy": "2022-07-24T09:02:43.247149Z",
     "iopub.status.idle": "2022-07-24T09:02:44.146944Z",
     "shell.execute_reply": "2022-07-24T09:02:44.145988Z"
    },
    "papermill": {
     "duration": 0.915029,
     "end_time": "2022-07-24T09:02:44.149887",
     "exception": false,
     "start_time": "2022-07-24T09:02:43.234858",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# REMARK shuffle = False [result might change with shuffle =True]\n",
    "train_loader = torch.utils.data.DataLoader(\n",
    "               dataset=train_data,\n",
    "               batch_size=CONFIG[\"BATCH_SIZE\"],\n",
    "               shuffle=False,\n",
    "               num_workers=2)\n",
    "\n",
    "test_loader = torch.utils.data.DataLoader(\n",
    "              dataset=test_data,\n",
    "              batch_size=CONFIG[\"BATCH_SIZE\"],\n",
    "              shuffle=False,\n",
    "              num_workers=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ecdc3aa1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# ds = HuBMAPDataset(tfms=albu_aug())\n",
    "# dl = torch.utils.data.DataLoader(ds,batch_size=64,shuffle=False,num_workers=NUM_WORKERS)\n",
    "it = iter(train_loader)\n",
    "imgs,masks = next(it)\n",
    "print(masks.dtype)\n",
    "plt.figure(figsize=(16,16))\n",
    "for i,(img,mask) in enumerate(zip(imgs,masks)):\n",
    "    img = ((img.permute(1,2,0))).numpy().astype(np.uint8)  # H , W , C\n",
    "    plt.subplot(8,8,i+1)\n",
    "    plt.imshow(img,vmin=0,vmax=255)\n",
    "    plt.imshow(mask.squeeze().numpy(), alpha=0.2)\n",
    "    plt.axis('off')\n",
    "    plt.subplots_adjust(wspace=None, hspace=None)\n",
    "    \n",
    "# del ds,dl,imgs,masks"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "589e8ecd",
   "metadata": {
    "papermill": {
     "duration": 0.011501,
     "end_time": "2022-07-24T09:02:44.173568",
     "exception": false,
     "start_time": "2022-07-24T09:02:44.162067",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Loss, Optimizer & Metric:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "b81023b9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:44.200841Z",
     "iopub.status.busy": "2022-07-24T09:02:44.200076Z",
     "iopub.status.idle": "2022-07-24T09:02:45.106576Z",
     "shell.execute_reply": "2022-07-24T09:02:45.105569Z"
    },
    "papermill": {
     "duration": 0.924541,
     "end_time": "2022-07-24T09:02:45.109416",
     "exception": false,
     "start_time": "2022-07-24T09:02:44.184875",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "#PyTorch\n",
    "class DiceLoss(nn.Module):\n",
    "    def __init__(self, weight=None, size_average=True):\n",
    "        super(DiceLoss, self).__init__()\n",
    "\n",
    "    def forward(self, inputs, targets, smooth=1):\n",
    "        \n",
    "        #comment out if your model contains a sigmoid or equivalent activation layer\n",
    "        inputs = F.sigmoid(inputs)       \n",
    "        \n",
    "        #flatten label and prediction tensors\n",
    "        inputs = inputs.view(-1)\n",
    "        targets = targets.view(-1)\n",
    "        \n",
    "        intersection = (inputs * targets).sum()                            \n",
    "        dice = (2.*intersection + smooth)/(inputs.sum() + targets.sum() + smooth)  \n",
    "        \n",
    "        return 1 - dice\n",
    "\n",
    "loss_fn = DiceLoss() #Dice()\n",
    "optimizer = optim.Adam(model.parameters(), lr=CONFIG[\"LEARNING_RATE\"],weight_decay=1e-5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "259eaca9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:45.137917Z",
     "iopub.status.busy": "2022-07-24T09:02:45.137022Z",
     "iopub.status.idle": "2022-07-24T09:02:46.008302Z",
     "shell.execute_reply": "2022-07-24T09:02:46.007427Z"
    },
    "papermill": {
     "duration": 0.886676,
     "end_time": "2022-07-24T09:02:46.011000",
     "exception": false,
     "start_time": "2022-07-24T09:02:45.124324",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def dice_coef(y_true, y_pred, thr=0.5, dim=(2,3), epsilon=0.001):\n",
    "    y_true = y_true.to(torch.float32)\n",
    "    y_pred = (y_pred>thr).to(torch.float32)\n",
    "    inter = (y_true*y_pred).sum(dim=dim)\n",
    "    den = y_true.sum(dim=dim) + y_pred.sum(dim=dim)\n",
    "    dice = ((2*inter+epsilon)/(den+epsilon)).mean(dim=(1,0))\n",
    "    return dice\n",
    "\n",
    "def iou_coef(y_true, y_pred, thr=0.5, dim=(2,3), epsilon=0.001):\n",
    "    y_true = y_true.to(torch.float32)\n",
    "    y_pred = (y_pred>thr).to(torch.float32)\n",
    "    inter = (y_true*y_pred).sum(dim=dim)\n",
    "    union = (y_true + y_pred - y_true*y_pred).sum(dim=dim)\n",
    "    iou = ((inter+epsilon)/(union+epsilon)).mean(dim=(1,0))\n",
    "    return iou"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c852e837",
   "metadata": {
    "papermill": {
     "duration": 0.010507,
     "end_time": "2022-07-24T09:02:46.032352",
     "exception": false,
     "start_time": "2022-07-24T09:02:46.021845",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# model save class:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "2ce29431",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:46.061366Z",
     "iopub.status.busy": "2022-07-24T09:02:46.060360Z",
     "iopub.status.idle": "2022-07-24T09:02:47.064875Z",
     "shell.execute_reply": "2022-07-24T09:02:47.063982Z"
    },
    "papermill": {
     "duration": 1.02092,
     "end_time": "2022-07-24T09:02:47.067781",
     "exception": false,
     "start_time": "2022-07-24T09:02:46.046861",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "model_path = \"./model_weights\"\n",
    "if not os.path.exists(model_path):\n",
    "    os.makedirs(model_path)\n",
    "    \n",
    "def save_model(model, optimizer, criterion, epoch):\n",
    "    \"\"\"\n",
    "    Function to save the trained model to disk.\n",
    "    \"\"\"\n",
    "    print(f\"\\n Saving model at {epoch}th epoch\")\n",
    "    fname = f'{model_path}/{CONFIG[\"model_name\"]}-{CONFIG[\"LEARNING_RATE\"]}-{CONFIG[\"AUG\"]}-{epoch}-{CONFIG[\"BATCH_SIZE\"]}.pth'\n",
    "    torch.save({\n",
    "                'epoch': epoch,\n",
    "                'model_state_dict': model.state_dict(),\n",
    "                'optimizer_state_dict': optimizer.state_dict(),\n",
    "                'loss': criterion,\n",
    "                }, fname )\n",
    "    return fname"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a0713166",
   "metadata": {
    "papermill": {
     "duration": 0.023173,
     "end_time": "2022-07-24T09:02:47.116268",
     "exception": false,
     "start_time": "2022-07-24T09:02:47.093095",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Train & Validation Function:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "9d7a4e21",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:47.146265Z",
     "iopub.status.busy": "2022-07-24T09:02:47.145922Z",
     "iopub.status.idle": "2022-07-24T09:02:47.995031Z",
     "shell.execute_reply": "2022-07-24T09:02:47.994096Z"
    },
    "papermill": {
     "duration": 0.863072,
     "end_time": "2022-07-24T09:02:47.997338",
     "exception": false,
     "start_time": "2022-07-24T09:02:47.134266",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def train_fn(loader, model, optimizer, loss_fn,epoch):\n",
    "    bar = tqdm(enumerate(loader), total=len(loader))\n",
    "    \n",
    "    \n",
    "    running_loss = 0.0\n",
    "    dataset_size = 0\n",
    "    \n",
    "    for batch_idx, (data, targets) in bar:\n",
    "        data = data.to(device=CONFIG[\"DEVICE\"])\n",
    "        targets = targets.float().to(device=CONFIG[\"DEVICE\"])\n",
    "\n",
    "        # forward\n",
    "        predictions = model(data)\n",
    "        loss = loss_fn(predictions, targets)\n",
    "\n",
    "        \n",
    "        optimizer.zero_grad()\n",
    "        # backward\n",
