From c8b30db42f08ae6fa78d2c58af32bf9542ab58f9 Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Wed, 6 Jul 2022 12:25:23 -0700
Subject: [PATCH 01/17] begin work

---
 .../swin_unetr_btcv_segmentation_3d.ipynb     | 422 +++++++++---------
 1 file changed, 213 insertions(+), 209 deletions(-)

diff --git a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
index eb5b57822b..82c9b0ffe4 100644
--- a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
+++ b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
@@ -79,6 +79,27 @@
     "If training from scratch is desired, please skip the step for initializing from pre-trained weights. "
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "False"
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import torch\n",
+    "torch.cuda.is_available()"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -88,11 +109,46 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 1,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
+      "Collecting nibabel==3.1.1\n",
+      "  Downloading nibabel-3.1.1-py3-none-any.whl (3.3 MB)\n",
+      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.3/3.3 MB\u001b[0m \u001b[31m63.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n",
+      "\u001b[?25hRequirement already satisfied: packaging>=14.3 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (21.3)\n",
+      "Requirement already satisfied: numpy>=1.13 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (1.22.3)\n",
+      "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.8/site-packages (from packaging>=14.3->nibabel==3.1.1) (3.0.8)\n",
+      "Installing collected packages: nibabel\n",
+      "  Attempting uninstall: nibabel\n",
+      "    Found existing installation: nibabel 3.2.2\n",
+      "    Uninstalling nibabel-3.2.2:\n",
+      "      Successfully uninstalled nibabel-3.2.2\n",
+      "Successfully installed nibabel-3.1.1\n",
+      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
+      "\u001b[0mLooking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
+      "Collecting tqdm==4.63.0\n",
+      "  Downloading tqdm-4.63.0-py2.py3-none-any.whl (76 kB)\n",
+      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m76.6/76.6 kB\u001b[0m \u001b[31m18.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+      "\u001b[?25hInstalling collected packages: tqdm\n",
+      "  Attempting uninstall: tqdm\n",
+      "    Found existing installation: tqdm 4.64.0\n",
+      "    Uninstalling tqdm-4.64.0:\n",
+      "      Successfully uninstalled tqdm-4.64.0\n",
+      "Successfully installed tqdm-4.63.0\n",
+      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
+      "\u001b[0m"
+     ]
+    }
+   ],
    "source": [
-    "!pip install git+https://github.com/Project-MONAI/MONAI#egg.gitmonai@0.8.1+271.g07de215c \n",
+    "# FIXME: check version compatibility\n",
+    "# !pip install git+https://github.com/Project-MONAI/MONAI#egg.gitmonai@0.8.1+271.g07de215c \n",
+    "\n",
     "!pip install nibabel==3.1.1\n",
     "!pip install tqdm==4.63.0\n",
     "!python -c \"import matplotlib\" || pip install -q matplotlib\n",
@@ -101,9 +157,51 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 2,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.8/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"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "MONAI version: 0.9.0\n",
+      "Numpy version: 1.22.3\n",
+      "Pytorch version: 1.12.0a0+bd13bc6\n",
+      "MONAI flags: HAS_EXT = True, USE_COMPILED = False\n",
+      "MONAI rev id: af0e0e9f757558d144b655c63afcea3a4e0a06f5\n",
+      "MONAI __file__: /opt/monai/monai/__init__.py\n",
+      "\n",
+      "Optional dependencies:\n",
+      "Pytorch Ignite version: 0.4.8\n",
+      "Nibabel version: 3.1.1\n",
+      "scikit-image version: 0.19.3\n",
+      "Pillow version: 9.0.1\n",
+      "Tensorboard version: 2.8.0\n",
+      "gdown version: 4.4.0\n",
+      "TorchVision version: 0.13.0a0\n",
+      "tqdm version: 4.63.0\n",
+      "lmdb version: 1.3.0\n",
+      "psutil version: 5.9.0\n",
+      "pandas version: 1.3.5\n",
+      "einops version: 0.4.1\n",
+      "transformers version: 4.19.4\n",
+      "mlflow version: 1.26.1\n",
+      "pynrrd version: 0.4.3\n",
+      "\n",
+      "For details about installing the optional dependencies, please visit:\n",
+      "    https://docs.monai.io/en/latest/installation.html#installing-the-recommended-dependencies\n",
+      "\n"
+     ]
+    }
+   ],
    "source": [
     "import os\n",
     "import shutil\n",
@@ -156,14 +254,22 @@
     "\n",
     "You can specify a directory with the `MONAI_DATA_DIRECTORY` environment variable.  \n",
     "This allows you to save results and reuse downloads.  \n",
-    "If not specified a temporary directory will be used."
+    "If not specified, a temporary directory will be used."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 3,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "/tmp/tmpo69cno4c\n"
+     ]
+    }
+   ],
    "source": [
     "directory = os.environ.get(\"MONAI_DATA_DIRECTORY\")\n",
     "root_dir = tempfile.mkdtemp() if directory is None else directory\n",
@@ -180,12 +286,13 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [],
    "source": [
-    "num_samples = 4\n",
+    "num_samples = 1\n",
     "\n",
+    "# TODO: EnsureTyped + ToDeviced\n",
     "train_transforms = Compose(\n",
     "    [\n",
     "        LoadImaged(keys=[\"image\", \"label\"]),\n",
@@ -267,196 +374,33 @@
    "metadata": {},
    "source": [
     " ## Download dataset and format in the folder.\n",
-    "    1. Download dataset from here: https://www.synapse.org/#!Synapse:syn3193805/wiki/89480\\n\n",
-    "    2. Put images in the ./data/imagesTr\n",
-    "    3. Put labels in the ./data/labelsTr\n",
-    "    4. make JSON file accordingly: ./data/dataset_0.json\n",
-    "    Example of JSON file:\n",
-    "     {\n",
-    "    \"description\": \"btcv yucheng\",\n",
-    "    \"labels\": {\n",
-    "        \"0\": \"background\",\n",
-    "        \"1\": \"spleen\",\n",
-    "        \"2\": \"rkid\",\n",
-    "        \"3\": \"lkid\",\n",
-    "        \"4\": \"gall\",\n",
-    "        \"5\": \"eso\",\n",
-    "        \"6\": \"liver\",\n",
-    "        \"7\": \"sto\",\n",
-    "        \"8\": \"aorta\",\n",
-    "        \"9\": \"IVC\",\n",
-    "        \"10\": \"veins\",\n",
-    "        \"11\": \"pancreas\",\n",
-    "        \"12\": \"rad\",\n",
-    "        \"13\": \"lad\"\n",
-    "    },\n",
-    "    \"licence\": \"yt\",\n",
-    "    \"modality\": {\n",
-    "        \"0\": \"CT\"\n",
-    "    },\n",
-    "    \"name\": \"btcv\",\n",
-    "    \"numTest\": 20,\n",
-    "    \"numTraining\": 80,\n",
-    "    \"reference\": \"Vanderbilt University\",\n",
-    "    \"release\": \"1.0 06/08/2015\",\n",
-    "    \"tensorImageSize\": \"3D\",\n",
-    "    \"test\": [\n",
-    "        \"imagesTs/img0061.nii.gz\",\n",
-    "        \"imagesTs/img0062.nii.gz\",\n",
-    "        \"imagesTs/img0063.nii.gz\",\n",
-    "        \"imagesTs/img0064.nii.gz\",\n",
-    "        \"imagesTs/img0065.nii.gz\",\n",
-    "        \"imagesTs/img0066.nii.gz\",\n",
-    "        \"imagesTs/img0067.nii.gz\",\n",
-    "        \"imagesTs/img0068.nii.gz\",\n",
-    "        \"imagesTs/img0069.nii.gz\",\n",
-    "        \"imagesTs/img0070.nii.gz\",\n",
-    "        \"imagesTs/img0071.nii.gz\",\n",
-    "        \"imagesTs/img0072.nii.gz\",\n",
-    "        \"imagesTs/img0073.nii.gz\",\n",
-    "        \"imagesTs/img0074.nii.gz\",\n",
-    "        \"imagesTs/img0075.nii.gz\",\n",
-    "        \"imagesTs/img0076.nii.gz\",\n",
-    "        \"imagesTs/img0077.nii.gz\",\n",
-    "        \"imagesTs/img0078.nii.gz\",\n",
-    "        \"imagesTs/img0079.nii.gz\",\n",
-    "        \"imagesTs/img0080.nii.gz\"\n",
-    "    ],\n",