    "        loss.backward()\n",
    "        # Update Weights\n",
    "        optimizer.step()\n",
    "        \n",
    "        # Calculate Loss\n",
    "        running_loss += loss.item()\n",
    "        \n",
    "        epoch_loss = running_loss / len(loader)\n",
    "#         running_loss += (loss.item() * BATCH_SIZE)\n",
    "#         dataset_size += BATCH_SIZE\n",
    "        \n",
    "#         epoch_loss = running_loss / dataset_size\n",
    "        # update tqdm loop\n",
    "        bar.set_postfix(Epoch=epoch,loss=epoch_loss)\n",
    "    \n",
    "        \n",
    "    gc.collect()\n",
    "    \n",
    "    return epoch_loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "34bdf758",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:48.020209Z",
     "iopub.status.busy": "2022-07-24T09:02:48.019863Z",
     "iopub.status.idle": "2022-07-24T09:02:48.998485Z",
     "shell.execute_reply": "2022-07-24T09:02:48.997571Z"
    },
    "papermill": {
     "duration": 0.992802,
     "end_time": "2022-07-24T09:02:49.001169",
     "exception": false,
     "start_time": "2022-07-24T09:02:48.008367",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "@torch.no_grad()\n",
    "def valid_one_epoch(model, dataloader, device, epoch):\n",
    "    model.eval()\n",
    "    \n",
    "    dataset_size = 0\n",
    "    running_loss = 0.0\n",
    "#     criterion = losses[CFG.loss]\n",
    "    \n",
    "    val_scores = []\n",
    "    \n",
    "    pbar = tqdm(enumerate(dataloader), total=len(dataloader))\n",
    "    for step, (images, masks) in pbar:\n",
    "        images  = images.to(CONFIG[\"DEVICE\"], dtype=torch.float)\n",
    "        masks   = masks.to(CONFIG[\"DEVICE\"], dtype=torch.float)\n",
    "        \n",
    "        \n",
    "        y_pred  = model(images)\n",
    "        loss    = loss_fn(y_pred, masks)\n",
    "        \n",
    "        running_loss += (loss.item() * CONFIG[\"BATCH_SIZE\"])\n",
    "        dataset_size += CONFIG[\"BATCH_SIZE\"]\n",
    "        \n",
    "        epoch_loss = running_loss / dataset_size\n",
    "        \n",
    "        y_pred = nn.Sigmoid()(y_pred)\n",
    "        val_dice = dice_coef(masks, y_pred).cpu().detach().numpy()\n",
    "        val_jaccard = iou_coef(masks, y_pred).cpu().detach().numpy()\n",
    "        val_scores.append([val_dice, val_jaccard])\n",
    "        \n",
    "        mem = torch.cuda.memory_reserved() / 1E9 if torch.cuda.is_available() else 0\n",
    "        current_lr = optimizer.param_groups[0]['lr']\n",
    "        pbar.set_postfix(valid_loss=f'{epoch_loss:0.4f}',\n",
    "                        lr=f'{current_lr:0.5f}',\n",
    "                        gpu_memory=f'{mem:0.2f} GB')\n",
    "    val_scores  = np.mean(val_scores, axis=0)\n",
    "    torch.cuda.empty_cache()\n",
    "    gc.collect()\n",
    "    \n",
    "    return epoch_loss, val_scores"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6d3f4ff",
   "metadata": {
    "papermill": {
     "duration": 0.010268,
     "end_time": "2022-07-24T09:02:49.022583",
     "exception": false,
     "start_time": "2022-07-24T09:02:49.012315",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Train loop:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "fe81f03b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:02:49.045176Z",
     "iopub.status.busy": "2022-07-24T09:02:49.044838Z",
     "iopub.status.idle": "2022-07-24T09:57:16.836873Z",
     "shell.execute_reply": "2022-07-24T09:57:16.835762Z"
    },
    "papermill": {
     "duration": 3267.80702,
     "end_time": "2022-07-24T09:57:16.840024",
     "exception": false,
     "start_time": "2022-07-24T09:02:49.033004",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[0m===================== Epoch: [1/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [02:50<00:00,  6.08s/it, Epoch=1, loss=0.813]\n",
      "100%|██████████| 7/7 [00:44<00:00,  6.35s/it, gpu_memory=3.50 GB, lr=0.00010, valid_loss=0.8255]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 1th epoch\n",
      "\n",
      " Final Train Loss: 0.8132  |  Final Val Loss: 0.8255\n",
      "Valid Dice: 0.1624 | Valid Jaccard: 0.0948\n",
      "\u001b[34mval dice increased: -inf ---> 0.16235117614269257\n",
      "\u001b[34mval jaccard increased: -inf ---> 0.0947938784956932\n",
      "\u001b[0m===================== Epoch: [2/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [01:12<00:00,  2.59s/it, Epoch=2, loss=0.803]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.91it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8124]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 2th epoch\n",
      "\n",
      " Final Train Loss: 0.8031  |  Final Val Loss: 0.8124\n",
      "Valid Dice: 0.1624 | Valid Jaccard: 0.0948\n",
      "\u001b[0m===================== Epoch: [3/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:06<00:00,  4.05it/s, Epoch=3, loss=0.799]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.07it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8102]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 3th epoch\n",
      "\n",
      " Final Train Loss: 0.7994  |  Final Val Loss: 0.8102\n",
      "Valid Dice: 0.2212 | Valid Jaccard: 0.1374\n",
      "\u001b[34mval dice increased: 0.16235117614269257 ---> 0.22124025225639343\n",
      "\u001b[34mval jaccard increased: 0.0947938784956932 ---> 0.13735531270503998\n",
      "\u001b[0m===================== Epoch: [4/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:06<00:00,  4.05it/s, Epoch=4, loss=0.797]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.99it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8069]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 4th epoch\n",
      "\n",
      " Final Train Loss: 0.7973  |  Final Val Loss: 0.8069\n",
      "Valid Dice: 0.2372 | Valid Jaccard: 0.1496\n",
      "\u001b[34mval dice increased: 0.22124025225639343 ---> 0.23722092807292938\n",
      "\u001b[34mval jaccard increased: 0.13735531270503998 ---> 0.14961457252502441\n",
      "\u001b[0m===================== Epoch: [5/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:06<00:00,  4.04it/s, Epoch=5, loss=0.796]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.05it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8059]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 5th epoch\n",
      "\n",
      " Final Train Loss: 0.7960  |  Final Val Loss: 0.8059\n",
      "Valid Dice: 0.2449 | Valid Jaccard: 0.1554\n",
      "\u001b[34mval dice increased: 0.23722092807292938 ---> 0.2449216991662979\n",
      "\u001b[34mval jaccard increased: 0.14961457252502441 ---> 0.1554485261440277\n",
      "\u001b[0m===================== Epoch: [6/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:06<00:00,  4.05it/s, Epoch=6, loss=0.794]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.93it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8057]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 6th epoch\n",
      "\n",
      " Final Train Loss: 0.7944  |  Final Val Loss: 0.8057\n",
      "Valid Dice: 0.2552 | Valid Jaccard: 0.1638\n",
      "\u001b[34mval dice increased: 0.2449216991662979 ---> 0.2552242875099182\n",
      "\u001b[34mval jaccard increased: 0.1554485261440277 ---> 0.1637946516275406\n",
      "\u001b[0m===================== Epoch: [7/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.95it/s, Epoch=7, loss=0.793]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.00it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8047]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 7th epoch\n",
      "\n",
      " Final Train Loss: 0.7929  |  Final Val Loss: 0.8047\n",
      "Valid Dice: 0.2506 | Valid Jaccard: 0.1601\n",
      "\u001b[0m===================== Epoch: [8/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.92it/s, Epoch=8, loss=0.792]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.15it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8027]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 8th epoch\n",
      "\n",
      " Final Train Loss: 0.7921  |  Final Val Loss: 0.8027\n",
      "Valid Dice: 0.2872 | Valid Jaccard: 0.1887\n",
      "\u001b[34mval dice increased: 0.2552242875099182 ---> 0.2871531844139099\n",
      "\u001b[34mval jaccard increased: 0.1637946516275406 ---> 0.18868796527385712\n",
      "\u001b[0m===================== Epoch: [9/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.91it/s, Epoch=9, loss=0.791]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.83it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8016]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 9th epoch\n",
      "\n",
      " Final Train Loss: 0.7912  |  Final Val Loss: 0.8016\n",
      "Valid Dice: 0.2819 | Valid Jaccard: 0.1842\n",
      "\u001b[0m===================== Epoch: [10/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:06<00:00,  4.00it/s, Epoch=10, loss=0.791]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.07it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8024]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 10th epoch\n",
      "\n",
      " Final Train Loss: 0.7913  |  Final Val Loss: 0.8024\n",
      "Valid Dice: 0.2988 | Valid Jaccard: 0.1979\n",
      "\u001b[34mval dice increased: 0.2871531844139099 ---> 0.29884931445121765\n",
      "\u001b[34mval jaccard increased: 0.18868796527385712 ---> 0.19786182045936584\n",
      "\u001b[0m===================== Epoch: [11/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.90it/s, Epoch=11, loss=0.791]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.17it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8040]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 11th epoch\n",
      "\n",
      " Final Train Loss: 0.7908  |  Final Val Loss: 0.8040\n",
      "Valid Dice: 0.3198 | Valid Jaccard: 0.2132\n",
      "\u001b[34mval dice increased: 0.29884931445121765 ---> 0.3198131024837494\n",
      "\u001b[34mval jaccard increased: 0.19786182045936584 ---> 0.21321061253547668\n",
      "\u001b[0m===================== Epoch: [12/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.91it/s, Epoch=12, loss=0.79] \n",
      "100%|██████████| 7/7 [00:00<00:00,  7.02it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8062]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 12th epoch\n",
      "\n",
      " Final Train Loss: 0.7900  |  Final Val Loss: 0.8062\n",
      "Valid Dice: 0.3257 | Valid Jaccard: 0.2187\n",
      "\u001b[34mval dice increased: 0.3198131024837494 ---> 0.32569172978401184\n",
      "\u001b[34mval jaccard increased: 0.21321061253547668 ---> 0.21871468424797058\n",
      "\u001b[0m===================== Epoch: [13/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.94it/s, Epoch=13, loss=0.795]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.15it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8021]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 13th epoch\n",
      "\n",
      " Final Train Loss: 0.7946  |  Final Val Loss: 0.8021\n",
      "Valid Dice: 0.2848 | Valid Jaccard: 0.1873\n",
      "\u001b[0m===================== Epoch: [14/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.93it/s, Epoch=14, loss=0.79] \n",
      "100%|██████████| 7/7 [00:01<00:00,  6.12it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8016]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 14th epoch\n",
      "\n",
      " Final Train Loss: 0.7898  |  Final Val Loss: 0.8016\n",
      "Valid Dice: 0.3283 | Valid Jaccard: 0.2206\n",
      "\u001b[34mval dice increased: 0.32569172978401184 ---> 0.32833418250083923\n",
      "\u001b[34mval jaccard increased: 0.21871468424797058 ---> 0.2206171602010727\n",
      "\u001b[0m===================== Epoch: [15/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.90it/s, Epoch=15, loss=0.789]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.91it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8028]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 15th epoch\n",
      "\n",
      " Final Train Loss: 0.7891  |  Final Val Loss: 0.8028\n",
      "Valid Dice: 0.3360 | Valid Jaccard: 0.2268\n",
      "\u001b[34mval dice increased: 0.32833418250083923 ---> 0.33599191904067993\n",
      "\u001b[34mval jaccard increased: 0.2206171602010727 ---> 0.2267662137746811\n",
      "\u001b[0m===================== Epoch: [16/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.70it/s, Epoch=16, loss=0.789]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.78it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7996]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 16th epoch\n",
      "\n",
      " Final Train Loss: 0.7888  |  Final Val Loss: 0.7996\n",
      "Valid Dice: 0.3288 | Valid Jaccard: 0.2203\n",
      "\u001b[0m===================== Epoch: [17/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  4.00it/s, Epoch=17, loss=0.787]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.08it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8022]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 17th epoch\n",
      "\n",
      " Final Train Loss: 0.7871  |  Final Val Loss: 0.8022\n",
      "Valid Dice: 0.3392 | Valid Jaccard: 0.2283\n",
      "\u001b[34mval dice increased: 0.33599191904067993 ---> 0.33915552496910095\n",
      "\u001b[34mval jaccard increased: 0.2267662137746811 ---> 0.2283368855714798\n",
      "\u001b[0m===================== Epoch: [18/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.97it/s, Epoch=18, loss=0.788]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.04it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7994]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 18th epoch\n",
      "\n",
      " Final Train Loss: 0.7879  |  Final Val Loss: 0.7994\n",
      "Valid Dice: 0.3587 | Valid Jaccard: 0.2439\n",
      "\u001b[34mval dice increased: 0.33915552496910095 ---> 0.3587295114994049\n",
      "\u001b[34mval jaccard increased: 0.2283368855714798 ---> 0.24394503235816956\n",
      "\u001b[0m===================== Epoch: [19/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.97it/s, Epoch=19, loss=0.788]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.28it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7996]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 19th epoch\n",
      "\n",
      " Final Train Loss: 0.7878  |  Final Val Loss: 0.7996\n",
      "Valid Dice: 0.3474 | Valid Jaccard: 0.2357\n",
      "\u001b[0m===================== Epoch: [20/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.96it/s, Epoch=20, loss=0.787]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.81it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7999]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 20th epoch\n",
      "\n",
      " Final Train Loss: 0.7866  |  Final Val Loss: 0.7999\n",
      "Valid Dice: 0.3524 | Valid Jaccard: 0.2390\n",
      "\u001b[0m===================== Epoch: [21/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.97it/s, Epoch=21, loss=0.785]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.14it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7997]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 21th epoch\n",
      "\n",
      " Final Train Loss: 0.7855  |  Final Val Loss: 0.7997\n",
      "Valid Dice: 0.3608 | Valid Jaccard: 0.2443\n",
      "\u001b[34mval dice increased: 0.3587295114994049 ---> 0.36084941029548645\n",
      "\u001b[34mval jaccard increased: 0.24394503235816956 ---> 0.24428865313529968\n",
      "\u001b[0m===================== Epoch: [22/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.88it/s, Epoch=22, loss=0.786]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.01it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.8012]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 22th epoch\n",
      "\n",