-    "    \"training\": [\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0001.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0001.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0002.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0002.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0003.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0003.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0004.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0004.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0005.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0005.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0006.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0006.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0007.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0007.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0008.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0008.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0009.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0009.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0010.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0010.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0021.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0021.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0022.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0022.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0023.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0023.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0024.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0024.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0025.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0025.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0026.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0026.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0027.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0027.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0028.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0028.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0029.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0029.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0030.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0030.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0031.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0031.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0032.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0032.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0033.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0033.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0034.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0034.nii.gz\"\n",
-    "        }\n",
-    "    ],\n",
-    "    \"validation\": [\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0035.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0035.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0036.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0036.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0037.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0037.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0038.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0038.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0039.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0039.nii.gz\"\n",
-    "        },\n",
-    "        {\n",
-    "            \"image\": \"imagesTr/img0040.nii.gz\",\n",
-    "            \"label\": \"labelsTr/label0040.nii.gz\"\n",
-    "        }\n",
-    "    ]\n",
-    "}\n",
-    "    "
+    "1. Download dataset from here: https://www.synapse.org/#!Synapse:syn3193805/wiki/89480. After you open the link, navigate to the \"Files\" tab, then download Abdomen/RawData.zip. \n",
+    "\n",
+    "    Note that you may need to register for an account on Synapse and consent to use agreements before being able to view/download this file. There are options to download directly from the browser or from the command line; please refer to Synapse API documentation for more info.\n",
+    "\n",
+    "2. After downloading the zip file, unzip. Then put images from `RawData/Training/img` in `./data/imagesTr`, and put labels from `RawData/Training/label` in `./data/labelsTr`.\n",
+    "\n",
+    "3. Make a JSON file to define train/val split and other relevant parameters. Place the JSON file at `./data/dataset_0.json`.\n",
+    "\n",
+    "    An example of the JSON file is provided in the current directory. If you would like to use this directly, please move it into the `./data` folder.   "
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 5,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Loading dataset: 100%|███████████████████████████████████████████████████████████████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Loading dataset: 100%|█████████████████████████████████████████████████████████████████████| 6/6 [00:10<00:00,  1.81s/it]\n"
+     ]
+    }
+   ],
    "source": [
-    "data_dir = \"/data/\"\n",
+    "data_dir = \"data/\"\n",
     "split_JSON = \"dataset_0.json\"\n",
     "\n",
     "datasets = data_dir + split_JSON\n",
@@ -469,6 +413,8 @@
     "    cache_rate=1.0,\n",
     "    num_workers=8,\n",
     ")\n",
+    "\n",
+    "# TODO: try ThreadDataLoader (and num_workers)\n",
     "train_loader = DataLoader(\n",
     "    train_ds, batch_size=1, shuffle=True, num_workers=8, pin_memory=True\n",
     ")\n",
@@ -501,7 +447,7 @@
     },
     {
      "data": {
-      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2kAAAF1CAYAAAByA6JvAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOz9ebQc130djO7qubqreu47AxcAARAgSBGcRFGSJZGWRMuanu0oy8/2spf8vOIkS46cl8RxsvzZeS9vJU6cvMSJP7980Wc78Wx9sWVLtmbTlGWZmkiKIkGCmAhc4I49d1d19Vzvj4v9u6cbFySmiwvgnr0WFu7trq4651SBPLv377e34fs+NDQ0NDQ0NDQ0NDQ0NG4NBLZ7ABoaGhoaGhoaGhoaGhob0CRNQ0NDQ0NDQ0NDQ0PjFoImaRoaGhoaGhoaGhoaGrcQNEnT0NDQ0NDQ0NDQ0NC4haBJmoaGhoaGhoaGhoaGxi0ETdI0NDQ0NDQ0NDQ0NDRuIWiSprEtMAzjmGEY79rucWhoaGhoaGhoEIZhnDUM491XcJxvGMb+a7zGNX9WY+cgtN0D0NiZ8H3/yHaPQUNDQ0NDQ0NDQ+NWhFbSNDQ0NDQ0NDQ0NDQ0biFokqaxLWA5gWEY/8owjP/LMIzfNQyjaRjGi4ZhHDQM418YhrFmGMZ5wzDeq3zuo4ZhvHLx2DOGYfz02Hl/zjCMZcMwlgzD+Cm1pMAwjKhhGP/BMIwFwzBWDcP4b4ZhmDd77hoaGhoaGhq3NgzDeLNhGM8YhlG7uK/4NcMwImOHff/FvUjJMIxfMQwjoHz+Jy/uV6qGYXzBMIz5mzwFjdscmqRp3Ar4IIDfAZAB8DyAL2D92ZwF8P8G8H8ox64B+ACAJICPAvhPhmE8CACGYXwfgP8ngHcD2A/gXWPX+WUABwEcvfj+LIBf3IL5aGhoaGhoaNzeGAD4xwDyAB4D8L0A/uHYMT8A4GEADwL4MICfBADDMD4M4F8C+EEABQBfBfAHN2XUGncMDN/3t3sMGjsQhmGcBfBTAN4O4G2+77/n4usfxPp/yFK+7w8Mw7ABNABkfN+vbXKePwXwV77v/6phGL8JYNX3/X9x8b39AE4COADgNAAHwJt83z998f3HAPy+7/t7t3KuGhoaGhoaGrcHuD/xff/LY6//LIB3+r7/Axd/9wG8z/f9z1/8/R8C+CHf97/XMIzPAfhfvu//xsX3Aljfgxz2ff/cxc8e8H3/1M2al8btB62kadwKWFV+9gCUfN8fKL8DgAUAhmG8zzCMrxuGUTEMowbg+7H+LRcAzAA4r5xL/bkAIA7g2YulCzUAn7/4uoaGhoaGhoaG4GLrxZ8bhrFiGEYDwL/Bxn6DUPcZ57C+DwGAeQC/quw3KgAMrFfwaGhcETRJ07htYBhGFMAfA/gPACZ9308D+CzW/8MHAMsA5pSP7FJ+LmGd8B3xfT998U/K931r60euoaGhoaGhcZvh/wfgONYVryTWyxeNsWPUfcZuAEsXfz4P4KeV/Uba933T9/2/3fJRa9wx0CRN43ZCBEAUQBFA3zCM9wF4r/L+JwF81DCMw4ZhxAH8b3zD9/0hgE9gvYdtAgAMw5g1DOPJmzZ6DQ0NDQ0NjdsFbLdwDMM4BOAfbHLMPzMMI2MYxi4AHwfwRxdf/28A/oVhGEcAwDCMlGEYH7kZg9a4c6BJmsZtA9/3mwD+EdbJWBXAjwD4tPL+5wD8FwB/BeAUgK9ffKtz8e9/ztcvli58GcDdN2XwGhoaGhoaGrcT/inW9xlNrH/J+0ebHPNnAJ4F8B0AfwHgNwDA9/1PAfh3AP7w4n7jJQDv2/oha9xJ0MYhGncsDMM4jPX/MEZ93+9v93g0NDQ0NDQ0NDQ0rgRaSdO4o2AYxg9czEPLYP1brM9ogqahoaGhoaGhoXE7QZM0jTsNP431LLXTWM842ayGXENDQ0NDQ0NDQ+OWxZaRNMMwvs8wjFcNwzhlGMbPb9V1NDRU+L7/fRddG7O+7/+A7/vL2z0mDQ0NDY3tgd6LaGho3K7Ykp40wzCCAE4AeA+ACwC+BeD/7vv+yzf8YhoaGhoaGhoaY9B7EQ0NjdsZW6WkvRnAKd/3z/i+3wXwhwA+vEXX0tDQ0NDQ0NAYh96LaGho3LYIbdF5ZzGawn4BwKPqAYZh/D0Af+/irw9t0TgQjUYRj8cRDodhGAYCgQACgQAMYz2PcDgcwvM8OI6DXq932fNMTEyg0+mg0+mg3W5v1XA1rgCGYSASiSAUCsm9DAaDck9934dhGPI7MRwOR14zDAO+74/8GQ6H6Pf76PV66Pe138idDN/3x0NJNTQ07iy84V4EGN2PBBF8KI7kzRmdhobGjkcbLrp+Z9P9yFaRtDeE7/v/HcB/BwDDMG54zWU4HMbs7CweeeQRBINBBINBWJaFRx99FJOTkwiHwxgOh+h0OnjllVfwP/7H/8Dx48cvOY9t23j44YcxNzeHb37zm7hw4cJVjSOTycDzPE3srhLJZBKWZcG2bdi2LcQrEAggEokgm82OkO9kMgnDMIRIBwIBxGIxIVqBQAC1Wg2xWEyIeigUQrfbxWAwEJLW6/XgeR7K5TIcx5Fr9vt9dLtd9Ho91Ot1VCoVDIfDbV4lDQ0NDY3rhbofSRpZ/1Hje7d5RBoaGjsF3/D/8rLvbRVJWwSwS/l97uJrNw2maeKRRx7Bnj178LnPfQ5nz55FOp3GzMwMCoUCgA1lZXZ2FvF4fNNz7N69G+9+97vxH//jf0S9XsdgMLiqcWSzWZTLZU3SXgfRaBThcBjBYBAA5J5MT09jamoKMzMzsu4kV67rotPpCHGq1WrodDpwHEfWOhaLwfM8+L6PUCiE1