      " Final Train Loss: 0.7859  |  Final Val Loss: 0.8012\n",
      "Valid Dice: 0.3646 | Valid Jaccard: 0.2493\n",
      "\u001b[34mval dice increased: 0.36084941029548645 ---> 0.3645763397216797\n",
      "\u001b[34mval jaccard increased: 0.24428865313529968 ---> 0.2492571920156479\n",
      "\u001b[0m===================== Epoch: [23/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.92it/s, Epoch=23, loss=0.786]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.99it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7982]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 23th epoch\n",
      "\n",
      " Final Train Loss: 0.7857  |  Final Val Loss: 0.7982\n",
      "Valid Dice: 0.3512 | Valid Jaccard: 0.2377\n",
      "\u001b[0m===================== Epoch: [24/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.90it/s, Epoch=24, loss=0.786]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.13it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7985]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 24th epoch\n",
      "\n",
      " Final Train Loss: 0.7863  |  Final Val Loss: 0.7985\n",
      "Valid Dice: 0.3607 | Valid Jaccard: 0.2448\n",
      "\u001b[0m===================== Epoch: [25/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.98it/s, Epoch=25, loss=0.786]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.83it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7982]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 25th epoch\n",
      "\n",
      " Final Train Loss: 0.7855  |  Final Val Loss: 0.7982\n",
      "Valid Dice: 0.3591 | Valid Jaccard: 0.2439\n",
      "\u001b[0m===================== Epoch: [26/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.98it/s, Epoch=26, loss=0.785]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.05it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7975]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 26th epoch\n",
      "\n",
      " Final Train Loss: 0.7849  |  Final Val Loss: 0.7975\n",
      "Valid Dice: 0.3465 | Valid Jaccard: 0.2329\n",
      "\u001b[0m===================== Epoch: [27/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.97it/s, Epoch=27, loss=0.784]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.95it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7971]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 27th epoch\n",
      "\n",
      " Final Train Loss: 0.7835  |  Final Val Loss: 0.7971\n",
      "Valid Dice: 0.3661 | Valid Jaccard: 0.2487\n",
      "\u001b[34mval dice increased: 0.3645763397216797 ---> 0.36611494421958923\n",
      "\u001b[0m===================== Epoch: [28/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.97it/s, Epoch=28, loss=0.783]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.89it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7971]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 28th epoch\n",
      "\n",
      " Final Train Loss: 0.7831  |  Final Val Loss: 0.7971\n",
      "Valid Dice: 0.3647 | Valid Jaccard: 0.2479\n",
      "\u001b[0m===================== Epoch: [29/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.98it/s, Epoch=29, loss=0.783]\n",
      "100%|██████████| 7/7 [00:01<00:00,  6.59it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7989]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 29th epoch\n",
      "\n",
      " Final Train Loss: 0.7832  |  Final Val Loss: 0.7989\n",
      "Valid Dice: 0.3609 | Valid Jaccard: 0.2440\n",
      "\u001b[0m===================== Epoch: [30/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.96it/s, Epoch=30, loss=0.783]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.08it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7972]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 30th epoch\n",
      "\n",
      " Final Train Loss: 0.7834  |  Final Val Loss: 0.7972\n",
      "Valid Dice: 0.3920 | Valid Jaccard: 0.2720\n",
      "\u001b[34mval dice increased: 0.36611494421958923 ---> 0.3920097351074219\n",
      "\u001b[34mval jaccard increased: 0.2492571920156479 ---> 0.27202779054641724\n",
      "\u001b[0m===================== Epoch: [31/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.96it/s, Epoch=31, loss=0.782]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.22it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7972]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 31th epoch\n",
      "\n",
      " Final Train Loss: 0.7821  |  Final Val Loss: 0.7972\n",
      "Valid Dice: 0.3544 | Valid Jaccard: 0.2385\n",
      "\u001b[0m===================== Epoch: [32/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.98it/s, Epoch=32, loss=0.781]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.18it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7974]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 32th epoch\n",
      "\n",
      " Final Train Loss: 0.7810  |  Final Val Loss: 0.7974\n",
      "Valid Dice: 0.3756 | Valid Jaccard: 0.2556\n",
      "\u001b[0m===================== Epoch: [33/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.96it/s, Epoch=33, loss=0.781]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.14it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7976]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 33th epoch\n",
      "\n",
      " Final Train Loss: 0.7809  |  Final Val Loss: 0.7976\n",
      "Valid Dice: 0.3944 | Valid Jaccard: 0.2732\n",
      "\u001b[34mval dice increased: 0.3920097351074219 ---> 0.394437313079834\n",
      "\u001b[34mval jaccard increased: 0.27202779054641724 ---> 0.27320823073387146\n",
      "\u001b[0m===================== Epoch: [34/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.91it/s, Epoch=34, loss=0.781]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.07it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7963]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 34th epoch\n",
      "\n",
      " Final Train Loss: 0.7810  |  Final Val Loss: 0.7963\n",
      "Valid Dice: 0.3591 | Valid Jaccard: 0.2437\n",
      "\u001b[0m===================== Epoch: [35/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.94it/s, Epoch=35, loss=0.78] \n",
      "100%|██████████| 7/7 [00:01<00:00,  7.00it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7963]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 35th epoch\n",
      "\n",
      " Final Train Loss: 0.7798  |  Final Val Loss: 0.7963\n",
      "Valid Dice: 0.3572 | Valid Jaccard: 0.2422\n",
      "\u001b[0m===================== Epoch: [36/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.94it/s, Epoch=36, loss=0.78] \n",
      "100%|██████████| 7/7 [00:00<00:00,  7.02it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7953]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 36th epoch\n",
      "\n",
      " Final Train Loss: 0.7797  |  Final Val Loss: 0.7953\n",
      "Valid Dice: 0.3785 | Valid Jaccard: 0.2591\n",
      "\u001b[0m===================== Epoch: [37/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.91it/s, Epoch=37, loss=0.779]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.00it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7956]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 37th epoch\n",
      "\n",
      " Final Train Loss: 0.7788  |  Final Val Loss: 0.7956\n",
      "Valid Dice: 0.3809 | Valid Jaccard: 0.2620\n",
      "\u001b[0m===================== Epoch: [38/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.99it/s, Epoch=38, loss=0.779]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.32it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7956]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 38th epoch\n",
      "\n",
      " Final Train Loss: 0.7789  |  Final Val Loss: 0.7956\n",
      "Valid Dice: 0.3667 | Valid Jaccard: 0.2493\n",
      "\u001b[0m===================== Epoch: [39/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.96it/s, Epoch=39, loss=0.779]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.22it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7960]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 39th epoch\n",
      "\n",
      " Final Train Loss: 0.7788  |  Final Val Loss: 0.7960\n",
      "Valid Dice: 0.3569 | Valid Jaccard: 0.2410\n",
      "\u001b[0m===================== Epoch: [40/40] =====================\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 28/28 [00:07<00:00,  3.92it/s, Epoch=40, loss=0.778]\n",
      "100%|██████████| 7/7 [00:00<00:00,  7.06it/s, gpu_memory=3.37 GB, lr=0.00010, valid_loss=0.7955]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      " Saving model at 40th epoch\n",
      "\n",
      " Final Train Loss: 0.7781  |  Final Val Loss: 0.7955\n",
      "Valid Dice: 0.3567 | Valid Jaccard: 0.2406\n",
      "\n",
      " ====================== [Training Summary] ====================== \n",
      "\n",
      "Training complete in 0h 10m 17s\n",
      "avg time per [train + val] epoch 0h 0m 15s\n",
      "Best: ~~~~~~ dice: 0.3944  || jaccard 0.2732 || val_loss 0.7953 || train_loss 0.7781 ~~~~~~\n",
      "Saved training summary...\n",
      "\n",
      " =============================================================== \n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>epoch</th>\n",
       "      <th>Train_Loss</th>\n",
       "      <th>Valid_Loss</th>\n",
       "      <th>Valid_jaccard</th>\n",
       "      <th>Valid_dice</th>\n",
       "      <th>model_file</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0.813244</td>\n",
       "      <td>0.825458</td>\n",
       "      <td>0.094794</td>\n",
       "      <td>0.162351</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-1-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0.803067</td>\n",
       "      <td>0.812416</td>\n",
       "      <td>0.094794</td>\n",
       "      <td>0.162351</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-2-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>0.799428</td>\n",
       "      <td>0.810205</td>\n",
       "      <td>0.137355</td>\n",
       "      <td>0.221240</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-3-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0.797331</td>\n",
       "      <td>0.806945</td>\n",
       "      <td>0.149615</td>\n",
       "      <td>0.237221</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-4-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0.796013</td>\n",
       "      <td>0.805911</td>\n",
       "      <td>0.155449</td>\n",
       "      <td>0.244922</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-5-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>0.794384</td>\n",
       "      <td>0.805677</td>\n",
       "      <td>0.163795</td>\n",
       "      <td>0.255224</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-6-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>0.792853</td>\n",
       "      <td>0.804670</td>\n",
       "      <td>0.160081</td>\n",
       "      <td>0.250584</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-7-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>0.792080</td>\n",
       "      <td>0.802681</td>\n",
       "      <td>0.188688</td>\n",
       "      <td>0.287153</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-8-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>0.791235</td>\n",
       "      <td>0.801557</td>\n",
       "      <td>0.184180</td>\n",
       "      <td>0.281867</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-9-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>0.791341</td>\n",
       "      <td>0.802432</td>\n",
       "      <td>0.197862</td>\n",
       "      <td>0.298849</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-10-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>11</td>\n",
       "      <td>0.790789</td>\n",
       "      <td>0.803966</td>\n",
       "      <td>0.213211</td>\n",
       "      <td>0.319813</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-11-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>12</td>\n",
       "      <td>0.790009</td>\n",
       "      <td>0.806191</td>\n",
       "      <td>0.218715</td>\n",
       "      <td>0.325692</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-12-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>13</td>\n",
       "      <td>0.794593</td>\n",
       "      <td>0.802148</td>\n",
       "      <td>0.187269</td>\n",
       "      <td>0.284757</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-13-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>14</td>\n",
       "      <td>0.789831</td>\n",
       "      <td>0.801567</td>\n",
       "      <td>0.220617</td>\n",
       "      <td>0.328334</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-14-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>15</td>\n",
       "      <td>0.789057</td>\n",
       "      <td>0.802806</td>\n",
       "      <td>0.226766</td>\n",
       "      <td>0.335992</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-15-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>16</td>\n",
       "      <td>0.788772</td>\n",
       "      <td>0.799610</td>\n",
       "      <td>0.220318</td>\n",
       "      <td>0.328823</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-16-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>17</td>\n",
       "      <td>0.787093</td>\n",
       "      <td>0.802207</td>\n",
       "      <td>0.228337</td>\n",
       "      <td>0.339156</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-17-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>18</td>\n",
       "      <td>0.787919</td>\n",
       "      <td>0.799369</td>\n",
       "      <td>0.243945</td>\n",
       "      <td>0.358730</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-18-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>19</td>\n",
       "      <td>0.787846</td>\n",
       "      <td>0.799584</td>\n",
       "      <td>0.235733</td>\n",
       "      <td>0.347427</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-19-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>20</td>\n",
       "      <td>0.786624</td>\n",
       "      <td>0.799855</td>\n",
       "      <td>0.239031</td>\n",
       "      <td>0.352381</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-20-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>21</td>\n",
       "      <td>0.785458</td>\n",
       "      <td>0.799725</td>\n",
       "      <td>0.244289</td>\n",
       "      <td>0.360849</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-21-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>22</td>\n",
       "      <td>0.785937</td>\n",
       "      <td>0.801234</td>\n",
       "      <td>0.249257</td>\n",
       "      <td>0.364576</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-22-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>23</td>\n",
       "      <td>0.785745</td>\n",
       "      <td>0.798175</td>\n",
       "      <td>0.237721</td>\n",
       "      <td>0.351161</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-23-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>24</td>\n",
       "      <td>0.786266</td>\n",
       "      <td>0.798543</td>\n",
       "      <td>0.244839</td>\n",
       "      <td>0.360731</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-24-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>25</td>\n",
       "      <td>0.785546</td>\n",
       "      <td>0.798250</td>\n",
       "      <td>0.243942</td>\n",
       "      <td>0.359060</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-25-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>26</td>\n",
       "      <td>0.784895</td>\n",
       "      <td>0.797538</td>\n",
       "      <td>0.232893</td>\n",
       "      <td>0.346468</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-26-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>27</td>\n",
       "      <td>0.783534</td>\n",
       "      <td>0.797087</td>\n",
       "      <td>0.248712</td>\n",
       "      <td>0.366115</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-27-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>28</td>\n",
       "      <td>0.783116</td>\n",
       "      <td>0.797122</td>\n",
       "      <td>0.247919</td>\n",
       "      <td>0.364745</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-28-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>29</td>\n",
       "      <td>0.783207</td>\n",
       "      <td>0.798944</td>\n",
       "      <td>0.244048</td>\n",
       "      <td>0.360881</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-29-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>30</td>\n",
       "      <td>0.783399</td>\n",
       "      <td>0.797247</td>\n",
       "      <td>0.272028</td>\n",
       "      <td>0.392010</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-30-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>31</td>\n",
       "      <td>0.782146</td>\n",
       "      <td>0.797155</td>\n",
       "      <td>0.238547</td>\n",
       "      <td>0.354406</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-31-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>32</td>\n",
       "      <td>0.781013</td>\n",
       "      <td>0.797440</td>\n",
       "      <td>0.255647</td>\n",
       "      <td>0.375636</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-32-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>33</td>\n",
       "      <td>0.780917</td>\n",
       "      <td>0.797637</td>\n",
       "      <td>0.273208</td>\n",
       "      <td>0.394437</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-33-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>34</td>\n",
       "      <td>0.781041</td>\n",
       "      <td>0.796269</td>\n",
       "      <td>0.243733</td>\n",
       "      <td>0.359115</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-34-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>35</td>\n",
       "      <td>0.779844</td>\n",
       "      <td>0.796331</td>\n",
       "      <td>0.242240</td>\n",
       "      <td>0.357183</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-35-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>36</td>\n",
       "      <td>0.779657</td>\n",
       "      <td>0.795305</td>\n",
       "      <td>0.259079</td>\n",
       "      <td>0.378497</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-36-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>37</td>\n",
       "      <td>0.778846</td>\n",
       "      <td>0.795577</td>\n",
       "      <td>0.262012</td>\n",
       "      <td>0.380871</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-37-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>38</td>\n",
       "      <td>0.778886</td>\n",
       "      <td>0.795634</td>\n",
       "      <td>0.249264</td>\n",
       "      <td>0.366675</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-38-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>39</td>\n",
       "      <td>0.778769</td>\n",
       "      <td>0.796031</td>\n",
       "      <td>0.240996</td>\n",
       "      <td>0.356938</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-39-8.pth</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>40</td>\n",
       "      <td>0.778119</td>\n",
       "      <td>0.795508</td>\n",
       "      <td>0.240570</td>\n",
       "      <td>0.356706</td>\n",
       "      <td>./model_weights/Vanilla_Unet-0.0001-No-40-8.pth</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    epoch  Train_Loss  Valid_Loss  Valid_jaccard  Valid_dice  \\\n",
       "0       1    0.813244    0.825458       0.094794    0.162351   \n",
       "1       2    0.803067    0.812416       0.094794    0.162351   \n",
       "2       3    0.799428    0.810205       0.137355    0.221240   \n",
       "3       4    0.797331    0.806945       0.149615    0.237221   \n",
       "4       5    0.796013    0.805911       0.155449    0.244922   \n",
       "5       6    0.794384    0.805677       0.163795    0.255224   \n",
       "6       7    0.792853    0.804670       0.160081    0.250584   \n",
       "7       8    0.792080    0.802681       0.188688    0.287153   \n",
       "8       9    0.791235    0.801557       0.184180    0.281867   \n",
       "9      10    0.791341    0.802432       0.197862    0.298849   \n",
       "10     11    0.790789    0.803966       0.213211    0.319813   \n",
       "11     12    0.790009    0.806191       0.218715    0.325692   \n",
       "12     13    0.794593    0.802148       0.187269    0.284757   \n",
       "13     14    0.789831    0.801567       0.220617    0.328334   \n",
       "14     15    0.789057    0.802806       0.226766    0.335992   \n",
       "15     16    0.788772    0.799610       0.220318    0.328823   \n",
       "16     17    0.787093    0.802207       0.228337    0.339156   \n",
       "17     18    0.787919    0.799369       0.243945    0.358730   \n",
       "18     19    0.787846    0.799584       0.235733    0.347427   \n",
       "19     20    0.786624    0.799855       0.239031    0.352381   \n",
       "20     21    0.785458    0.799725       0.244289    0.360849   \n",
       "21     22    0.785937    0.801234       0.249257    0.364576   \n",
       "22     23    0.785745    0.798175       0.237721    0.351161   \n",
       "23     24    0.786266    0.798543       0.244839    0.360731   \n",
       "24     25    0.785546    0.798250       0.243942    0.359060   \n",
       "25     26    0.784895    0.797538       0.232893    0.346468   \n",
       "26     27    0.783534    0.797087       0.248712    0.366115   \n",
       "27     28    0.783116    0.797122       0.247919    0.364745   \n",
       "28     29    0.783207    0.798944       0.244048    0.360881   \n",
       "29     30    0.783399    0.797247       0.272028    0.392010   \n",
       "30     31    0.782146    0.797155       0.238547    0.354406   \n",
       "31     32    0.781013    0.797440       0.255647    0.375636   \n",
       "32     33    0.780917    0.797637       0.273208    0.394437   \n",
       "33     34    0.781041    0.796269       0.243733    0.359115   \n",
       "34     35    0.779844    0.796331       0.242240    0.357183   \n",
       "35     36    0.779657    0.795305       0.259079    0.378497   \n",
       "36     37    0.778846    0.795577       0.262012    0.380871   \n",
       "37     38    0.778886    0.795634       0.249264    0.366675   \n",
       "38     39    0.778769    0.796031       0.240996    0.356938   \n",
       "39     40    0.778119    0.795508       0.240570    0.356706   \n",
       "\n",
       "                                         model_file  \n",
       "0    ./model_weights/Vanilla_Unet-0.0001-No-1-8.pth  \n",
       "1    ./model_weights/Vanilla_Unet-0.0001-No-2-8.pth  \n",
       "2    ./model_weights/Vanilla_Unet-0.0001-No-3-8.pth  \n",
       "3    ./model_weights/Vanilla_Unet-0.0001-No-4-8.pth  \n",
       "4    ./model_weights/Vanilla_Unet-0.0001-No-5-8.pth  \n",
       "5    ./model_weights/Vanilla_Unet-0.0001-No-6-8.pth  \n",
       "6    ./model_weights/Vanilla_Unet-0.0001-No-7-8.pth  \n",
       "7    ./model_weights/Vanilla_Unet-0.0001-No-8-8.pth  \n",
       "8    ./model_weights/Vanilla_Unet-0.0001-No-9-8.pth  \n",
       "9   ./model_weights/Vanilla_Unet-0.0001-No-10-8.pth  \n",
       "10  ./model_weights/Vanilla_Unet-0.0001-No-11-8.pth  \n",
       "11  ./model_weights/Vanilla_Unet-0.0001-No-12-8.pth  \n",
       "12  ./model_weights/Vanilla_Unet-0.0001-No-13-8.pth  \n",
       "13  ./model_weights/Vanilla_Unet-0.0001-No-14-8.pth  \n",
       "14  ./model_weights/Vanilla_Unet-0.0001-No-15-8.pth  \n",
       "15  ./model_weights/Vanilla_Unet-0.0001-No-16-8.pth  \n",
       "16  ./model_weights/Vanilla_Unet-0.0001-No-17-8.pth  \n",
       "17  ./model_weights/Vanilla_Unet-0.0001-No-18-8.pth  \n",
       "18  ./model_weights/Vanilla_Unet-0.0001-No-19-8.pth  \n",
       "19  ./model_weights/Vanilla_Unet-0.0001-No-20-8.pth  \n",
       "20  ./model_weights/Vanilla_Unet-0.0001-No-21-8.pth  \n",
       "21  ./model_weights/Vanilla_Unet-0.0001-No-22-8.pth  \n",
       "22  ./model_weights/Vanilla_Unet-0.0001-No-23-8.pth  \n",
       "23  ./model_weights/Vanilla_Unet-0.0001-No-24-8.pth  \n",
       "24  ./model_weights/Vanilla_Unet-0.0001-No-25-8.pth  \n",
       "25  ./model_weights/Vanilla_Unet-0.0001-No-26-8.pth  \n",
       "26  ./model_weights/Vanilla_Unet-0.0001-No-27-8.pth  \n",
       "27  ./model_weights/Vanilla_Unet-0.0001-No-28-8.pth  \n",
       "28  ./model_weights/Vanilla_Unet-0.0001-No-29-8.pth  \n",
       "29  ./model_weights/Vanilla_Unet-0.0001-No-30-8.pth  \n",
       "30  ./model_weights/Vanilla_Unet-0.0001-No-31-8.pth  \n",
       "31  ./model_weights/Vanilla_Unet-0.0001-No-32-8.pth  \n",
       "32  ./model_weights/Vanilla_Unet-0.0001-No-33-8.pth  \n",
       "33  ./model_weights/Vanilla_Unet-0.0001-No-34-8.pth  \n",
       "34  ./model_weights/Vanilla_Unet-0.0001-No-35-8.pth  \n",
       "35  ./model_weights/Vanilla_Unet-0.0001-No-36-8.pth  \n",
       "36  ./model_weights/Vanilla_Unet-0.0001-No-37-8.pth  \n",
       "37  ./model_weights/Vanilla_Unet-0.0001-No-38-8.pth  \n",
       "38  ./model_weights/Vanilla_Unet-0.0001-No-39-8.pth  \n",
       "39  ./model_weights/Vanilla_Unet-0.0001-No-40-8.pth  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x576 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "history = defaultdict(list)\n",
    "best_val_dice = -float('inf')\n",
    "best_val_jaccard = -float('inf')\n",
    "start = time.time()\n",
    "\n",
    "\n",
    "for epoch in range(CONFIG[\"NUM_EPOCHS\"]):\n",
    "    \n",
    "    print(f'{sr_}===================== Epoch: [{epoch+1}/{CONFIG[\"NUM_EPOCHS\"]}] =====================')\n",
    "\n",
    "    # train and validation loop\n",
    "    train_loss = train_fn(train_loader, model, optimizer, loss_fn, epoch+1)\n",
    "    val_loss, val_scores = valid_one_epoch(model, test_loader,CONFIG[\"DEVICE\"], epoch+1)\n",
    "    val_dice, val_jaccard = val_scores\n",
    "    \n",
    "    \n",
    "    # save model\n",
    "    model_file = save_model(model, optimizer, loss_fn, epoch+1)\n",
    "    \n",
    "    \n",
    "    # weight and baise Log the metrics\n",
    "    wandb.log({\"Train Loss\": train_loss})\n",
    "    wandb.log({\"Valid Loss\": val_loss})\n",
    "    wandb.log({\"Valid dice\": val_dice})\n",
    "    wandb.log({\"Valid jaccard\": val_jaccard})\n",
    "\n",
    "    # logging\n",
    "    history[\"epoch\"].append(epoch+1)\n",
    "    history['Train_Loss'].append(train_loss)\n",
    "    history['Valid_Loss'].append(val_loss)\n",
    "    history['Valid_jaccard'].append(val_jaccard)\n",
    "    history['Valid_dice'].append(val_dice)\n",
    "    history[\"model_file\"].append(model_file)\n",
    "\n",
    "    \n",
    "    # print loss and scores\n",
    "    print(f\"\\n Final Train Loss: {train_loss:0.4f}  |  Final Val Loss: {val_loss:0.4f}\") \n",
    "    \n",
    "    print(f'Valid Dice: {val_dice:0.4f} | Valid Jaccard: {val_jaccard:0.4f}')\n",
    "    \n",
    "    \n",
    "    # print best val score\n",
    "    if best_val_dice < val_dice:\n",
    "        print(f\"{b_}val dice increased: {best_val_dice} ---> {val_dice}\")\n",
    "        best_val_dice = val_dice\n",
    "    \n",
    "    if best_val_jaccard < val_jaccard:\n",
    "        print(f\"{b_}val jaccard increased: {best_val_jaccard} ---> {val_jaccard}\")\n",
    "        best_val_jaccard = val_jaccard\n",
    "\n",
    "        \n",
    "        \n",
    "# print time\n",
    "end = time.time()\n",
    "time_elapsed = end - start\n",
    "avg_time_per_epoch = time_elapsed/CONFIG[\"NUM_EPOCHS\"] \n",
    "summary_df = pd.DataFrame.from_dict(history)\n",
    "\n",
    "print(f'\\n ====================== [Training Summary] ====================== \\n')\n",
    "print('Training complete in {:.0f}h {:.0f}m {:.0f}s'.format(\n",
    "    time_elapsed // 3600, (time_elapsed % 3600) // 60, (time_elapsed % 3600) % 60))\n",
    "\n",
    "print('avg time per [train + val] epoch {:.0f}h {:.0f}m {:.0f}s'.format(\n",
    "    avg_time_per_epoch // 3600, (avg_time_per_epoch % 3600) // 60, (avg_time_per_epoch % 3600) % 60))\n",
    "\n",
    "print(\"Best: ~~~~~~ dice: {:.4f}  || jaccard {:.4f} || val_loss {:.4f} || train_loss {:.4f} ~~~~~~\".format(\n",
    "    best_val_dice, best_val_jaccard, min(history['Valid_Loss']), min(history['Train_Loss'])))\n",
    "\n",
    "\n",
    "summary_df.to_csv('training_summary.csv')\n",
    "print(\"Saved training summary...\")\n",
    "\n",
    "print(f'\\n =============================================================== \\n')\n",
    "\n",
    "display(summary_df)\n",
    "\n",
    "# plotting training and validation loss\n",
    "plt.figure(figsize=(15,8))\n",
    "\n",
    "fig1 = plt.subplot(2,2,1)\n",
    "fig1.plot(history['epoch'], history['Train_Loss'], color='lime', marker='>')\n",
    "fig1.set_title(\"train loss [dice]\")\n",
    "fig1.set(xlabel='epoch', ylabel='train loss')\n",
    "\n",
    "fig2 = plt.subplot(2,2,2)\n",
    "fig2.plot(history['epoch'], history['Valid_Loss'], color='cyan',  marker='>')\n",
    "fig2.set_title(\"val loss [dice]\")\n",
    "fig2.set(xlabel='epoch', ylabel='val loss')\n",
    "\n",
    "fig3 = plt.subplot(2,2,3)\n",
    "fig3.plot(history['epoch'], history['Valid_dice'], color='orange',  marker='>')\n",
    "fig3.set_title(\"Val Dice\")\n",
    "fig3.set(xlabel='epoch', ylabel='val dice')\n",
    "\n",
    "fig4 = plt.subplot(2,2,4)\n",
    "fig4.plot(history['epoch'], history['Valid_jaccard'], color='lightcoral',  marker='>')\n",
    "fig4.set_title(\"val jaccard\")\n",
    "fig4.set(xlabel='epoch', ylabel='val jaccard')\n",
    "\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7b4863ea",
   "metadata": {},
   "outputs": [],
   "source": [
    "import gc\n",
    "a = gc.collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f3e59b89",
   "metadata": {
    "papermill": {
     "duration": 0.246894,
     "end_time": "2022-07-24T09:57:17.343143",
     "exception": false,
     "start_time": "2022-07-24T09:57:17.096249",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Model File Check:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "adbe50d4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-07-24T09:57:17.848382Z",
     "iopub.status.busy": "2022-07-24T09:57:17.848020Z",
     "iopub.status.idle": "2022-07-24T09:57:19.724048Z",
     "shell.execute_reply": "2022-07-24T09:57:19.723024Z"
    },
    "papermill": {
     "duration": 2.131207,
     "end_time": "2022-07-24T09:57:19.726648",
     "exception": false,
     "start_time": "2022-07-24T09:57:17.595441",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total 6937288\n",
      "drwxr-xr-x 2 root root      4096 Jul 24 09:57 .\n",
      "drwxr-xr-x 4 root root      4096 Jul 24 09:57 ..\n",
      "-rw-r--r-- 1 root root 177593833 Jul 24 09:04 Vanilla_Unet-0.0001-No-1-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:16 Vanilla_Unet-0.0001-No-10-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:17 Vanilla_Unet-0.0001-No-11-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:18 Vanilla_Unet-0.0001-No-12-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:20 Vanilla_Unet-0.0001-No-13-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:21 Vanilla_Unet-0.0001-No-14-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:22 Vanilla_Unet-0.0001-No-15-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:24 Vanilla_Unet-0.0001-No-16-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:25 Vanilla_Unet-0.0001-No-17-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:26 Vanilla_Unet-0.0001-No-18-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:28 Vanilla_Unet-0.0001-No-19-8.pth\n",
      "-rw-r--r-- 1 root root 177593833 Jul 24 09:05 Vanilla_Unet-0.0001-No-2-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:29 Vanilla_Unet-0.0001-No-20-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:30 Vanilla_Unet-0.0001-No-21-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:32 Vanilla_Unet-0.0001-No-22-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:33 Vanilla_Unet-0.0001-No-23-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:35 Vanilla_Unet-0.0001-No-24-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:36 Vanilla_Unet-0.0001-No-25-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:37 Vanilla_Unet-0.0001-No-26-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:39 Vanilla_Unet-0.0001-No-27-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:40 Vanilla_Unet-0.0001-No-28-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:42 Vanilla_Unet-0.0001-No-29-8.pth\n",
      "-rw-r--r-- 1 root root 177593833 Jul 24 09:06 Vanilla_Unet-0.0001-No-3-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:43 Vanilla_Unet-0.0001-No-30-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:44 Vanilla_Unet-0.0001-No-31-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:46 Vanilla_Unet-0.0001-No-32-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:47 Vanilla_Unet-0.0001-No-33-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:49 Vanilla_Unet-0.0001-No-34-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:50 Vanilla_Unet-0.0001-No-35-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:51 Vanilla_Unet-0.0001-No-36-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:53 Vanilla_Unet-0.0001-No-37-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:54 Vanilla_Unet-0.0001-No-38-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:55 Vanilla_Unet-0.0001-No-39-8.pth\n",
      "-rw-r--r-- 1 root root 177593833 Jul 24 09:08 Vanilla_Unet-0.0001-No-4-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:57 Vanilla_Unet-0.0001-No-40-8.pth\n",
      "-rw-r--r-- 1 root root 177593833 Jul 24 09:09 Vanilla_Unet-0.0001-No-5-8.pth\n",
      "-rw-r--r-- 1 root root 177593833 Jul 24 09:10 Vanilla_Unet-0.0001-No-6-8.pth\n",
      "-rw-r--r-- 1 root root 177593833 Jul 24 09:12 Vanilla_Unet-0.0001-No-7-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:13 Vanilla_Unet-0.0001-No-8-8.pth\n",
      "-rw-r--r-- 1 root root 177593897 Jul 24 09:14 Vanilla_Unet-0.0001-No-9-8.pth\n"
     ]
    }
   ],
   "source": [
    "!ls -la ./model_weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bd81d964",
   "metadata": {
    "papermill": {
     "duration": 0.248845,
     "end_time": "2022-07-24T09:57:20.230560",
     "exception": false,
     "start_time": "2022-07-24T09:57:19.981715",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f139e730",
   "metadata": {
    "papermill": {
     "duration": 0.259865,
     "end_time": "2022-07-24T09:57:20.794504",
     "exception": false,
     "start_time": "2022-07-24T09:57:20.534639",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e36cfc42",
   "metadata": {
    "papermill": {
     "duration": 0.241472,
     "end_time": "2022-07-24T09:57:21.287311",
     "exception": false,
     "start_time": "2022-07-24T09:57:21.045839",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ec12cca5",
   "metadata": {
    "papermill": {
     "duration": 0.249454,
     "end_time": "2022-07-24T09:57:21.788957",
     "exception": false,
     "start_time": "2022-07-24T09:57:21.539503",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.7.0 ('.venv': venv)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 3332.697837,
   "end_time": "2022-07-24T09:57:24.763901",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2022-07-24T09:01:52.066064",
   "version": "2.3.4"
  },
  "vscode": {
   "interpreter": {
    "hash": "21a2a557654fc1676068684031cf9bb9dfda94e124d3623f4e9c9ed764d794ac"
   }
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {
     "15fcc8486d184eb38e54498e8c8404d2": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "VBoxModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "VBoxView",
       "box_style": "",
       "children": [
        "IPY_MODEL_e53d0b17ab73405398a50e5eaee95e42",
        "IPY_MODEL_9f58bfd913b74e339cd507a4fc97ebeb"
       ],
       "layout": "IPY_MODEL_1f3677f6546e40368542835bad7a616d"
      }
     },
     "1f3677f6546e40368542835bad7a616d": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "30dd58e4f3c0480ba9776ac529132d4d": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "30ec4205cc144341ae4a852e1b541a14": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "4155a9d4e38544dc84417849e9fc5b94": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "62cc7aab6dea4c34ba0e5f4112e18efa": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "ProgressStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "bar_color": null,
       "description_width": ""
      }
     },
     "9926b22c86044a2c9050c12ed14af7f4": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "9f58bfd913b74e339cd507a4fc97ebeb": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "FloatProgressModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "ProgressView",
       "bar_style": "",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_f2ee84d627d549bd9389c643e144ead3",
       "max": 1,
       "min": 0,
       "orientation": "horizontal",
       "style": "IPY_MODEL_c63357b7e4ef436cbd86f0d868915b1a",
       "value": 0
      }
     },
     "a780dd55d62d459aad0a02cec435c2b6": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "VBoxModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "VBoxModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "VBoxView",
       "box_style": "",
       "children": [
        "IPY_MODEL_f3c9f8bc4fce4fb3807ffdfb160d0462",
        "IPY_MODEL_e1a7a70a73e149c783dc2fc6fbb695e8"
       ],
       "layout": "IPY_MODEL_4155a9d4e38544dc84417849e9fc5b94"
      }
     },
     "b90386d424e04302a7c4858ae9c30d58": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "bf208c556485412aa1cd0be497a249bd": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "c63357b7e4ef436cbd86f0d868915b1a": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "ProgressStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "bar_color": null,
       "description_width": ""
      }
     },
     "e1a7a70a73e149c783dc2fc6fbb695e8": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "FloatProgressModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "ProgressView",
       "bar_style": "",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_bf208c556485412aa1cd0be497a249bd",
       "max": 1,
       "min": 0,
       "orientation": "horizontal",
       "style": "IPY_MODEL_62cc7aab6dea4c34ba0e5f4112e18efa",
       "value": 0
      }
     },
     "e53d0b17ab73405398a50e5eaee95e42": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "LabelModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "LabelModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "LabelView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_30ec4205cc144341ae4a852e1b541a14",
       "placeholder": "​",
       "style": "IPY_MODEL_9926b22c86044a2c9050c12ed14af7f4",
       "value": ""
      }
     },
     "f2ee84d627d549bd9389c643e144ead3": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "f3c9f8bc4fce4fb3807ffdfb160d0462": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "LabelModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "LabelModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "LabelView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_b90386d424e04302a7c4858ae9c30d58",
       "placeholder": "​",
       "style": "IPY_MODEL_30dd58e4f3c0480ba9776ac529132d4d",
       "value": ""
      }
     }
    },
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}