dVVWJaFUCiEYDCISCQi5IznjcVisG0bqVQKwWAQpmkiGo2i3W6j3W7DcRwsLCyI4gZAfu50OtC5gxoaGhq3DLZ9L6KhoaFxrdgqkvYtAAcMw9iL9f8g/jDWU9tvGkzTxJvf/Gb84i/+IjzPAwA4joNf+qVfwi/8wi/g4MGDSCaTiMfjOHPmDDqdziXnePe73427774b/+E//AdUq9VrGsfp06evax53OkKhEB566CHcddddmJiYALCuXjabTbiuC8/zsLy8jFqthlarhX6/j1gshna7jVBo/fEdDAbo9XoYDAbo9/uiesXjcSFsgUAAzWYTnufBMAyEQiFEo1G4rjsynlgsJkQrFAohkUggEAggHo/DNE3E43EcPnwYDzzwAKLRKACg2WzizJkzeP7559FoNG7W0mloaGhovD62fS+ioaGhca3YEpLm+37fMIyPAfgCgCCA3/R9/9hWXOtyaLVa+NrXvrZpX9GJEyeQzWaRyWTQ6XTwb//tv0W5XB455ud+7ufw6quv4rd/+7dRq9Wu+vqhUAj79+/HqVOndG+TgoMHD+LgwYOYmZlBMpmU8sTBYIDV1VWcP38e0WgU3W4X/X4fvu9jMBgIQTMMA4lEAq1WC9FoFLFYDLFYTPrLqGqRqAUC6944w+FQ1DqStHA4DABC9vhZErlgMAjP86Q8MhwOIxwOo9vtIhKJIBKJIBqNIplMYteuXdi/fz8Mw0C/30e9XsdXvvIVNBoNff81NDQ0tgG3wl5EQ0ND41qxZT1pvu9/FsBnt+r8bwTP8/Dtb3970/LEqakp2LaNYrGIT3/60yiXy7KRDofDeOKJJ/Ctb30Lr732GiqVyjWVsPm+j3a7vePL3/bv34+JiQnYto1AIIBoNCo9ZtVqFY1GA47jiFlHsVhEJBKR0kcSL95HGoKQbJHEqaQL2CBePN4wDPR6PQSDQQwGAwQCASQSCTlONQ4hVIMZVa1rt9vodrsIBoMIh8NotVqoVqvSI8eyyTe/+c0IBoMYDodotVo4efIkisXi6xrUaGhoaGjcOGz3XkRDQ0PjWrFtxiFbjX6/f4nJRygUwr333ou9e/ei1WrhxIkTeOqpp4SgJRIJTE9PI51O48tf/jJqtdpV96AB60QvFouh2WzuOJKWSCSQSCQQi8UArJO0yclJIWalUkncNFutFur1OrrdLobDoRi5DIdDMfggaSIBIzGiGqa6O0ajUYRCIRiGgWg0il6vN+LmSZLW6/Xg+z4SiYQQLypu/X4foVDokvtGVW84HCIQCAg5HAwG6Ha7qNfriEQiorblcjlkMhkkk0mEQiG0Wi0MBgPE43FxjtS9ihoaGhoaGhoaGpvhjiVpKliqls1m8a/+1b+C7/v45Cc/iT/5kz+RXrRAIIBdu3bhHe94B377t3/7ukwgkskkCoXCpm6RdyJUZWvv3r04cOAA5ubmAADdbhftdhtra2toNBpCnEOhEIbDIbrdrpCiwWCAaDSKwWAgZIzkzbZtAOuKl2malyhtJEg0IfF9H47jCLlTHRz5OSpaLIMMh8PodDpIp9NyDN/v9Xro9Xpot9uIRCLodrsj6lu32wUAKd2kk2QymRTSevToUVHums0m/vIv/xLLy8sYDAYyTw0NDQ0NDQ0NDY0dQdIef/xxvOtd78ITTzyB4XCIn/mZn8Hx48dHzEIee+wxDAYDfOITn7hu9WtychIPPPDAjiBppmmiUCjgPe95D7LZrJT2ra6uwnVd1Ot1tNtt9Pt9BAIBOI6DeDwuqplalhgKhaTXi8QmEomg1+shkUiMlCMGAgEEg0FRydLpNAaDgZA1z/OQSCQwHA7lcyRZVM+CwSDy+fxIZl6v10M4HEa/38dwOEQoFMLMzIzY+bdaLXGUZAkk58nXe70eut0uWq0Wer0eKpUKAGB5eRnRaBTpdBrZbBZ/5+/8HRiGgUajgRMnTuCv/uqvtucmamhoaGhoaGho3FK440na9PQ0HnvsMdx7771YWVnBL/7iL17i5viBD3wAJ06cwNLS0g0pT7zTVRHDMPDAAw9g3759sG0boVAIvV4PFy5cgOM4aDQaqFarCIVCQtAAwLIsABDFi6SMyphhGFLmqBK2druNaDSKaDSKSCQi7weDQfi+j36/j3g8jm63K7lmyWRyJCONmWitVuuSe6OqcvF4XEgkxxoMBoXwqT1sVOdItNrtNlzXheM4cBxHxsOSSJY2MoNtZWVFxrpnzx589KMfRbfbxenTp3H27FmsrKxs/c3U0NDQ0NDQ0NC45XBHk7RAIIAnn3wSu3fvxurqKr7yla/g2LENYyfTNHH33XejWCyiWCzCcZzrvub+/fuRTqdx/vz56z7XrYaJiQkUCgWk02ns3btXSJfneVhdXYXjOOj1epJXxrLD4XAoqlcikYBt26KMGYYhmWQsS+XPtMnvdDoSSM5j2PNGU5FQKDRi7kHSZxiG9ImpfWe+7484PVJNo4pHsk6HR16/1+tJLxz7FUkmScRc10W1WpXSx16vJ+6Uqvuk67ro9/tot9tIJpNIpVJIJBKYn59HKpXC/Pw8HMfBK6+8ckeTfg0NDQ0NDQ0NjVHcsSQtEolgZmYGH/zgB1GpVPDMM8/gD/7gD+T9aDSKfD6PI0eO4JOf/OQNc9x7+OGH0el08Ld/+7c35HzbDcMwRC07cOCA2Oezz6xaraJWq6FWq8HzPLHDJ3mKRCIIhUKSOZZIJJDNZqXHrNvtIpfLjRAlGoOwT4z3huSGZI0h1DQYoSEHXwc2XB45F7otso8tEAjIGGk4QiWNJiE0HPF9f6R/jUpaOByWufX7fViWdUkfWzQaRavVGnGk9H0fnueh1+uh2WyiXC4jl8shm81ienoaoVBIvkDodruiBGpoaGhoaGhoaNzZuCNJWiAQwO7du/Fbv/VbKBaL+I3f+A18/etfHznmyJEjmJ+fxx//8R/fUEv0AwcOYHl5GaurqzfsnNsJ0zTxQz/0Q5idnRX7+cXFRZw9exalUkmcEkmSqEyFQiFRhxgEXSgUJBg6GAyi3+9jMBggmUwKCfJ9XxweaaihRhnQpZEEigSNfWF8fTAYSOkkCV2n00Gj0RBiFgwGxbpfJV8smVTXQCV76nsklqojpGmaEswdiUTEybHT6Qgp63a74v4ZiUQAQIgvTUds20YymcRP/MRPoNvt4tVXX8WXvvSla3Ic1dDQ0NDQ0NDQuH1wx5G0YDCIj3zkI/iRH/kRLC8v4+Mf/ziKxeLIMY8//jgGgwGOHTt2wyzQDcPAQw89hEqlgsXFxRtyzu3Eu9/9bhw4cAC5XA6+76NYLGJtbQ2VSgW1Wg39fn/EGZOh0vF4HLZtI5PJiKMmCRGdFqmIsXSR/WUqBoOBlALW63UkEglRvphTRuWs0+mMEG2VMNFSPxqNAliPCCCRNE0TAKTskXNR89H6/T5isZiUUwYCAZimiWazKZ+jekeVzDAMxONxABsh2pOTk+h2u9KfRvdQz/PQarXQarXQbDbR7XZRrVYBAMViUchuLpfD/Pw8Pvaxj6Hb7eILX/gClpeX4Xneltx/DQ0NDQ0NDQ2N7cMdRdIikQg++tGP4k1vehNWV1fxP//n/0SxWBQDiWAwiKNHj6Jer6NSqaBUKt2wawcCAbzvfe/D3/zN3+DVV1+9Yee9mbBtG/fffz+mp6cxOzuLWCwG13VRKpWwvLwMx3HQbreFnJH4GIaBQqEgQdXJZBLJZHKkpJDh3iwrpDGIqnzx53q9Ds/z0G63pccLWFeaBoMBms2mqHAsGwwGg6JIUdWj4kWlT1W9aFzCc7AUku6OJI4sW6QBCAkmCRvPqapvKuHj9dXzc6wsJW2322i1WkgkEvA8D41GQ9REZrANBgM4jgPLshCLxfDwww+j1WqhVCrh29/+tjzjGhoaGhoaGhoatz/uCJJGJefRRx/FY489Btd18dWvfhV/8zd/I8fEYjEUCgWYponXXnsN1Wr1hvb3GIaBQ4cO4VOf+hSWlpZu2HlvBhi6PDk5ifvvvx+5XA6dTgfNZhOVSgUrKysoFosj9vcsWwyHw4hEIpiamkIsFoNlWWKf3263RxQp1ayDyhlJi6rMra2tCSFk0DXNOvr9PjzPGzkXSV80Gh0hSKrLpnoOEkQSRtM0Yds20um0KG8kasxTY99YrVYbOT+vR/D1zVxCSeRIABkjEIvFYJomTNOE53kIhULwPG+kPLJer8N1XTSbTViWJQYutm2jVquh1WphbW1Nh2NraGhoaGhoaNwBuO1JGjfZ8/Pz+IVf+AW89tpr+PznP4/PfOYzckwwGMTU1BQeeOABfOUrXxFl4kaOIRqN3nYOfIZhIBKJ4MiRIzhy5Ajm5uYQCoWwsLCAcrmMer0+ompRWQqFQojFYsjlckgmk7AsSxwaVYUK2CAtAIQMsdyv3W7DcRzEYjF0Oh24roter4czZ84IqQMgBhq8NklUPB4XIkWSpqp27GWjfT/DsWOxGBKJhBw/Pz+PPXv2YO/evVKSOY5qtYqVlRWcP38etVpNShVpZEIySdJIdY59duylU9deJYORSASWZaHVaknOG9ekWCwKYfU8D7VaDdVqVfLW3v/+96Ner+PLX/4yFhcXpZ9PQ0NDQ0NDQ0Pj9oRxI3LBrnsQhnHNg5icnMQ73vEO/MzP/AxOnDiBf/2v/zXOnTs3cszjjz8O27bx8ssv49SpU9c93nHs3r0bP/mTP4lvf/vbeOaZZ1Aul2/4NW40wuEw8vk8fuqnfgqmaaLb7cJxHKyurmJpaUkcBaPRKHq9npQtxmIxTE9Pw7IscWmMRqNSrkhlidbzdErsdruIRCKIx+NoNBqoVCpYWlpCo9GQ42nV77ouIpGI5KKlUikhZ3ydmWpqLxsVuXa7jVqtJseZpol0Oi2GIZOTk7jrrrvw6KOPjrhKXil830epVMLzzz+PEydOYG1tbcRNks6P46SdmW5q3xsdMDkPz/NEAaRRS7FYRK1Wg+M48DxPCBjX2zRN7Nq1C+l0Gu12GxcuXMCXvvSla3swdhB837+6G6+hoXHHI2lk/UeN793uYWhoaOwQfMP/SzT8yqb7kdtaSZufn8cP//AP495778WLL76If//v//0lpYbvec97EAgEsLCwgIWFhS0Zh2VZePDBB/Ff/+t/lXK4Wxlvf/vbsX//fjH3IFmq1+uoVqsj/WG+78O2bRQKBdi2LWWBNNCgUsZAa4Lkn/1cgUBA+q1YNlipVCTzjAHXiUQCrVZLyB+wrsCplvbxeHwkO20wGIgrIxUs27bFvINljul0Grt378a+ffuwa9euy6pmbwTDMJDNZvHYY49h7969+Na3voVXXnlFrk3Fjv1y/IxaBkmVjWvFPzRTUdVHrjVDsuv1OtrtthDDdruNkydPIpVKIZlMYnp6Gj/6oz+Kp556CuVyGd1u95rmqaGhoaGhoaGhsT24LUlaMBjE5OQkfuAHfgB79uzB6uoqvvSlL40oaNFoFIcOHUImk8Hp06extLS0JZvVZDKJbDaLVquFRqNxy5aZMTdufn4ehw8fRi6XQyAQQLFYxMrKChzHQavVguM4opixfDGXy2FqagqJREJ60FTTDABCSEhOGOzMvrFAIADHcTAcDhGNRjEzM4PJyUlkMhlZM56PeWuBQADdbnek34xlgiSE/J3EkpEAoVBISJHaS8drXK+CHAwGYds2DMMQoxqqgnSdVMsf1dJH/uEc1JJQlkbS8p+Ek/eC5aR0gux2u+j3+1JuyQDtQqGAI0eOoFqtolwu4+zZs9c1Xw0NDQ0NDQ0NjZuH246khcNhpNNpPPzww3jyySfx8ssv46//+q/xhS98QY6JRCLI5XI4evQoVldXceHCBaytrW3JeGZmZjA9PY0XXnjhujf+W4V4PI5CoYCjR4/ioYceAgC0Wi3UajUsLi5K3hmdC8PhsPSaJRIJTE9PI5/Pi1JFZ8NutytGGCRazADrdDqo1WoS9GyapoQ8T01NIZfLIZ1OI5fLodlswnVdOI6Dcrksoc90UyTZUjPPVCfFQCAg5IS9bOFwWIgLyYvjOFheXsZwOITnedizZw9s2xbSeS2wLAuHDh3C8ePHsbCwgGazKS6UJFsqMVSdJunIyDmopiMklGquXCQSEZIcDoclBoAgaaN5yPz8PObm5rC6uiomMLfqM6qhoaGhoaGhobGB246kzc7O4rHHHsPHP/5xPP300/iDP/gDvPDCCyPH7N69G4cOHUKr1cJTTz21peVejz76KPbv349f/uVfvmVt0B988EG8/e1vx8zMDMrlsuSdlUol1Ot1ITMARC2bmZkRB0GqOLSepyrkeR6CwSASiYS4NFLNounF5OQk5ufncfToUTz++OOX5KEBQKfTwdraGs6ePYtarQbDMCRTLBgMiusjA697vR6i0ehIuSDdEkniSBw5ll6vh2KxiHK5DN/3kUwmceTIEfzAD/wA7rrrLliWJeO52h61YDCIe+65B81mE81mE47jSHkjHTBJNll6CWzkualqH8tMqb6pZC0cDo/Y+rPs03VdVCoVcb9st9ti/jIxMYFdu3bhyJEj+PVf/3W02+3bzuBGQ0NDQ0NDQ2On4bYiaY8++ije97734a1vfSu+8IUv4L/8l/9ySQ/Ygw8+iGQyiVKphOeee27L+3Hi8TgSiYRked0qCIVCeOc734mjR4/CsiwMh0OcPHkSFy5cQLPZRKvVEsdAltCZpolCoYBsNotsNithz77vj8QVBAIBuK4rig9t5CuVCrrdLuLxON70pjfhIx/5yIjt/WYEDVgvTY3H4yMmIwx7JtlRDTfYq0ZCE4lE5D77vi9qH9W1Xq+HcrmMr3zlK1IOaRgGPvvZz+I//+f/LCWZ4XAY99xzD97//vfjHe94Bw4fPjxC3jZDr9dDtVrF17/+dZw8eRKlUgmu646MKxwOj4R5cy1834dpmuj1eiMKJQCJKWDZI9eajpaJREIcLNPptPTyMcfOcRwUi0XU63Wsra1hbm4OH/3oR7G0tISTJ0/ipZdeuo6nS0NDQ0NDQ0NDYytxW5C0QCCAN7/5zfjIRz4C27bxta99DX/0R3+EWq02YvV+6NAh5PN5lEolnD17dssJGksHT58+vaXXuVrMzMzg0UcfxV133YV4PA7XdVGv11EsFqWssN1uS3BzLBYTO/d8Pi/lf7Te7/V60hdFUBUaDAao1+sIBAKYnp7G4cOHMT8/L4HWV4pSqYRXXnkFq6urI8oeAFHHVEt+kkbVpAPYIGlqGDSJWTweR6VSkXP3ej14nifXoYK3traGv/iLvxBXydeD7/vodDpYWlqC4ziiAHLcapZbPB7H9PQ0CoUCMpmMkE7VCZKqIMdDgsvf6QYZDofFbIWumJZlSayBqkY6joNz585hMBggmUzinnvuQTqdHskR1NDQ0NDQ0NDQuHVwy5O0WCyGubk5vPvd78bExAROnz6Nv/qrv8Lx48flmFAoBNu2MTc3B8dxpJxvq/Hggw/Cdd1bhqSZponJyUkcPHgQ9957LyKRCEqlEsrlMmq1mhibsP+MZX+ZTAbZbBbpdBqpVEpID/vLut0ubNseKctTSVMkEsGBAwcwPT2Ne++9FzMzM1c17rNnz+KVV17ByZMnJYOMzo0sY1SNNgzDEKt6khjXdWXcDMZmXxoDqqempoR8djqdS0i87/uoVquoVqs4duzYDbgjo4jFYqjX6yiVSkgmkwiHwygUCpLZphqgsPeMpimRSEQy2WjIAmCkly4ej6Pdbgup6/V6aLVa6Ha7aDQaWFlZweTkJOLxOObn57G2tobl5WXJe9PQ0NDQ0NDQ0Lg1cEuTtFAohJmZGXzf930fnnzySXzmM5/BU089hW9/+9sjxyUSCczPz8M0TXzjG99AvV6/KeN74IEH8Bd/8Rf42te+dlOu93oIh8OYnZ3FE088gT179sAwDKytreH06dOoVqtot9tiVkGiFYlEMDU1Jc6NqoGGSoBoAd/r9US1oTplWRZ27dqFn/iJn7hkTCwr3AxUvNrtNr74xS/i2LFjEljN8kkqRRw3FbZ2uy0h2/1+X0oiSdRIHhOJBICNsPF9+/Yhm80KiaeRhvpnK9Fut3Hu3LkRF9LZ2Vns3bsXhUJBSiuZB8fogUwmI6RsOByiWq3KvNgfmEwm0e/3Zd1YTlmtVsWyv1arwfM8JJNJTE5O4oMf/CC+8IUvYGlp6YYHvGtoaGhoaGhoaFw7bmmS9j3f8z1497vfjccffxyf+MQn8Od//ucoFosjx9i2jV27duFNb3oTfud3fuemutfRZGO7kUwm8Z73vAdHjx4FADGSOHv2rIRSD4dDRCIReJ6HaDSKbDaLmZkZ7Nq1C6ZpyoafuWO0e08kEmKpz76pZrOJTqeDj33sYzh06BAikcim42o0GrAsa9NetGq1iueeew6/+Zu/KWQqFotJMDbLAEm4a7WaqEKxWAzValXUwF6vJ6YZwEbJ4HjJZCwWQy6XQyqVQqfTkcgBqme00L+ZWFxcxOLiIoB1A5KjR49KJh37Ag8fPoz9+/dLPh0AOI6DEydO4OWXX8bS0pKQZ/arxeNxxGIx+VOv19FsNuF5HjzPk/V797vfjdXVVZw8efKSLz80NDQ0NDQ0NDS2B7csSfuxH/sxfM/3fA9M08Rv/MZv4E//9E8v2UTfc889mJ+fR6/Xw5/92Z/dNIJmmiY+8pGP4Nlnn932/KmDBw/igQcewJEjR9DtdlEqlVCpVLC2tiYGFgCEwNi2jXw+j4mJCUxPT0t/E1WvaDQKz/NkLdnfxVyuSCSC6elp/ORP/iQmJiYuS9B4rXFre8dx8PTTT+PFF19EsVhEMBhEPB4f6aHqdDrwPA+9Xg+u6yIUColy1uv1kEwmUavVpKyPZYC+7wshpZ29mkFGwkJ1MBwOY3p6GplMZoS8lEolMR25mRgMBjh27BhCoRAmJiZw4MABHDlyBDMzM0gkEiNrmUgkcM8992Dfvn3wPA+f+tSn0Gg0ZM5cUxrCWJaF1dVV6Ztrt9s4deoUut0uJiYm8PDDDyOfz+OLX/yidn/U0NDQ0NDQ0Nhm3HIkLRaL4W1vexsee+wxtFotfPe738VXv/pVKfECNsog77//frFuv1kljsB6aeFb3vIW/P7v/z4uXLhw0647jqNHj+LgwYOYm5vDYDAQW/1qtQrXdcUunwYUAJDP5zE9PY10Oi2B1QBGSBnt9Wlf3+/34TgO8vk89u7di0ceeQSzs7NvmC2m2sV7nicljceOHZNw8VgsJmWU3W5XSAQNP7rdLoLBIFqtloxF7aljOWWn0xkZD8siVZJGqOSOP7PvKxQKwbIsuf5wOESr1RLy2G63t/TLAGacMevttddeA3BpLIBhGEI6bdvGW97yFrz88suoVCoy7nA4LCopw8FJ3nu9HtrtNiqViuTiTU1N4S1veQtefPFFNJvNLZujhoaGhoaGhobG6+OWImmJRAJ79uzBBz/4QSSTSTz11FN45plncOLECTkmGAzCtm0cOnQI6XQazz333E037giFQti3bx9OnTqFUql0U68NrG/QJycn8Za3vAWFQgG+76NUKmF1dVUIGskNe7YikQgikQgKhQJyuZwoLSwLZI8YyY3ruuLuyN60PXv24JFHHsHb3/72Kxpno9EQwlWr1fDMM8+MuCAC6/ez2WyK2QfVIBp+MCuNQdu+74sqyLmFQiFR00hKSVBVp0jVbINz5rwZFm1ZFnK5nBiQDAYD1Go19Pt9eJ4niqL6Wf5+I+G6Li5cuIDvfOc7qNVqSKfTolq2Wq2RWINgMIgHH3wQAHDmzBmsra3B8zyJK6DzI81HGP7Nks/V1VW0Wi3k83k89NBDqNVqWFhYgOM4N3ROGhoaGhoaGhoaV4ZbhqQZhoGHH34YP/IjP4L77rsPv/RLv4TvfOc7l/SgpdNp7N27F/fffz9+7dd+bcRC/WaO9Wb2vo0jGo3ip3/6pxGNRtFqtVAul7GwsCCEkSYbpmmKWpVMJpHL5ZDP5xEKhSR8mn1MNPFwXRerq6tivQ+sW/ofPXoUTzzxBO699943HB/Jy9/8zd/g7NmzKJfLaDabMAxDYhPoPDgYDET1Gw6HQi5Up8ZQKIRwOCzKHoljt9sV50cAIzlj/X4f0WhUrPg5RwZjc5xUGUn+SO5CoZCURjLQm+eu1Wqi6jUaDSwsLIzkyN0ouK6Lp59+Gl/72teQTCZhWRb6/T5eeeUVRKNRybRjX+SDDz6IiYkJPPvssxKLwDVhf2G320UoFJKeNPYY1mo1nD9/Ho888ggef/xxHDt2DF/5yle29TnX0NDQ0NDQ0NipuGVI2sc+9jE88sgjiEaj+Nmf/Vl897vfHdlQA8D09DQOHTqEqakp/Oqv/uqW56BthoMHD+Kd73wn/tf/+l/bEmB911134Qd/8AdhmiYqlQrK5bKUOFJhYjlgp9ORgOqpqSnpQVPdDAOBANrtNsrlMsrlMiqVCgaDAVKpFFKpFGKxGI4cOYK3vvWtb2it3+128bWvfQ0vvfQSarUaisWiGIG0Wi1Uq1U0m02x8md5I7CuToZCISFUzP9S1zgSicAwDDiOg3Q6LSHXJHYAxL2S12XpH8mYGorNcVF5Y4mh53kIh8MyJqpsajlkPp8X05LZ2VkZF3vayuXyDbnfvu/jZ3/2Z2HbNj70oQ8hHo9jcnISxWIRCwsLWFlZwdTUFCYmJmAYBmZnZ5FOp9HtdnH27FkxUaHj49TUlASHk6CFQiEYhoFWq4VXXnkFu3fvxuHDh3H48GH84R/+ocQiaGhoaGhoaGho3BzcEiRtYmICjzzyCM6dO4e//Mu/xCuvvIJOpzOyMcxkMti3bx+azSZOnDixLQSNY33ggQfwb/7Nv9kS9eT18Oijj+Lee+9FPB5HrVbD6uoqKpUK6vW69GoFAgEYhiEkgyYhLHGkkgJAlCqSPZYn5nI55HI5mKYpVvDtdhsrKytoNpvSv9VqteC6LqrVKo4fPz4SmB0OhyVYudvtwnVd1Go19Ho9UegYRA1AXAnD4TBs2xaCRXt91cSDxh/AhhpG90qSNPXZoaW/Gi/Az1I5AzbCrIHRUkrf9yXHjOewLEsMV4B1k5REIoFUKoVCoYBmswnHcVAul+E4znXlkLXbbfzKr/wKFhYW8PM///MoFAoolUqylmtra7Ju4XAY8Xgcb3/72+F5HlZWViQXjuOn4yZLUTudjpREdrtdLC0todPpIJ/P4/u///vx9a9/HSsrK9vypYSGhoaGhoaGxk7ELUHSLMvChQsX8I1vfAPf+ta3RjaDLNW666670O/3sba2Jpbl24F4PI5sNnvTDUMOHjyIw4cPY3p6Gp1OB6urq1JGyJI+kg+SC8uypCSOPUk8Ftgw9GCfFbB+L1KpFOLxuOSm9Xo9NBoNuS/hcBiBQADlchn1eh3VahWnT59Gr9eT8rpwOCxj8zwPjuOg1WqJ+sUSRvZWmaYppNC2bTECMQxDFFWSKZYu8jW19HEz8Bp0PiTYv8ZjmMdG4sWyVjo9qiSN4wDWSR/dF03TRDQaRTweF9LZaDTEOfJa8Z3vfAfxeBz33HMPPvShDyGbzQKAEOVGo4FIJCJlmdPT09i7dy+63S5WVlZknGpIeKFQQKPRQL1el/vf7XZF7QwEAti1axcOHz6MYDAo/YQaGhoaGhp3KoxwBAEzJr8PWy341/FFq4bGteKWIGkA8Lu/+7t45ZVXLimrikajmJycxKFDh/D5z39+W4w6CJIWOvDdDNCZ733vex+SySR6vR7K5TIWFxdFoeHmG9jYiIfDYRQKBRQKBSSTSUSjUSEiqgrJgONAICB9T1SJSKjq9brYtzebTfR6PTiOI69RxWKOV6/XQ71eR6lUQr/fR7vdRrvdllJDlWwB6yoUM9JILlXVq9vtynh83xey6fu+lCRyTiylZF+bSsR4fT5j7MdTXSLHQ62p6HG8HLMaJN3v94UgqSWStm2Leuk4DhzHkfFdi839N7/5TXz84x/Hu971LszNzSESiWB5eVkUS963aDQKALjvvvvkPZVocvzZbFZ6OqvVqhDdTqcjZCwYDOK+++6Te3Py5ElZDw0NDQ0NjTsKhoFAygayaXkpsFbCoN4AdNm/xk3GLUHSzpw5s+nr8Xgce/fuxeOPP45f//Vf3/b8pscffxyTk5P40z/905t2zVQqhR/90R9FPp9Hs9lEsVjE4uIiKpWKkEUqV9FoFL1eD7FYDHNzc9i7dy9SqZRkiNFwA1hXYGq1GtrttvQoRSIRWeNms4lAIIBIJILBYIAXXnhBVLe1tTUMh0Mkk0mEw2GEw2EkEgl4ngfXdSUkulQqST9XLBYbCZsGICV4iURCTDri8bgoVoFAQFwmE4mEkEa1p2w4HKLX642ULpKIABDHR75HAqc6PZK4UQkERgkZCS4AOUYtXwwGg0Jq1N64WCwmvWC2bePNb34zOp0OSqUSVlZWrprs9/t9LC8v42d+5mfwy7/8yxKlwH8/1WoVw+EQExMTANZNdnbv3o1Wq4WzZ8+iVqvJ+oRCIXieh9nZWVH+FhcX4bqurJfjODh16hQMw8DMzAwmJiaQz+fx1a9+VfeoaWhoaGjccQjtnoMfHct/ncgjlEigv7i0PYPS2LG4JUjaZgiHw3jiiSeQSqXwu7/7u9tO0ADgkUcegeM4+OM//uObcr3Z2VkcOXIEExMTaLVaOH/+PFZWVmQzDkDUJZo/JJNJ5PN5zMzMwDRNITIkIXRQdBwHrusim80ikUgA2FCh+v2+lCICwKlTp1Cv1yWrbDgcwrZtpNNphEIhtNttNJtNlMtlIYyZTAamaSIcDsOyLCFxgUAAjuNI3hiVnGAwKL1utNsfDodC7jguuiqSJJFgqWoYSVQkEhnpMWMpINU8ElCCpiUkZyR+NFvh+Xkevk8DEpZV8noAJAqB5IiqYTabxdraGpaWlq7q2e52u/jTP/1THDlyBO973/tw7733Yv/+/Wi1WtI7p6JQKMB1XZw5c2bEoVI1WMlkMmKqcv78eYTDYelxHAwGOH/+PHq9HvL5PB555BEh7br0UUNDQ0PjdkIwmYSRy4y+6LXRX1ndngFpaLwOblmS9v73vx+dTgcvvPACarXadg8HwHpZnupIuJWIxWKYnp7GXXfdBcMwUC6XxR1R7b+i6mQYBkzTRD6fRyaTQSwWk1I8Kkgse+Pmm+V4JDMsIWRpoZqj1Wg05NrsGwMgtvaNRgMAYJomEokE4vE4gPUS0Xg8Dsuy4DiOuCeyR21lZUWcFVmiqPbWkTx4nncJOVOhlityXai68ZwsQ2Q+HHvZVOWN0QQsZWT5KFW/SCQipijMVxsPmmauGwkgCR774hgeTUOUer2Odrt9RWTN9300Gg38yZ/8ieS67d27V8ZLhZCwLEuIeLValWeFa0u1NJFIYGJiAq7rwnVdIfW8HnvzAoEADh06hE6ng3Pnzl0SkaGhoaGhoXGrwAhHEEinNn6PReGHR7e+BmII5nMYlMrwq3UYSRt+PDZ+Kg2Nm45bjqSFQiEp1Xvqqafw0ksvbfeQAAC5XE76q24GpqamMDc3h0KhAM/zxH2Rm3mW7LEPyjAMpNNpZDIZKXGkikbVp9vtyue5MWefnVrCR3WMbo10bOTmnb1nDHEm8bNtG6lUSvraeG6WU7KksdvtIpVKoVarodFoiGpDtUyNCOCYVCdN9poBGCm7Ixli6afjOCPljMBGoLX6eRI5AHJ9Epnxa9EchESXhE81FlEVPb7P+fR6PTFNsSxLxlSv14WIXgm+8Y1vIJvNIp1OSyTAOFkEIIQ6nU7j7NmzYuzC50c1bslms3BdFysrK6I8Ur1Un4977rkHBw4cAIARgq6hoaGhoXErwIhG1/8/bppAJimvb1ao74dDQCaFgONi0GwiGA4BmqRp3AK4pUiaYRjIZDL4Z//sn+Ff/st/iXq9vt1DArA+rve+971YWVnBiRMnbsr13va2t2HXrl0YDAaSvUWiEovFRJFhyZ1t25ienkYymRRyxA22Gs7Mz5I80MadKgyzySqVCs6fP4+TJ0+i2WwCWC9BTaVSmJqaQiqVEjXJMAwUCgVMTEwIGbMsS4gJ1adkMinKHklKt9vFuXPnRGVjXx0JZa1WE7JAuK4rhE91ZAyHwzBNc0QtC4fDI8fQGZIkRCVWLHXkZ1Rli0Ss3W5LLx5JJcdLpYzXInHjdQBIXyAz1zKZDOLxuJizXI0xzuc+9zkcO3YM3W4XP/7jP45YbPP/qYTDYczMzMgXHqq5Csko+wFnZ2eFlFElZelru93G4uKiHMfIgaeeekr3qGloaGho3DIIzk7DDwXf+EDCMGDsnkVgYfvcwzU0xnFLkbSHHnoIb33rW/GP//E/3rYctM1gGAY+8IEP4D/9p/+E559/fkuvFQwG8YM/+IPYtWsXhsMhisUiTp06hWazOaIyAesbfiolu3fvRjablTJDz/MwHA5hWdZIOWE0GkW32xUC0m63hbhRETtx4gROnDiBRqMhx0WjUezduxe7du0S+3eajuTzeWSzWWQymRGXyWQyiXq9LsrUYDCQvDIqOvv370e1WoXjOKhUKiNmJcz3Yn8Zz2uaJrrdrpALkjOqhXyd7oUkeLTzV9Uvrg1LItVyQTprqv1ovV4PlUpFyB0JqOM4iEajci4SSI652+2OqHSDwUCIXTAYRDqdhmmayOVyWFpaQqvVGumXuxzOnz+Pf/SP/hE++9nP4hd/8Rfx4IMPynu1Wg2O44hiSVLM+2+aJoCNvj4+LzMzM4jFYigWi9Jrx75GwzBw6tQpVCoVzM/P48EHH0Sn08E3vvEN7fqooaGhobGtMEIhBPbuhr9JZckVfX7XzA0ekYbGteOWIWl/9+/+XfR6PXz605++pQgaACklBHBFG+drhWma2LVrF/bt24fhcIhyuYylpSW4ritkQlVoBoOBKFTZbFaUFKpiqqEIN9AMiOb7LIlstVooFou4cOECXnvtNbHWpxnJ3r17xc6f/WBUzlKpFFKplARo8zqquUm/34dpmkIe1Tnv3r0bS0tLEtDMXjKCRiaqsQlJHskWiRgVMhIQEjEAI2WNtNYPh8NyDOerHq+C5JKEa/wYqoQ8P+8FyVmn0xFjD7WUE9gI3CbpW11dvaJ+L573a1/7Gn7u534ODz30EN72trdhfn4e7XYb3W4XrVYLlUpF+tDGTWdI+tWgbq41Q7lJYjn/crksRi/3338/HMfBa6+9dsuo3xoaGhoaOwfBQgFGNAIYxjUTNADAxc8a/QGGK2vy8nAL934aGpfDLUHSLMtCMBjEwsICzp49u93DGUEkEsHMzIz0Tm0VgsEgMpkMjhw5gkQigWKxiHK5jHK5LJtobpxJBGKxmPSBMceKm26SF2A004tlgQCEbFSrVSwuLuL8+fMoFouy0aY9/uzsLCYnJ8VkhLb7yWRSyhtZYkmQ7KjkisRSLbULhUJIp9NwXRe1Wk2ClVWXxXEHRs5vnESppiPMUgM2yBnJLckp+8NIgHkf1L85F6pNqjqmmpKQKKrvM9KA41eVs3Ebf9VxMp1OSwklTTveCKVSCX/7t3+LxcVFrK6u4ujRo7AsS+bruu6ImYlK1tQx0PGRZCyTyUj5I+dK1bVarSIajWJ+fh4HDx4UYs7yWA0NDQ0NjS1DIIigte5ObVjxSwxBrhdGKISB7rnW2EbcEiQtk8ng2WefxalTp7Z7KJfAsiw89thjePnll7d082lZFubm5vDAAw8IcSqVSqjX68hkMmKZDkDITSqVQqFQQDabFcJCQkGQ1PA9AKIedbtduK6L8+fP48yZM1ha2sgACQQCsCwLU1NT2L9/v+R9BYNBxONxJBIJZLNZRC82524WCM3yRtWVke+pBIG9TclkEktLS+j1epfMQQV72ng+YMPshOTJNE00Gg0hKezPY7+ZqnSpShKJoaoIqgSN5FHt5+JnVbfHfr8vQdZUBkl0eG2uSzwel3Hz/uRyOYRCIZw5c0biCt4Inufh5MmTWFxcxNmzZ3HgwAHkcjkkEgnp6+NYOD+Oiz+TTLOvMJ/PS2g6/3BNWYIbCARw1113CZk7efLklirOGhoaGho7HIaBgBkDpgoANjcEuR74oeD6uR0XGOr/n2lsD24Jknb+/PntHsJlkc1m8aEPfQgf//jHUalUtuw6jz76KB544AEYhoHV1VUsLy/DcRxRV1QHRG6i5+bmLilzVP8YhjFSOqoSG0YJPPPMM7hw4cIlKuHk5CRmZ2cxNzcnZhx0AZyamkI8HheCBqz3P7FXjb1dakkiyQlJjEomY7EYMpmMGFMwdFktxyTRokU+yRLnqtrJm6YpawZslPHxNTUrTA3ApvEH+/pIDkkAY7GYzJsB3HS5XFpaEuXOMAy02220Wi0pz+Q8+Tvt/Nvt9khAN9U9OmJaloWXXnrpkjLRy4Eh1M899xyazSZ27dqFqakpcYHk/efzwPJXvs4y2GAwKP2MrVZLjFy4Xlwfz/Nw+vRpAMDMzAxSqRT6/f4t+YWLhoaGhsadgeBEAUjZ2z0MDY0txS1B0m5lhEIhFAoFURO2Au9///tx4MABRKNRVCoVrKysoFQqwfM8IQ4ARkoEJycnxRmQZIXByWrIskpCSApqtRqWl5fx0ksvbUrQdu/ejenpacnXikQiSKVS4g5pmqYQFJVoqQSIhEM1kxi3y2fpHJ0UaWZRq9VGCBrL72h7H4/HRfWhEQYDmOPxuFjkq+PZrGSSZIwEjWWbpmmO5KSRpNKRsd1uwzRNpFIpWdPBYIDV1VWJJaC7JteC81TLQn3fl5JEmpokEomREO1sNou7774b58+fl1LQN4Lv+3AcB+fPnxdyTCKskniSfpV4cX34vNAAJpVKYTAYwHVdIXacl+u6eO2119Dv9zExMYEnn3wSp0+f1o6PGhoaGho3DoEgQvNzAAA/GHiDgzU0bn9okvY6ICmpVCpbUr5lGAbi8Th2794tQdnFYhGrq6sjqgwDpLnpV0sNSdwuFwRNVUTtDVtYWMDp06dx/vz5S/qi4vE4stksbNsWm3v2vZmmKb1vPJ6be5JJEjeqRez34nioFLHni31cHCffU8kb14BKHM/Dc5KkqqSReXZcDxIfOjBy/MC6wkZyxr47KoHNZlOy3zhHEiUqY4wrqFar4gjJewBA+s82MyIJhUIol8tizhEOhzE9PS1h4cB6Kezk5CRCoRCWl5ev6NliCDXX2bZtZLNZWQ+uMck9+/VIRKl4hsNhJBIJJJPJkbgGzoskj8YzgUAAc3NzeM973oNnnnlG96dpaGhoaFw3ArEYAunUDe8709C4laGf9tdBoVDA5OQkTp48uSWqQCgUwq5du5BOp2EYBlqtFsrlMmq1mrgUsr+Jm+dQKCRh0arNu6qQkKzwffaSUfU4d+4czp07N+LEFwgEEIvFkM1mLzEESSQSolBRCVJJJAmWanUPrBMBEhrVsIO5WySV44YohNpv5jgOAIgVPx0fOVcqTzyvGjqu9uVxHVTCFIvFEI1GpZyRa93pdFAulxGLxUYCvGOxmMzD8zxUq1WsrKyg2WwKCWQPnzqX8XLFWCwGx3FQLBZHAqF5fWaxARD7/FqtBs/zruj56nQ6UjqaSqUwOzsrpi+qkQzz0KjEqmWk4XBYDGpYIquqefzyYjgcwnEcKdV805vehKWlJSwsLOiwaw0NDQ2Na4YRjsCwbfhJa7uHoqFxU6FJ2uvggQcewIMPPojPfe5zW1LqaJomvvd7vxfRaBS1Wk026zQJUUvO+v0+YrEYEomEmEGoyhFLAdWeL5YJUu3wfR8nTpzAuXPnUK1WLxlLJpPBzMyMELN0Oi2lbtFoVEgHyRSJUigUGrHc5xgASG6Zqn6RIPR6PVEISRZUcwpCVelIWlVCRwJmWZYoXaoTIbBOImKxmBAU1WjFtm1Rz7jmHFe9XhfCpxqnkIByLL1eT4wzgsHgSHA3r6Per2AwiEQigWefffaSfDE6nDLTznEcOWb37t04ceLEFX9p0G63USwWMRwOMTU1hZmZGUSjUUSjUSFmvCetVgvxeHyk95FrkUqlAKybkzQaDZkLSzgjkYiYiZw7dw6WZeE973kPvvWtb+Fv//Zvr6ifTkNDQ0NDYxzByQL8eOzmX1iX7GtsMzRJex1MTU1hYmJiyzaZgUAAmUwGvV4Py8vLOH/+PMrl8kjf0mAwkNwq27aRy+WQyWRkY00CRrMOOheSFFA5arfbaDabePrpp0dUJmCdBE1PT2PPnj1S5shwauafsbSPpXGxWExK30jeTNMUe/ZgMCg5ZCRgJCbsL2NZJMfcarVQr9elPJNErdlsjrgyqu6EwLoipRJHllSS7KolmFQteb10Oo1IJCLEhfb3NEXJZDL43Oc+h9nZWczMzGB2dhb5fB6GYaDRaKBcLsPzPIkoANbJFUsLOc5Wq4VwOIx0Oo1QKIRut4sLFy5c9rmq1+tYWlpCLBbDuXPn4HkebNvG/Pw83va2t+GFF16A67pX9Fz2+32sra3hi1/8It71rnfhrrvuQjAYhOu6UtJrWRZM0xwJGlf/ULntdrsolUqibNKZkuvOEtFoNIpDhw7hnnvugW3b+NznPnfF/y40NDQ0NDS2E0avj/5r57Z7GBo7HJqkvQ6ojmwFQZuamsLhw4cRiUSwuroKz/PEbZC9ZqpdO3vFaFahWtlTDWq1WiO9aCQo3W4Xq6ur+PKXvyxliiomJiaEjCWTSSSTSViWJaV/qk27WoLJzXm32x2xb6fC4nkeTNMUhcz3fUQiEXE9JNkiYarVauh0OkIcVBJHQsoxAJA+LypgVLbYX8VST1U5U4Ol+/0+2u02bNuW8k6uNVWzXbt2YWVlRUggyRufDbpN0lCF10wmk+KQGAqF0Gq1JFw8EAjAdV0sLCxcttex2WxiOBxiz5494gDZ6XTQbDYxOzuLgwcPolgsolKpCGF6PbBH7cUXX0S/38c999wjSivXx7ZtuRbvLe8p712hUMDU1BTOnDkj76sRD7w3nudhZWUFuVxOAtrPnj2rFTUNDQ0NjStGaG4WfixyU69pNF0Mq7Wbek0Njc2gSdplsGfPHgC4YqOGq0UqlcL8/DwAoNFoiMU6e64AjJSU0dSCfWHjAcz8w9dJzobDIUqlEs6dOzeSg0ZEo1FMTEwgnU4LUWHPG8kiyw3VDTaJIADpB+McVFKkjlMloMBGH53runAcB67rotPpiCLHYwjVFEUlqSpp4/HjhFdV/MaNQ1T7/lAoJKWdw+EQ+XweBw8ehGEYSKfTME1T5kLFif1s0WhUzpdIJGCaJizLktw2EpperwfHcV63t4zkl313JMbNZhOBQAC5XG7EMKbVal3+YbsIOlDSeGZ+fl5UTpXcq1BDuxlunslkEIvF5Jp83tTYBRqJMKT93nvvFbdJDQ0NDQ2NyyEQi8GIRQEjAN+MAmP96lsNvz/AcKziSENjO6BJ2mXw6KOPIhAI4OWXX96S85umiVwuJxtvbthN0xT1jr0+LBOMRqNSBqg68qluilQ3qPa0Wi0sLCzg5MmTl4zBMAwkk0nk83mk02nJ/iIhZKkeCQuwYYChlh+yJJMOhcAG8Rm3a+fPqsV9pVJBrVaTPC41W0yFajZCdY1ujwBGrk3CRpJJIxbOm6SW2XEAxIafhCMcDiMSieC+++6Tz3FdufbRaFQIIQkc+9+y2aysIa/T7/evOM6BfYXxeFxKRBkgzRgCzo3Pwxv1qjWbTSwuLsKyLExPT4s5CUtXN/v8+D1lGayqtlGx5N/qHA3DwJEjR/DMM8+gWq1uWZSFhoaGhsbtD8O2gUxyu4ehobHt0EETl8E73/lOVCoV/Pmf//mWnD8ajcKyLOnDYtkcc7JIGGh/nk6npRyPDoM0wqBqBmyoauwROnHiBF599dVNVbRgMIjZ2VmYpgnTNGHbtvRb8fzAej8Vx0cVj/bsakC26izJjb8KlmR6njdCCpaXl7G0tCROhI7jSNkdySfJaCAQkJwytZSQKhZJHAltNBodMQVRrfxZwsieu3A4LMHUtMX3PE8+TxWMaxGPxyWegD18tm3LWFhySbWNJhu9Xg+1Wu0NnxG6cU5OTiIWi40oh8yqM00TyWQS09PTUm75RiiVSnj++eexuroqSpnv+yNqLlWxer2OTqczolAWCgXs27cPhUJBLPzV8PF+v49WqyXOlefPn0elUsGP/diPYd++fW84Pg0NDQ0NDQ2NnQ6tpF0GDIreCtx///24++67pR+t2+2K4hQMBmWjzxK8fD6PVCo14g5IqEoGCQ0Vjkqlgm9/+9uoVCqXjCEcDiOVSiGRSEh5YyqVkjwsKlHsRyLpCoVCSCaTaLfbYtihOhaqbo/8WY0HYDnccDiUvqULFy6g0WhgOBwikUhI3xx7x6hG0emS5ZFcj3A4jHa7LYRNJXZqTAEVLqqSvV5PwrqpUnE92I9Hi33btoWQdjodxONx6ZdTVTvVsIX5drTTb7VaCIVCKJVKOH369Bs+J51OB8ePH8fb3vY2uK4rtv+dTkdIe6fTgWVZoqKx146q5OWUtX6/j5MnT4qCqiqkXLder4dqtYpYLCaxDFxvKrCe56HdbksWHjPW1EB1z/OwsLCAgwcP4k1vehNM08QLL7xwhf9aNDQ0NDR2CkK75m56D5oKf3EFwytoH9DQuBnQJG0TUOHazGTjehEMBrF//35MT0+LxbtqTc+NODfKkUgElmWNmIUAG6V3/JkqDwA57ze/+U3U6/VNzSnoZMgeNBIXVbEiaaRyRKLGzTh7rFRHQACSuxUOh2VMLMvjuTudDpaXl3H27Fk0Gg10Oh0YhoFEIiGq4GAwEDIIbKhkJKQ0NeHvHC/HRNUP2OhDC4fDiMfj4krILDDOOxqNwjRNcaXs9/solUojDpJUGVXCyVLK8f42KktUo3q9HjzPg+u6b/isUKViXxfNU0iIqGaqSqEav8BIAM6b94/9ZY7joNlsjpR2Un1UM/kajYasD9fUsixks1khsaojJMfOcwBAuVzG2toacrkcXNfVJE1DQ0ND41KEgje9Bw3Aut1+sYyh19bW+xq3DDRJG0MwGMRdd92FxcXFS7LErheGYWB+fh6zs7OIx+NwXVdc/EgABoOB9INFIhEJlCYRU8/F8RqGgXa7LQpYq9XCysoKTp06tSlBY3A1jTBYNqeSHn6OpXDhcFjKD/v9PqLR6CUkjUSFrpTRaFR67TjGfr8Pz/NQq9WwuLiIlZUVMaBQywN5HgBCStTx85zjxIWvq2V8PDd7yFQnxmazKWvNHi+uN3vP1tbWpL+MBEglJMyLI/nhcZ1OB9VqFZ7niZ2/mq92JSCxSiaToqLx2mpfHcc+nvWmkjQqbSRU3W4XjUZDSBnLPlkq6vs+crmcqGUAkEwmhazZto1MJoNEIjESjM57oKpqjuOgVCohnU4jm81ienp6y0x5NDQ0NDRuPwTi8e0haMMhDK+Dfq3+xsdqaNxEaJI2hmg0ig996EP49re/fUUlaVeDSCSCD3/4w8hkMvA8D+VyGbVabcSpkGOgWkNXQOaKqaBjoWrX32q1sLa2htOnT1/W3j0ajSKVSqFQKCCTyciGnooV+8V4DWDD5Y8KGhUi1UxjfK4st6OSRAJZrVaxtraGhYUFMUkhsQA2iCHJpOrCyNdVAkcTDxUseSRZIGGwLEuICEkaSQd7yVRFst/vY2FhARMTE4jFYrBtW4gW1z4UCom5R6fTQbvdRq1Ww9LSEl5++WWsra0hEong0UcfRaFQAADpf7sS1Ot1ZLNZWJYlhFB1rGSv4LgDJ9VXlkJSIeRx3W4X1WpVehKp2qqh3+yFu3DhAiqVipA5El72S7Ick/eJx7A3EgCKxaIocO9973vxO7/zO9rtUUNDQ0MDCARhzE3jpmpYrEzyOugvXtq3r6Gx3dAkbQyhUAgPPvggfuu3fgsXLly4YeednJzEe97zHmSzWXS7XVQqFSwtLUlmGEsBk8kkotEobNse6Y0iqJTYti1KidoXVKlUcPbs2dd1pcxkMsjn80gmk5iYmBDrfaplqmtgvV4XsqOGQ0ejUVFXAoGAhC23220Mh0NYloVisSjlc57n4cyZMygWi2g2m2g2m2g0GnJdtX9NdVlstVoj1wkEAojH4/A8T0gBVTtgQ3Wj0kSiS+VMLXukkkglrt1uiyLI+TuOg5MnT2J2dhaJREJ6BmkoQtLC0j/mw509exbnzp1DpVKRPLVarSZkK5fLXbGSpMYJqISaBKfb7cozQIVO7RmkmsX1Ya9YMBhEtVqFZVmS62bbNhKJBGzbhmmaaDQaiMfjMAwD586dw5kzZ3DkyJERojYxMYFOpyPB3nQNJQklCfc8D4uLixgMBpifn8fDDz+MF1988XWjCDQ0NDQ0NLYCw9cW4Gu3YY1bGJqkbQLTNN/QyvxqEI1Gkc/nsX//fnQ6HZw/fx7VahXtdhuGYchmlo6PpmmO9IdRsWGZWiaTwcGDBzEcDlGr1bC6ugoAQtCWlpYuO/5EIiE9WWpotWqUoZYcUrlS+6C63S6SyeQlJZckCGpum6qCRaNRUX9IcILBoJiDDIdDOI6DeDwujoFUfEgc+TcVMo6RaxMOh2GaphA69tpRKaN5CI1DWPJH9TAQCIyoSRMTE/jIRz6CyclJRCIRsbunQkhVyvM8IY6hUAi5XA6JRAIHDhwQUpROpxGJRJDJZDA/P3/FJI3zYjkpCSkJWCKRQKlUkrGoqifXB9hQBtlDR1LMsVMpzOVykpPn+74YixiGgXK5jMXFRUxOTgq5np2dRb1eF3dIlu+qOX68d1QZs9ksHnjgAZw6dUqTNA0NDY0djEA8jkAhd3NVNAD+UPeeadza0CRNAcvW2Cd2ozA9PY19+/YhFotheXkZa2trUh5GsxCSGgYGx2IxISMkQaFQCJlMBocPH8a+fftkU7y8vIx+v49arYZSqYRGo3HZsdi2jXQ6LeHVpmnKZp6ba5IBYNSghH1Kao4ZiVQ4HBZli0ROPR5YJ7/MQVOJmmrxzuNJNFSyCGDkc1T9WP4HYMS9cTgcIh6PC1m0bRuGYUjPH+dGEtrtdhEKhdBut4XkJBIJ7N+/X153HAfBYHCEzLL/jD1sgUAAqVQKmUxG1pX3GgDi8TjS6TQKhQJKpdIbfiGgrg3JEnPfuO4s6VTVTrW3kPeJa8jjqcS6rislonS4pJsljWMMw0CxWMTKygrq9bqobfxyoVKpjGTgqf2AwEYAuOu6qFar2LdvH/bt2yfkT0NDQ0Nj58EIheBHt8/R8WYieORurHxP9qo/Fy8Okfjjb2zBiDRuZVwXSTMM4yyAJoABgL7v+w8bhpEF8EcA9gA4C+Dv+r5/Yx04tgjxeByTk5M4c+bMJT1W1wrDMLB3714cOnRILM0bjQba7faI0qQaTsTjcTGnUIlAJpPBXXfdhaNHjyKXy4mKxj6oUqmEer0u5YGbgb1oLPVj3xI39xwLN/eqa6Fa9kgDC6pYkUhEyA0AeJ4nRhfcnMdiMXieJxt4ki21x0ztLWOvGEsoVaMM1ZSE1vBUwWhm0e/3RVUbDAZIpVLSd8e5sQ9LDWMmqfJ9fyQ/zXVdsfpvNptyb0g0B4OBzINGHzwXw6hVU5WZmZkrCnfmWNRSRfYock2oFjIO4HKh1Go0AUkdSRodJ1XDmmQyCcdxkE6nYds2ut0uPve5z4nhCkkwHUjVwHGuMdeUbqntdhvlchkzMzM4cuQIut2uJmkaGhrXhTttP7JjEAgCyhfDdypC+/bADwaw/K4c3vmT37zqz3/+9GEkv7ueM+qfX8LwdfZ5GncOboSS9rjv+yXl958H8Je+7/+yYRg/f/H3f34DrrPlmJubwxNPPIFPfepTcBznhpyTxC+Xy6Fer6NSqUhp3XA4RLvdFgt8EgqWOqqOhbZt421vexvuvvtuyW9T+6ZarRbOnTv3htbumUwGuVwO2WwWiURiROFh6RtfByDEir1bJAFUV0gcAIgroupWCUBs50m2WL6oEiY1p4tkgkpfKpWC4zhyHZpnqApjLBaTMj2ST5bv0aKe16SJCQkG3RNJLEi2otEostkser0eHMeR6/Z6PXEzZMkmw7VJalU1kqSW83ZdF+VyGa7ritr1emoaVVWSTdUwhiTIsix4nodutytEkrl2gUAAvV5PLPZptw9A8tQYPs1nkkHddGMEgNnZWczPz2NxcRHPP/88SqUSotEoMpmMhGp7nidrRaWT68S/2TtZLBaxb98+ZDKZEaKuoaGhcY24Y/YjOwWh6Un4CXO7h7G1CARh/04DU7EGDmLhmk7xfXe9Avzh+s8v/ZM3Ifj0czdufBq3LLai3PHDAN518ef/CeBp3Cb/UcxkMrjnnnvwa7/2azdMSXviiSewe/dudLtdFItF6VtS+4aYzcXeMDWAmSTo6NGjmJubg2lu/McskUiIQ2GtVkO5XL6sY2AoFEI+nxcVjZvvdrstzoDMv6KBBpUymj6MuwoCG/1ozWYTqVRKXu/1ejBNUxSaarWKxcVFIZEspaPrIABRpizLwp49e1AoFBCPx1GpVER9pDskTS7YX8aS0UAgANM0kc1mpQyQxETNkiN5oVLGsXS7XelXY4/cYDCQPsXBYIBisThiMmJZFprNJur1uqhYJNAkk77vIxqNYnFxESdPnsTp06dlzS+HcDiMyclJBINBtFqtTeMRWO6p9vjRMEVFNBpFq9WSclCSbgBC7BzHEft9rs/4M2TbNp588kkUi0UsLS1heXl5RE0j2WMwOJ8XrgmwEa2wtLSEXC6HQ4cOIZlM4jOf+cxl10JDQ0PjGnDb7kc0tg5Gr4/+uQvA8PL//71RCNx/GP5/amAiunbDzmn9vxbRHsziXDmL3R958YadV+PWw/WSNB/AFw3D8AH8H77v/3cAk77v0xFhBcDkZh80DOPvAfh713n9G4pIJIJUKnXDCBoATE1NIRgMotFooFarjShd7IEi+WF5mWEYUrpGp8T9+/ePmHWo56jX61haWnrdDX8kEsHc3Bzy+TwSicQI4QJGN/zjShjVMTV7bLz3SC3L5GudTgfNZlM29M1mUxQd5nRlMhn0ej0phcvn85iYmEA6nUY4HBaCp5YT9vt96duLRqNCgrh+LDUcL7XjeqrW9VTwWEapEgr2rNHYJBaLIZPJwHVdUQQJ1VCE98V13RGjlFQqhWKxiGq1esXPGJ8PVUFjHp5KTHkPmHUGYETlVHsceW8ZCM41LZVKcF1XngWuG41KgHVSzl6yWq2GlZUVxONx5PN5IYqxWAyNRmOE5A2HQ3Ej5ZjoCJnNZpFKpa5oPTQ0NDQugxuyH4khfjPGqgEgODkBbFcv2k0gaK0ffBQXnhziw6mTN/S8e6z19oB8zMXXf+Wx9df+ogOtrt15uF6S9nbf9xcNw5gA8CXDMI6rb/q+71/8D+YluPgf0P8OAJc75mZCNaO4EQgEAsjlckilUmJR7zgOer2ebJJVNU211ucmma8fOHAAExMTI1b8wPom13Vd1Ot1rKysXLZcjGYQhUIBqVRKzEI2C4kmEVCdAsePU48FNsxFxklbs9lEuVzG2toaisXiSK8Sf6YSRxJiWRYsy8JgMECtVkOj0UCptF69wt45AEJGSKxisZh8VnV3BDYMK1QSxXJAtTxTJW7q5wkqVrSr73a7orhRFSPRbbfbcF1XyCd7DNUMuisBDULUfjcSQvYTkiBTaWOUgjoPvsd1ZPYeAJkzSxAPHjwo5LfT6UgZK5FOp3HgwAFcuHABCwsLKBaLQtKY7zf+rG42Z0Y8xONxhMNhzM3NYXFx8YY6q2poaOwY3JD9SNLI6v8A3SzYCfhj/5/dahi9PuBtfT/X8J0PYPFx4MMPP79l18hHHXzgiW8BAD7vPYL59puAr393y66ncfNxXf86fN9fvPj3GoBPAXgzgFXDMKYB4OLfN07j3UIw1PlGZaMxvJh9Sgya5kaZikMsFhsxxGDgMI+ZmJjABz7wgUs2vQBQrVZRKpVQKpWwurp62c2tZVmiTtFenhtz1RmRRIbvUa3hcWreFufEwGMAI6QtFAqhWCzi/PnzWF5eRqPRQKvVkpJCmps0Go2Rz1mWheFwiAsXLuDEiRM4c+aMOFbW63U0m03pMaOBCHvHCoWCZJmpWWfdbndEwaHCRFWP86YyxZI/EjiuF9VOloxStUskEjJ+0zSRyWQkL83zvBE1kZ8ZJ4CbYTAYoNlsjpQ38nWqXwwU55g5TqqakUhEjFW4ziSVvV5PyjYHgwEqlQpOnjwp6m0ikbgs8X/ggQdw+PBhJBIJ1Ot1NBoNhMNhpFIpWJaFdDp9yRzb7baQWj5fzMsLBoN48sknLwkl19DQ0LgS3En7EY0tgu/Dd1ror6xuyemNUAgB20bAtlH5py4+9NZnt+Q6m+H73v8tnPwH69fXuHNwzSTNMIyEYRg2fwbwXgAvAfg0gJ+4eNhPAPiz6x3kzcDDDz+MfD6Pz372szfkfJFIBI888ghisRiazSYqlYr0f6n9XOzjsiwLqVRKzB2AdYfAmZmZy25cjx8/jpdffhkLC6/fiJpKpTAzMyN9WmrmGf9w087Nv5qjRXVN/ZnlgyRBJGu0dGepY7lcFiMLEiteg6HeVGssy0Kn00Gj0UCxWMTa2poQWyo5JCRUhhKJBGZmZrBnzx4hKWpPVb/fR7vdHnG8pJFJvV5HtVpFq9WSXqrBYCBGH5wjACFtvV5PyAjL9GhDzxLVWq0GwzAwOTmJ2dlZTE1NCVmnpf8jjzzyhs/QcDiE67poNpsSXaAay7C/jAome+/4/JEss7SU5JhujiSnJGK+7+PZZ5/F8ePHUSwWEQgEkE6nL1FRAWBiYgL3338/3vnOd2LPnj3odruwbRuFQkGI8/jaqSWxXOdarYZarYZut4upqakrIq8aGhoaKu60/YjG1sA/v4RBsbhl53c+/BD2PtXD3qd6eNv0a1t2ncvhA0deROrzYRjhnRFnsBNwPV9bTwL41MUNXAjA7/u+/3nDML4F4JOGYfw/AJwD8Hevf5hbj71792IwGODZZ6//m49EIoHp6WmYpol2u41qtYpqtTpCNFS3PjoDxuNxUW4ikQjuvvtuPProo5e9zvHjx3H8+HGsrKxc9phUKoVcLifhytxAczNPsEyPhE21/2eZHm3bAcjmXu1BojsgiRqJKfO3gHU1hSTDMAz5XQ2YVss9VfdJYEOBnJmZkZwuBnGznJFllN1u9xI1sN1uS38bx0RTDgZq856Ew2Gx+fc8b8S4g0SRn+O95vXVEkwaf1BNNAwDyWQSR48exYsvvvi6vYQcu2ppz+eHa0jSreafqWsxHm5NhY3PGq9DIvvcc8+hUChg9+7dlx0XANx9990wTRMLCwsjtv+ZTAaO40gEA/vR2EvH+8gvKxzHwcrKCtLpNB5//HE8++yzKG7h/0g1NDTuONxR+5E7HUY4guDM5M0rdfT9ddv6i+0HNwrBg3fhzP9no4cxY5cQCdyYlplrQcDwkY+4eOH374bvA/GnLBT+2zPbNh6N68c1kzTf988AuH+T18sAvvd6BrUdyGQyaLfbaDab132uVCqFffvW8yyo0NBEAtjoRaOqFIvFRvrEgHWiRxfGy+HcuXNYWVm5rKMjx8IStEgkIuSCYwAgxEjtyWPJIFUPXkO1vlfJk2pAQfWKmWjqnNXSTjpb2rYtZZdUWFhaSJMQmmKwnHBqakr6CEn6xlUYkhTVTVMNd+73+xJA7fu+KH7sUVPLAlVHS0K1wee68DO01ydisZgcS2WPmW71ev2yRiJqeSkAGfO4usX14boyIFztA1T77NSAbZZCUkG8cOEC1tbWNu1FVGGaJiYmJjA5OYlmszny7Ko9nuMkVCXEtPxvNBpoNBqYn58XJU9DQ0PjSnCn7UfueASMmx5efSNyxYxoFCt/7yHg4v8WO1ng++66+syzrUQoMMCT+14BAPx5/14MI28FAEz/xgsYvkFEk8atB90AchEkGTcCmUwGBw4cwGAwgOM4cBwHnU5HVA8qULymZVli1ACsb6DT6bT0Om2GXq+Hc+fOoVKpXPYYlqul02mx9lft5lUjCpZhsnSRRIymJ+x9IrlUc60YZs2NPw1N1Kw5KlccQzQaRTwel748ukGS5AGQNUqlUgiHw7AsC8lkEpZlIZ/Py7lZiqmurxq0zfOpJiiBQEBcG0nISNICgQDa7ba4b3Y6HVmP8Sw3khiVkI2rdHyfrpqNRkOUyMnJSTEg2QwkgrwOCTWATVU0Omeq5EpdE5WYkzyphIkuj6urq6hUKsjlcpd9vvjZ2dlZLCwsyFx5T1W1jMRXNTIh8Sepr1arOHjwoJC81/vyQUNDQ0PjNkQgCGOTHvstg++vm4VcB0Jzs/CTCQysKN7+488iFNh6Z8gbgQ/c/RJwNzD0Dbzy/L0Iv3gGg1p9u4elcRXQJO0iSqWSOPFdL0zTRKFQwHA4lJ6nwWAgIdG0ZI9EInBdV8gKidpgMMDBgwcxOzu76fl938fi4iJeeuklLC8vb3oMTTgymQySyaQEPKtByMzxYh8cN/tUstQ+NG7kSeI4DhIgKl8s66NZCMmBWkLJoOp8Po+5uTnpu2KvGMfGzLRcLoddu3YJyQE2CBywTorK5bKMiRltnufJ2sdiMZkfTUP4WZZV8v6TIFCBSyQSI3NVS1VJArlGXCfbtsWUg/1v+Xwe+XweyWQS7XYbKysr2L17t+TAjSMUCiGbzUrJJO8F3RBVN032qjEwnMdx3K1WSwxLSIipdPEPz+U4Dp577jlEIhH8/b//91/3WSdJu3DhgjwzfPboMhkIBCQ3TTUNoYLJctqVlRXMzMxg9+7d6HQ6OHXq1OteW0NDQ0Pj9kIwmway6Zt2PcProH9h8brOcfyf7sIHv+fbN2hENx8Bw8eR//wivv7/fRip3/v6dg9H4yqw40maYRj40R/9UTQaDZw5c+a6z3f06FHcc889ANaJX6VSkR4mKjGmaYpqRct40zRHAohnZmYuW+rY6/XwyU9+EuVy+bKOjoZhYGpqCjMzM0ilUqL60KmQG+RWq4V0Oj1CQqgqUTVjr5hqs08lqd/vw/M8ZDIZIWjMgzNNE9VqFQCEUOXzedi2LWWY7Nei0yLLPFnamM1mkc1mMTU1JSWJVKFUpYWGHiwH7PV6chxJZafTkeNIKFVljOV/JKLMCVPVNjWfTF2Pfr8vodokpKoRTDgcRqvVQrlchmmaSKfTME0TvV4P+/btQyKRuISUGIYhREtVWkl6STBN00S/34fjOOh2u3Lf1PMA632EqVRK5phMJtFoNKRvjOf2PA8rKys4deoUarWafIZoNpsYDofyTMfjcSl3ZKZbv9+HfdFligqbalAyGAxkjFT3XNdFuVzG3Nwc+v2+JmkaGhoaGteOYgX9Wu2aPho8eBdCn1gvD3xX9KUbOKjtw75/+Cq+/ua34MDHNVG7XbDjSRoA3Hffffj85z+PY8eOXfe55ufnMT09jUAggFarJUSCZX5UFlj6RRVNLTnbs2ePKD+bYTAY4PnnnxdTi3EEAgGYponZ2VnYti15Yjyfaumu9ijR1VHtZVKztNTctG63K0YnLFfr9XoSC0BzDoIKHkvhTNOEbdtYW1uD67pSzpjL5RCPx8XxMp/PS1+cWpKnOk2ytJAEjS6GqtpE8xKaiqhlmwy5Vo9X14fvqYTY8zwhe/F4XO4Fx8XP8Dy+76PRaMhxkUgE2WwWjUYDkUgEExMTAIBTp06JKsnA7k6nI0orSZ9KHNV7w3tFRZSEiTb84z15qlkKx0/iXS6X8corr+Chhx6S0l32bPb7fSSTScRiMSwvL4tSxueHSiLvB599gs8iiS3vaaVSEXKroaGhoXHnIFgowEiYuGlBdP4QuIbczfYH34yFD/n4sHVuCwa1fZiINeHHb49STY11aJKGddWq2WxKaPK1IhAIIJPJSNZXo9EQ0wlu+KmIUIVJJpNil87N/Z49exCNRje9xnA4hOd5OHny5GXLM8PhMNLpNLLZrJQ5kpgQai8TN+kqwVCvx1I+/s7yOJa3qYoaN/JUdGj+QcdE1Vmy1+vJpp8KWjQaRSKRQCAQEHdENVdNLTFU57MZkeO68neVaKklm3yNKpD6vkqeac5B8qpmp6k9XoRqvkJlkp9j/ALnT9LG/jsGQ6u9kqoDp2rCoq7B+B/eq3g8LmvOe6/2wXEtScA7nQ4cx8HZs2dxzz33iFtnrVYT5Q5YVyfL5TK63a7cN8MwkEgk4LquzJvPjmo4M25Mwlw4BnXncjkpY9XQ0NDQuL1hJEz4kfAbH3gjrtVqY9jdvNf79TB414NYfFcAH37w9i1vfD2YGQ+d9z+C6F98a7uHonEF2PEkzTAM5PP5y5KiqwEVIpbSVavVkZI8dbPM0rpCoSBmD1RF7r777pHSRxXdbhdra2s4c+bMpiSNG+S5uTlxdKSKN+5OCGCkt4oEjQRLJZMkKIPBQJwqfd+XsbfbbTiOA8/zxNnRtm0EAgEpw0ulUqKStdttuK6Lfr+PfD4vAdG+7yOXy40YgYRCITSbTSG6/X4fruvCsiwZz7gCpq4HyVs0GkWn0xmJFaDTINeG5Y2qKkUSy6Brlgdyjfi36ozJnkMSqXa7jXg8Lj1yg8EAqVQKxWJRFNd+v4+9e/eKkQrJrerayHmSsLGfjv1w4w6XqhELFUOVcG7mAMneQNd1sby8jGq1OpJ5R9LLNeD6k3AFAgEkEgkh5SrB5zqz5JHkm18G0GgnFovhvvvuw9NPP735PzYNDQ0NjdsHr+MUvBUYrKzC71+dYUgwk8GFn+ngA3u+u0Wj2n68d+9xrP1LG7VvFjAoV4ChVtZuZex4kgZgRDm6VhiGgXe84x3IZrNCIpgbRlIwbj6RTCaRTqflfVq5szRtM7z22mv4xCc+IaYc40in05iYmEA2mxUCotrm8/qqUyMVFZpxcAPOjTzPw5wrNdh5MBjI2JeWlqQ3iefnxjwWi4lrpeM4aLVa8H0fMzMzmJubQyQSGSm5AzZCp1W1jvb1ruuOZKrV63VEo9GR3jKaiLCsdFzpIulUSYRKSqkqMe+Lc2auGwk5sE66SUDZ+8f3VROS1dVVKdNMp9OYmprC6uqqECX2zfGcLElUnTn5hYJKrmi5zzGr5IxOlp7nCZFjlhr709SeOp7DdV2srKxgZWVF7PkZQs45ca58tvklAcsfO52ORFuodvwqMR5XMmmysmvXriv956ehoaGhcasiEERw/56bV+Z4DQjYNmY/38X+4IntHsqWYyLWRPazQZz6B4fhf/vO6Le7U7GjSVosFsP+/ftx5syZEbv4a4FhGLj77rvFYn11dRXAutMjN74kG57nIRKJwLIsAJASsUgkgkwm87rZVM1mEydPnoRt25sSNZI827YxMTEhRMkwDDF6IAnjpp7Xo2GIZVmilqlhyBwrjUTUPLNKpYJSqSTKCABR0BhA7fu+qIuGYSCTyWByclJIULfbRTqdln4vYIOoUKliGR6z27h247EAalmj6urI49XeNpUE0m5ftbkn4eTvdOnkOdT+LlVNUp0Wg8EgksmkqGYcTyaTEfLDNSUJoxOmamfPsap9hiTLnA/71ur1utxX3meOj+NV7zFdILn23W4XjUZDCB8NWdhvSCJaKBQkB45EjKSVpM5xHBkn56V+YQFA+teq1Sqi0SgKhcIV//vT0NDQ0NAw+gMMFpevSkUbvv0oVv5pBweCpxAwbmUqeeMQCgxuurqpcfW4SXHvtyZM08TRo0dx7Ngx2dBeK0KhEDKZjLg2qqYhwIZyxY00yZS6eY5Go5icnLzsNarVKlZXV1Gr1UQ5GQfzx+LxuJQ6qpbx40qRqkYBG5t3kgEqaMAG8eGxJAYMI2bvmFoOmEwmkUwmYds2QqEQXNcVa/1MJiPqF8/LTDa1v0ztk+J7JGQkDzxOXQdVmePc1fuhljLyvCRw6vlUNZTnoKJHm3+VwNAwo1arodFoCHE1TVPcPFkCads2LMuSHDsqcbyuSsz4vlrqyPvK8THcm6Ww/CzLPvmFAe+1us4ARspih8MhWq2WxCKwP47mJ4wtYEC6+owAEPOTVCoF27ZH1p9fFKj3h7/zGQoGg9izZ4+QOA0NDQ0NjTeCfxVxSt6H34wzPxTD47tO7hiCRpz5QQv9731ou4eh8TrQJO3oUXznO99B7RptWoF1gpZKpSR8WiVp6qaYwcgkZMyNUsczOzt7WSVtcXERZ8+eFXIwTtKocsXjcVFjqKpQiVI/w82xSj7UzTqJAD+nEhp1o+84DiqViigh3IDbto1cLodMJjPSi2eaJjKZjBA3kiMqaiQ8NLlg7xLPrfarcTxq7hbXV3UTVI1D1H4zAHItKnacJ8elmomwZ4v3wHVdsfKnEud5npQL8jmgEQr/0P0zHA4jkUggHo9LeaRKFHlNkjS1XFEdv2riogZF8/Pq/eU9V/PRSLKo2vFeMZic42fGmloSqX6ezyHHYpqmEHUSUNWgRiXbhJpjd/DgQe30qKGhoXG7wjAQuElmIdeCpbcH8cF33JkmIW+E73/vt7Dw3giCd+/f7qFoXAY7+ivqWCyGu+++G9/97ncly+lakEwm8da3vhWRSASNRkMcDqPRqGxqadXebDZhWRZs20YymRwhZIlEAvv3X/4fy/PPP49vfvObIwqdCuZv0WWv2+3Ctm0xb1AVMdW1keV4tE9vtVowTRPAhpEGP2sYBhzHkfO6rotSqYS1tTU0Go2Rcrs9e/YgGAzCdV0J9c7lcpiamkI6ncZwOIRlWeJwSJWo1Wqh3W6PmG6oRJHlgmqkgEpagA0XTNXZMpFIiHEHM9QY8M21oEEISYNpmiMW/hwPSZRKplSXTj4XqnGG6hjp+z7W1tbQbreRSCQwMTExYn9Pgjqej8ZoBq4F+x9JxFQzjkgkIqqeZVkyDrUstN1uixKmWvEPh0MhYM1mE1NTUyPumKoKx+w5lYTZti0Om8C6spnL5VCtVtFut2W9ac2vRkDwnnY6HezevXsk801DQ0ND4/ZBwLJgTE9s9zA2h2EAO0w9G8f73/0tfPWefch/cLtHorEZdjRJC4VCmJycfN0esCtBPB7HPffcg1qtBsdxhFRwA8p8KyoH2WxWVCRVwSAxuhzW1tawuLh4Wev9TCYjpWXsWVL7kdg/RZWp1WohHo+PZJqRdLDvisYdNNBQHRYBoF6vY2lpCUtLS4jH43BdF8PhUEKr6/U6ms0mHMcRgsbYAZa0JRIJGaNlWWLlD0D641ieyLUiUaNa02w2hfyw7JJzIMrlsmz4Vet9tW8tFArJtdXQb35mOBwiFouNmHIYhoFWqyWkIxKJIBqNot1uX6KEZbNZ6dFqtVqoVCpCkGzblnJJHs/nhGSPJiBU+dQewWazCdd1RV3sdrtCwFUFlO6TXHuqh+wfZK9hKBRCrVbDcDjExMQEbNtGqVS6pF9RLZfkGvA1lqmm02nU63Wx5R8MBrAsayS7bTzSoV6vo1AoXNZER0NDQ0ND41pR/fP9eDz/4nYPQ0PjstjRux8aZVwP1BBmqkDs4YnFYiN27AwlzmQyoqKpVuabWeQTr776KorF4uvmudFyn6WKdEyk6kESo4Y3j5fKsaSNeVUkjuPKFQApc2MZJw1BmIXG8kVa7c/MzCCdTsu6qH1wqhEH+9SoJqlmFIRKKkk+eYxqmU8SxftAqKRPdR0cN2JR8+DUfC+VNLHHkGvN8kkSMdd1sba2huPHj6NWq8ma27YN13VRqVTQbDbFBTORSIjzJ3u5uFYsl+XYSFZVYkmFkIRJJeucMz/HEk0qiyr5U69BhdS2bSHtLJvkeVnqqbqIct3Ze0fFLRgMSnks/x2pvXfqfSGh1NDQ0NDQuF6Epqdw4tffjPtyy0iErrx/7U7FgWwJJ//3RxFMJrd7KBpj2LEkjZvharW6qQHHlSKVSmFiYkJ6wKikkRzQyY7ZW4lEAolEQn5neRjDiy+HV199FWtra2i1WlIeSBiGgXg8LuHHqhMhsGHwwfGQiJDYkIDQHGK8HBLACKnipp+bdMuyRDki0aC6xN65WCwmOWlU5Ego1J4wx3FE5dqsL4ykQFU/h8OhrOf4ceq5xs8zbqJCgqcaqXAdVBJN4qY6XlKhI0lTySLJBlVFZrUx9JzlsSwtjcViYv7CktnxoHHVDVItFRzvWVPLB/neODlW562ayxCu68LzPFln1fRDdbOkgjmujPELAzVY3TAMMYJRzXPU3jn286k9jRoaGhoatwcCsRiM19nXbBd8K44PP/oc7PD1fUl/p2Aq1sAHHn0OuAF5wRo3Fju23DGfz2NmZgYLCwvXlZM2NTWFffv2AYD0B6kBvcAGIWCpI3t7aGPu+74oJ5vB930cO3YMy8vLcF33kp60UCiEiYmJEdMH9pDRpl11AFTL36hUqH1ZJCVqKZpqxGFZlmTAhcNhcW90HEfIYiAQQLlchuM4CIVCSKfTUgbI3jeacKhhy9VqdaQHS1V8+BqVLK4hACknBSD9YVSH+DlgI0Sc6zquEJLoqOfqdDojGWi9Xm/EQp7Kn+o8aRgGksmkODTG43HUajV0u11Eo1FEo1GYpilGKezfogKm9p+RRKm9gyxDBTDSo8c5EDThGC/TDIVCaLVaI18MqIoiEQwGUa1W0Wg0kE6npWeOY+TakbiTPPK8qutjLBYTlZUkjM8Cj+U51FLJ+fl5eJ53yZcTGhoaGhq3LgK5LPzE67dxaNxCyKZgNBpX5Y6psbXYsSRt3759OHToEL70pS+NOMtdLfL5PKampsSGHsDIhllVQhKJBCYnJ6UXRy0vy2azmJi4tLnW9314noeXX34ZCwsLm7pQhkIh5PN5FAqFkZK+fr+PVCol5Wc0eOj1ekJkWI4XDoclR42fpULEOfAcAIRcUmEh6eDmPRKJYG1tTfqRcrmclPEB68SCpXGJRELs69VQawBCItVyxMFgANd1pfSu1+shEonAcRxxiSSZ4u98n4SUCo5q1a9+hvO0bVtiAah8klior/E6NOFgeR+VRMuykE6n5XN0h7RtW6z6q9XqSI8in6VUKjViSkJVlCWCLHGl4kXSx94xfkHAXka1x4tlir1eT1RQrgcJ62AwwOnTp6XPj/dJDQFnWDlLHnu9nqhfJJi2bUsuXLlclmeIz7BhbARrUw0sl8uYnJzE4uLilfxz1NDQ0NDQ0LhKBAwf+3/vHJ79dw/C+r++sd3D0biIHVvuODs7i927d+Opp566ZpKWSCTETbHdbqNWq0m5FzfiqqphmiZSqZRsrtV+Idu2N1XSGo0Gfu/3fg/FYhHVanVTNYG9ddyMkyRxM89jAIxYrodCoZF+H27sOY5OpyP9Sq1WS4hBrVaTzT/nsVkcQLVahWmasG1bSt3U0j32Q433XFHNMwxDTEFInJrN5ki/H4mFmikGQBwP1ZwxKlp0fmSZqVpGCEAULSp4KtHls8K5q66Y7E0jQeP6U5EMhUJSwhiJRDAxMSHElaHnjBpoNpvihKgqfVwvNVqBa8/IBGCjb47ETlX+1B5EPgskSONZclQJa7UalpaWsLq6ing8PtIDOBgMkE6nxcmzVCrJ+qllsqFQCJZlifsoVUfOg4Y7fF6HwyEajYaMVUNDQ0ND43rQ+JG3oPpfdIjzZthpOXG3A3bszocZXjRcuBbs3bsXExMTotI0Go2Rnh61nykUCsnmFNjo5aE6pRIVwvM8rK6u4tixY9KztFlppu/7Qg7HN99qr5JKMFhOxo0yyU273RYliuDYuHFmL5lqNqKWWtI5ku9ZlgXLskYIGgkNsFEqx/LJccWHm33OnQQNwCXmFcCG2sbyPjptsldMVZe4fmpWm0pk1XnzPtKYgwSGREjtvyIR5JhZkkiDEd5zjoUmHyQrBE07TNOUXj7+zfvJMasqlOpYSRJKNY5rxGPHc+O4Hmo5KV+n4qb27FHN49o0m80RxU910mQvaCwWQ6vVErLG66lrzufoesqRNTQ0NDR2DvyAgWChgEGpBGziN9BJGnjHxLltGJmGxtVjx5K0WCwG27av6xz79u1DLpcTtzuaXpA0cLPJTWw8Hh9RLlS7+PFeIGBdRVtYWMCFCxfEWn0zqA6Iak8V8834GomBShD7/b4QLI6Hn1E3x+pmWu0ZAyCmJyy1ZNizaZowTRPxeFxs9gFIj9Vm5hYkKRwnSeZ4/xOwQWDUnibVLj8UCskxXDsGMfOc6pqybJNOixwfySxJFMOpeT2WRqqmHAy65njUYGp1TdXSRfb9eZ4na0FiyXlTJaWySIWO/VvjpibqFwYsXVTvq/pFgtqrpz6f48Ye45EEJKscH0nmZmNhWayazabGRKjrw/JJtRz1ekx+NDQ0NDTucAQCQCYJo1qFfx2tLBoatwJ2LEkDcN0bvunpadi2PaI8qFlkJAIqieF7LF0jueNGX8Xa2hpeeuklVCoVCVDeDCx/4yaXCgo3wtxMc5NO5YabZG5+WUbY6XRGxkjnRpawNRoNUd84Z2ZuNZtNtFotRCIRsZDnppxrosYUmKYJz/PQarVkQ06VhcrQ+OadhETt/et0OuKayXUfJ8S2bYvKQ0LS6/UusfwHIKWg44SOKh/7+0iAqLINBgM0m80R63quN/vQWKJpWRYMwxjpzarX6yP3h9fvdrui3lFN43PH0lr1iwGqdQBGjFdIeFh6yc9zDup4Ga2gPt+RSERy6/hc8TxUVNmLyfJJEv54PC5fAJDQkaAB63mD7OskqGoyD+96Quc1NDQ0NHYwAkH4AV3qqHH7YEf2pE1PT8P3/es2I2CGleM4EvpLJYpmCoZhwLZt5PN5pFIp2exTQajX66KcqKhUKlhdXUW5XBZ1aTMlLRBYDzOmwyKJC3O8PM+TXp9OpyN9cCQezEHjpp6vcQNOksm5sX+O4yfJsCwL8XgcpmkiEAjAdV1ZIzVygKWHanA3yRh7n3zfF5dIjpkbdfZDARhRYEhqSKRImEi2qCTynqhliKqVvfqHiqDa68fzqVlxLO10HAeu60o2HBU4qn1TU1OYnZ3F7OwspqamhOBTXaL7IcFr8TkjYWUPHskNnTYZv8DnafxeAxAixfusGqtQxeN5eS0S+1QqhcOHDyMcDktEQCQSkQgFzsP3fZRKJVljtXSS88lkMvIM8HpcR5WQs5Q3n89j7969V/+PVENDQ0NDA8DJ3zyKB3/8u9s9DA2NK8aOJGn33XcfotEozp8/f02fD4VCOHTokGyWWbKoql3cVJOQcGMMbPQD8XObhfUuLS1hZWUFruteUvamQu31MU1zZEOuZqCRzNBFkQSCmW6EWh5HAkjSqaou3NRTfUqlUqJyRKNR2exzPFwXEiCWWpJIURFTSzcBSPmcSszUNWZJXjQaFZdCw1jPjSP5Iymm2qQqUqohhfq72iPF9zgXtf9rvKyV4wyFQiOh3nwGEokEbNuGbdsIh8NSajqeQzdOrjhfNY9svHdOJayqUQzLM8ejG9TyRh7P3kWuG41MAoHAJe6katkjsOFqymuqz5NaIhuJRGBZlpTY8tnizySgPG+73UYsFkMul7vk+dfQ0NDQ0LgSBCMDnY32Bqj9sIOVn33rdg9D4yJ2JEk7ePAgAOD06dPX9PlgMIi9e/eKWtPtdkf6j1jOpdrYsxxRJSkkP8wtY3lar9dDsVhErVYTcsZjx8GNO/u0xt8bH7eaj6YqJioZUPuWVDt6QjV2IGlLJpNC2Hg+roNqXjGei6UGMPM11XSFJYfAhqEFN/zj5ibqZl8doxrYvFkP27gip/ZebfYZEkp1DGq/FM8TjUZFuVKJF+ei9uWpLozq2vEP3RyBjb46NcJB7Rkc7y8DIKR6HFQWVYMbdXw0q+H7/CKAz6Xaq8a5cFzq+nCOVO6olI5/OcG5q72TzKVLJBKXjF9DQ0NDQ0PjxuA9e15F46Du5btVsCNJ2oEDB9Dr9XDs2LFr+nwwGMSuXbtks8qsMTWkmcoZN9tqD1iv10O320W32xVXP8/zxHjEcRyUy2XpbWI52WY9dNygkzByc81NM7CxUafCopIOta8KwEhvltrbpNrth8NhKb/k+ejeqJZ6ciOuBmqrJIh29KqBCOdC0qTayHMtWBLItWY5JkmXquSQ2LGUE4Csp2q+oTockkxRFeJn1H4uYJRQkpyQaJFYsQRUJTC9Xk/KUNU+Qt/3R/LRaNXPtSRh5XOjkkWV+LMsUg2uVvPhiOFwCMuyxPCFfWjquUnQSQD5M/sH1XvGPzy3+u9FJV6MQojH4/Jvg+9z7fhccL0u90WEhoaGhobGGyE0NYlgaHPzNQ2NWxU70jhkenoaL774Io4fP35Nnw8EApidnRVzBG4kuTHudDpot9uIx+PI5/MjgdbsYWPe1fz8PDqdDlqtFsrlMkzTRKlUwtraGlqtlmz2abAxroaQGJimiUQiMWKkodrTqwYMfK3ZbMpmmNbwjuNgbm4OzWZTSiJpBKG6LNKog1bzLJvkHNPptJAAhlVzrbgZZ28XyVgikUCr1RKSoapO6s9UGKlOquSThhSNRkPMJjhGEhU1KgDAiCsl58XSVdu2L1lDGqlQFbJte8TJkGMnYSXR6vf70j/GNVTJIUlkPB6X+AJmtrEskmSUa8G+OdWBke+ppZx021TnzZJJ9q9x3blOvO/tdlvmwntlWZasWbValWuSyDWbTRQKhZFx8tpqWLYaD6GWY1JxZkafqqBqaGhoaGhcKQLxOGb+zMGBUHm7h6KhcVXYkSQtm81ed+kUlQ869tGkg1lodP/LZrPirhcIBNBqtVCv1xEMBiUCIBKJoFarYXFxEZFIBPV6XUgSFRW130kFCRw37yRCJEZqrhYdAtXcL2DDKl0ta1ONPuguqCol3HRzQ8+NuqpWWZYlpISESR0P+7v4ZzzPjedjXxzn6jiOmKmQJFUqFaTTaSEKiURCHCB5HtXZkqWfamSBamBBdSwYDKJWq4m6R7JDVYcKG6MMaCSiKl6JRALRaBSlUmkkUoDEmL2L2WwWzWZTXB/HQ5zVsfI+Axv2/pwrDVL43PGaap8eVVKadagmJY7jCCElQeIXD4Zh4NChQyiVSmg2mzJ/ruVmPXb9fn+kT45OoIVCAdVqdcQ5kgYvPCfNTVQrfw0NDQ2NWx/95VUEU0mgkN3uoWho3JbYkSRN/Tb/amFZFubn5yWbqtVqifrT7/dFYVDVIJISGjGogcP82XVdVCoVITaqKyBNLjYjaarpA7CR1UWyopozcAzqcbwON9WxWAye5430wNGKfdxOneqSSuZUlYhqEknIeKA2x0cyGYlExMafYA4XSzD7/T4ikYgQTpUAjzs0ktiqa6eaqrAcVSVNJFu0iidBVi3vSeY4V9WghfeE95ZlqLVaTeIM2F9GwkeFlNcaVxLVUj8aeah9her952tUR6muUUXjsWqAOteLRji8Z7yfNMWp1+tYWFiAZVlIpVLo9XpoNBqIRCLyzHDO6s/qddWexEQiIfeQzwKfDbU8kl+EkJCrz4eGhoaGxi2K4QB+vw9teq+hcW3YcSSNeVWb5ZJdCZLJJPbt2yfKGfvRuAlVXe+oDLDHRy3dAjaIiu/78DwP9XodpVJJCBL/kORtRtLGVTGqGCRDar+TSihU10V1PHRJVN351BI+XkdV5MaNRsbd/viHTo2qy6VqvBEMBmGapmSFARCFhcoKS/BIzIB1G30SLEI18Rg3GFHXRTWqUNeAZELtKVRNUdTrcIyqiyXXmOdS+894zl6vN6KCUs3isSoBDYfDI88EFTiSIZWkUg1VS2xN00S73ZZ+Pq69qsqNOzvy/Xa7jVarhWazifPnz+Puu+8eCaQOh8NwXVcIJEk0lU5VDVRdOklY+YUBxz7uCMpcvFAohFwuJ72bGhoaGhq3OAYDGJ0u/Ghku0eioXHbYUfVDxmGgUwmI8HJ14JkMok9e/ZgOBzCcRw4jjNCPNTSrVQqNVIi2G63UalUZBPNniaGETcaDaysrIhhiOpOeLlyR85LLUkkKeB1xgmZWoZmWZY4/KnZX9ysA5C+KhKeWCwmc+Z5WbpHxYumEOpmnqoP50YSol5PJWA08AiHw5JnRjVLXTceTwWJjopU/9iHR/WIrouqayJt4efn52XMJEYq6WWWGckFiZCav8Y1tixLQrxTqRSy2ayUQXIOwIaxRygUgm3bIyHgLPFUc+1oSkNnURJqkhqWoHqeJ+RJ/bxt24jH43L/SEj5BQKJlaqi8VydTkd6Arm2LJvk+1yXRqOBVqs1Qrj5PFO1VXvtVNMT/rtRy2MjkQh2796tyx41NDQ0bhMMWy30zy9t9zA0NG5L7CglLRwO44knnsDCwgKWl5ev6RyxWAyFQgGGYaBWq6Fer6PT6cC2bTEIIdmhayNLz9rttvQbbdZLRmWEBiDsgVIVt3H0+324rotarSYbb5pNDAYDuK4rRMSyLJTLZSllA9ZJZSwWQ6vVGlFUqPR4ngfTNIUQUY1h3xddKTkn1W2QxiokdpwTe+3S6bSEHlN54ue5WQ+HwyiXy3LMOBlTg5Tb7fX8ExIU9geGQiFkMhkpL1Qt40OhEJLJpChRAGR+JFwMdAYgpiQkmwBGiK4aN9DpdGTtg8EgKpWKEC9VVQM2+rTYe6f2ZFHVYm8YjThI0FiKyfHxXnHtg8Eg6vX6yHVJjFSirCqbnF88HpdSRhLSqakp6TUk6XJdV76wUEk7P0PjFs6V189kMiOB65wPy0WpeHe7Xfi+j2xW9zZoaGhoaGho3PnYUSQtGAzi3nvvxZkzZ66ZpAWDQVEq1E0t3Ry5CaYJg1oWCAC2bcs5qChwE87NLBUZlfhQjdoMg8EAlUoFExMTI1lXg8EAlmWJIuL7vhABZk9R1VHJIq/PTXIqlQKw0c9FpZAOkpFIBI7jiOIFrDsUJpNJ2VyrZiq8lmrpT1JDUqlmZKmufipZ49hpyMIeKrUfTY0voFpH0sQQbBq5ABA1Uu33YrwCx9TpdIRs8/zsrVL7+JrNpsQM0PGS82YJrLquAEbUKfbH8Vy8Z57nicLEPjK1Tw7YMOvg+dV7y+taliWlvyTuXCM+w4FAQEgUS3KpNMbjcdi2jWeeeUZ62mzbRigUguM4sp4cL9eezzpdSXmd8f5NVcllySO/wNDQ0NDQuE0wHMC/sIzAZAF++OZuO42HjuDVj0dxV+DaIpd2Gr78yTfj0OeruDSVV2M7sKNIWiAQwL59+/DVr34Va2trV/15Zl7RGIObXhIQbuq5eVbz0YCNHCqqLWrelbp5VnuL1Iyzy21OWXpJhYnHUiHhubvdLuLx+Eg/npqdxY20qriovXDq9am0qTliamBzPp+Xcasbb86dm3CV8JIYktRxfACEQLDMjqWC3PCzJJPnIbnlPSLxYd8V14XHU6lqt9uwbXuEuLAfSiVTJGT9fh/VahWWZYmaynHxGSDBo7pEdYtlqeo6q+/xXqr3S+3vU5VF9flRSbf6jFDB5H3iPPg+n73xObL3stlsYm1tDY7jwLZt2LY90k9nWdbImFSFjmqf2o/ILyvGTWXUvjreH5JhnZWmoaGhcfth2GohsA29xL10DB868sJNv+7titSZAYYvvLLdw9C4iB3V3BEIBLBr1y4sLCygVCpd9edzuZz0FbVaLVGnVOWD5V7sG1LJDkscWfbFcjWVCLEMjOVuLGsj0dkMw+FQen/UzT1JDVUXWpurZiCqaqeanpBURKPREcKoGjuoAcc0jGBv18zMjJQXshRQJV7M9wqFQrKWalgygBE1jGNjlhnXnaRPdUMEIMojTVhIkFTCyPvIsTD6gGvOnjYSNpb28f75/nrw+IkTJ7C4uChOnCSeqvqmOm7yPvOeqH1XnAOJPueglihy3qqDo/ossFRUJURqLxz78nifVUdQFXym2W/GmIhz586h0WiIShoKhRCNRpFMJpHJZERhUwkq5zhO9PmMqdb9XFt1zmofo4aGhoaGxuVg9Afwh9pc6mpxrDaNkKfX7VbCjlLSDMNAoVAYMcW4GszPz2NmZkZK4Lh55eacChvNJQCMuNTxfbV0sdVqIRKJjChIJALs6XIcR1wCN0O/38fy8rL0Bvm+L+V4JFIkAFT8aFbBc3MTzI20as3veZ7MD4CoK6Zpjrgesu8pHo+LmjVe1sfSTRKUbreLWq0m7zuOIyV86ppQdbQsS4gUyUcoFILruojH40LaqHRSyeHxVKVUEsVeNppZqISRqhZVHBqiUHUrlUp4+eWX5QsA9rd5ngfP8+QcXFOWRap9daqxitqHaBiG5JhxHuxTY3mp6qxIM5VgMIhkMinkk6+pZiecT7fbFYLE9eY9VtVJ5qT1ej18+tOfxvz8PDKZjLhRqm6N7OPjHOksyedCLRO2LEu+uOCXHWq/JA1+qA5qkqahoaGhcTkY/QH6Z85u9zBuOwx9A+EfbiNQ/OZ2D0VDwY4jaVQnrgW5XA6ZTAaGYcBxHFEHSILYc5NIJJBMJqVcjCoJFRESAJIwNbON5W88p5oj9Ua24wsLC9izZ49slKm0ABsllJFIZEQZSqVSshnmZ0ql0kg/HUkklQ6qViQNJFs0v7AsSwgHyWK73Uan0xF3zeFwKJv3iYkJNJtN1Ot1WJYF13VHbP2pTLFsMZlMynss46O6pX6OpYKqdT2fAyqFJGRUbVTnRRJTljqyx4sGF+fPn8e5c+fgui6ef/55HDhwANFoFLlcTvoJqVSqdv7AhlKkqnIkT2osQDQalYgHHkOlVI1fKBQKkh/GMlISMRJz9v6pWXVq7xkAeJ43QlBJkHl/pqamUCwW0el0pPyXzwoJrdqrR1dOlYCp//5UFZGKMZ9P9iNS0WOenOq2qaGhoaFxe6C/sIhgIQek7O0eioaCb6zNI/sP+xiUFrD8T94K96h3yTGBpRj2/fNntmF0Oxs7iqRxE3yt4KaUageVBW7gubmntTjdHtk3pdrcc6PKTSg3uABGytVUe/Y3IpcXLlxAOp1GIpFAIpGQsVJFYQmfaslPEsHNOgkY10k1k+BmmsSGm3vVkp7XSiaT6HQ6I6YRXA9CdWbk3GnSQUIx3j+nmpDwZwZCkyCQGHD91FLS8aBoqlIsI1TXmWoZx9XpdIQ89Ho9FItFlMtlUe5Yism+NrUU0zAMMb0g6eBYVALOZ4jPDI9VSwXV/rbhcAjLskZMQmKxmFxbzXzjerLUlUqe53kjBFYtOVTLealOtlot+SKBDo38HNfOMAxRNtVeQEKNlWAZMJ8TPkP8YoHqH0lpLpdDqVQSkxoNDQ0NjdsAwwGgfCmtsf34sxeOIvv1MBZ+yAAwC/feDiZyzUuOW23tKLpwy2DHrDqDkqniXAui0ahsNrlB5eZRNd2gGYJqQw5sbHoJkhqVlKkqgbqRVh0PL4dyuYxarYZsNiu9RyRHVNdUkkbSqpb+0blRdUnkMby+mstFEsm14e+maaLT6UhuFomFOjdVHaKi4jjOCOngdbh+3MRTxWM/k+u6cl6SG5I0XoMqjXr9aDQqpiuqmQjJHgl5o9GA67ridMgS0VgshnQ6LX1yjFKIxWIANpQ9khHOQ1X8uCaqOsY/fH74bKhGKurzoD5n7CNUSSqJpGq8wTGxt1I9v9qbyNdJQvkc8R6pZEstIVWJl3pfOC8+83SKpJqm/kzjFz4HvV5PlFdN0jQ0NDQ0NK4OQ99Asx/DwDcQKoaRWBsg/H+rAAAS2zw2jVHsGJIWj8cxNzeHUql0XaVSg8EAnufB99cDmxkATJt7KkoqQWBfGK3NgQ3iEY1GpX+J5YKqWUY0GhXzjTfqx+l0OqhUKshms9IvRHLGTb9abje+mSZp4jEcq+u6kukGQIgfN+xUASORiBBKrpVqMtLpdKR8kiWMJF6qY2Or1QKwnklnWRZqtZqULjKfjcSBZILXZK+gaZpCklWio95HGpcQJAGO4wjR7vV6WFtbQ6PRGCF6MzMzME1T3l9eXh5Ran3fHzH+YBkhyRuwUY7JtWZeHXu8VEMU1emQ8+Xc1D8ku67rCsln3ABLTEnY1FJcllbW63UhlcFgUNafJZ7NZlOs+03TlOeC9489e7VaTYgic9ZIgKmQqSW9jIoANr6cYK8ej+c6Tk1N4cKFC1LeqaGhoaFxG8H3gZsQpWKEQhhEdpQ/3hti6Bvo+0Ecr07A9w3k7iuie992j0rjctgxJC2VSuHQoUN49dVXxYzgakDik0gk0Gq1UKlULlE2WHpHJYoqjFpCp/ZP0QVSDRTmudjHQ5ONRCKBVCqFZrMpJGYzrK2tIZFIYHZ2Vnp8XNdFr9dDoVAQF0huelULfI4pFAqJUhKLxYREkWRSvVJL0YLBIDKZDJLJpJRWUqWhYsI+vVarJXlw6XQaa2trUlaYz+dRLBZHnDITiYQYnIRCISwuLmJubg6e5wn5iEajaDabCAQCsCxLiJqqZIVCIbkf7DlzXVfWvt/vY2FhAY7jIB6PI5vNYvfu3Uin0yPqK0PJbduWHDSuK5VEKnG8n8lkUgg9n4dGozESlM3SQQZ8q9EJaq8ae71Ugso14DOjKqK9Xg+u646YolBhI5njfeazyS8ZHMeRZ5EkkaHoLIVUnUCZocZ+MhJAlouSlLIkmPeDai97AnmPVJWZ65FKpbQVv4aGhsZtiEGpjIDjwtg9u+XXeu2XHsG7nvzOll/ndkKtZ+JkrQDf13mjtwN2DEmLx+OYnZ3Fa6+9dk19aTMzMyPmH9w8q5t3qgLcQNJeX7U4p/kBS7g4lnF1geegG2Q6nUY6nUa1Wn1dkua6LsrlMkqlEpLJpPQ+0aGPJXncWDebTQmrpsphWdYIcUin00IsqAiqJXlUmFKplDg+9no9TE1NoV6vi8OgWkaolthROWy1WmLhDmBE/SKx5AZeVWGAdfWNjoJUbThfrj9JKeegXqdaraJUKqHT6WDv3r1Ip9NIJpNiAtJqteRc/X5fyAjJCTPPSC4cxxGnT85B7a0CMEKoqFT5/nrAM58JhkKT8PA1flngeZ6UByYSCblvfI2/JxIJccCk66Na2sh7r5a68j7F43H5MoF9eXxGo9GoBKQTfLZVBZDlqCSwavA5SRh7C/l6LBYbyR/kM6KqkRoaGhoatxd838fNoAh+EDCD1+5DcCfhldoU+n4Ag2FAE7TbCDuKpM3NzeFLX/rSVStptO7nxrLVaokzIF0PSTTU7CqWVXKTSTIzXm6pkp1OpwPTNEccHakmkTgUi8XLjnUwGKDZbOL8+fOYmJgYIXx0jlTL6NT+LdWcgaWCvV5PVBauhUqYxnOv1LwtqkokITwHf1Z7uNR+J7VfjWSLxICllWpfIOc9bk4xns02bryiGnfQUXB6ehq7d++GZVlCQMYdGV3XlQwwkj81E2y8B1ElZnxeOBd1nqrSR6hujlwnlo6qIeQshaWbJfvS1BgCfrnAZ45rxN+pvvHfB58FWufTCIbzpaENr8USyn6/Lxlzai+dGtegKrOqy6TqRslniMfxS44rMdHR0NDQ0LhFMRgA9SaQtG5K2eNOxKefeQiw+njwrnMAAKcXwWCo/795u2HHkDTLsrB79268+OKLUlp3NSgUCmIWwYyqcUJAYwbVeh0YDTKmNbrq8sjSMG5c1c/yNdM0kc1mryiE23VdvPbaa9i9e7f0iDHvzDRNIWj8ezwsm5lk7LOjCQgACVHm2MYt5oENMqNuqFU3Q5WUqkSNJFU9RrXJJ1lSTVBIPnhdFTyeJCyRSEhUAAC5FseeTCZx4MABTE9PCwFjzhdVOapqnD+JkLp+qqGKSn7VrC9V3VPdNdnvyPfZK6aqW+FwWI7j86b2Kw4GA9i2PUK2qACrRE+9h1w/VeFSA7Mdx0G73RbCzueK5a8sqeQzQ1MXfknAZ00l+/xygGMg6VMJrUp8OX/VsVNDQ0ND4/aC3+9jsLqGoJ3QJO0GY+gb6AxD2PUFH825CC5MpLd7SBrXgR1D0ljueO7cuUs281cCmlXQfp0h1NzIsp+Im2WW/Knugiz142f4t5pLxYwxEjdu3A3DgG3bSKVSYtpxOfT7fTQaDZw9exaFQmEko0otQRsMBkgmk6hWqwiHw2Ig0mw2JYxbLRvk5p3qRiaTQa/XQ6VSETLB8r5Op4N6vS4GHiwJJJHt9Xqo1WpIpVJiFc9Q5VgsJllclmUhlUqJgsMx0/iDfX0MtFahBiNzXYAN1YhxAolEAtFoVEiUqoCSEJGE0SWUBBFYJ9r1eh2xWGyErDFGAFgv3avX65eEgqvW+izDVJUiPnOqEyLPxxLDcDgsxIXqHA09SBa73a58ycBQcMdx5DO8p2quHrD+74Z9e/zCoNvtwrKskQw/z/NQrVbFwMS27RHTHPVnNU+NBJn3gGWxnDvXgWoayy41SdPQ0NDQ0LgU3y3PwP8HFUS2eyAa140do31SFblW2LaNwWAgfWhqADUAKUmjIQIVArVkkaRCVc3Yp8VzMh+Km22qUcwem5qawvz8/BuOdzAY4NixY7hw4QIcxxElioSDCpXneTK3er0u71GhYY+SuimnWYbaL0Yyxj4mkj5VNaJxRDAYHDFEIfmkwx+Vw3g8LmSB662amZDEqmWVkUgEiURCPkNCTmLK5yAajcK2bUxOTmJqagqpVAqO4+Dpp5/GysqKWLyTHFDxMgwDxWIRKysrqNVqcBxHlFWSa6qCXEe6aPLesvSP5itUEKvVqjwzzFtj7h5jDujOqK4jnzfmy/E1lg1SDbNtWwgtFS61dDIej8tzrJ6XXxKEQiHJvVNdPJvNppBlloeqKhwt9lWVWf2Sg394bf47dV1X7itRq9UAQHoCNTQ0NDRuTwzOLMBoXX1lk8al+MpvvRnP/6sH8Vxp15b0nGVn6jj9ew8gYOsg8puJHaOkBYPBazYcoNGBSkqYr0WSwN4fNZcKgKhp3OSqmVFqGRqPZwkYFRRVWQiFQkilUpifn8eJEyfecNy9Xg+nT58GAOzevVtK6piVRhdHy7JEMaNRBEkJ56v2sLXbbZkvDSuotrGnjAqWWtZJt0eWEOZyuZEgafZ7UdUCMGJ0AmyYm6jX4HFUDNUxqLlc9XpdCAaVUd5b5p3t2rVLDEWo3Kn5Xr1eD4uLi1LyRzKp9pWpfVgkaezt4zqRZKpEnaSFcyapp+JH1UntLeOacC3ULyPUskA6Xaq9hiy9ZQmsOgcSJ/ZfkjypJiAsPyS5owkKDUuoItO6Xw0tJxEnkRy/x7yfaoksVcBOpwPbtt+wP1NDQ0ND4xbGcLBux69x3agd7cLZFUZui/rOwsEhsmlHV7HcZOwYkqYaaFwLVPLFDaj67T+VEgAjLo3ARiYZN9eqoYS6SSa5UHuEeD6eIxaLIZ/PizL0RlhaWhIVhptt2uPzZxWcG008uOFXDVI8zwMAUXO4FlS91FI2lneyj4ygIlcul0fMVtQeMx7XarVEuRsOh2i1Wpdk0HF9qFSpJYTj68TrqdEH4XAYmUwGtm0jkUiMfEYlyuy9InlQFUW1B433UR0nr8l7O+72yFJIGtKwf4xqFrBBtsYJKn/nc6ba6bOctN1ui9qmllPyvnIdqXZR8VLVLQBiy8/xqaWYfK753NJJsl6vI51Oj4xbVZB935dSRrUcVS3/5PNFc51EIqFJmoaGhsZtDL/TgREOwY/oWJWrRb1n4umX78ZDB85i/95VNGd0dcmdhh1D0tTN8LWAm1puOrnp5sab1umqOsbfVXdHtXSOr8fjcSkjowkDVRRCJXAsqaR5xOvBcRycPXsW3W4XmUxG8qhU1YLzATDSF0T1kaoOLfYbjQYymcyILXooFJKgYypRw+FwJN+sUqmI3T/JEfuyqMpw808rdpbtpdNp+L6PVquFZrOJqakpIT9cT5JFvt5qtURZCoVC2L17t9yzccJHJ0fTNKXMkWvEcj+Satu2JbqA/XG83+zPIhki+VHz2lSzDJVgjjs0BgIBIdHq/Or1umTO8VxUudQMNT6bavyC6qJYr9dHSgzpmKmWEZIAqqpqu92W0G2WX7LckQpwOBwWZdZ1XdRqNYkWoJrJZ4/rn8vlRgg/CRoVcJbN8rrail9DQ0Pj9sagVEaw2wOmCjfupIZxxyt03iCMl0rTOPSxl/Gd//Mg8mnnplzXiJuA466roBpbjh1B0ljOd/bs2Ws+h6qkqQoCwQ26Wj6n5kABG7bnjUZDHCbT6bSQAMMwhAA1Go0Rcwv2YQ0GA1iWhSeeeAJ//dd/jUaj8YZjdxwH58+fx/PPP4/HHntMNv8kha7rIpFIIB6Pi02/Wo6njp9zbDQaspkn8alWq3AcB77vo1AoYHFxEalUStQkjiWRSIixCrCRU1apVJDL5WQ9qEzxWJXIkACwj6/Vao2YiDBfLJlMwrIs6ZWyLEuIIMkHLevz+Tzq9brEAbAXD1hXJHmNXC430k+nKkg0FnEcR8wzSEpM05QxDwYD6Z1TlSySVVUBpBJJ0kulUnWA7HQ6UmapKmYAxOyG5Izndl1X7g9LMjudzkiOXiaTgeM4oqSRdKkZbY7jIBAIYHJyEjMzM8jn86hUKnj55ZfRbrfhOA5c1x0pjWT5oppVp5YJk4Dyeex2uxJ/4XkeLMuSe6OhoaGhoQEAfiiI4P69GL62sN1D2VI8/cWj2P9/LmLlj+aQD94cghYwfKx8IoXUr84h/OVnb8o1dzp2BEnbs2cP9u3bd13nIGFhDw0dClmCxZ4j9uOwh0gth+NmOZlMSj+QGh7MTT7NQoANYqe+HwgEMD09Ddu2xfXwjdBut3Hy5ElkMhkcOnRIFCASLGCjf4xmEtz0cyPOzbiau0bTDjr6kQBQQSORYx8c3RxZVker+Gg0iqmpKVFYGo2GKDLdblfiA6LRKGq1GtrtNmzbFhVuOBwilUqh0+mIRf6FCxdGTE3YE6eaZuTzebG15/1lLxpdEGkkwnDqeDwu5im8NySiqjFJtVqVLwg8zxMyx3vJ+6/2M6r3kyoS15T3Xy2ZpErG40mk+fyQEHKdLMsSgptOp4WAU7lSVTSOWy1lpYpLosY1z+fzorClUilRCF3XFYWVpZVqJIP6nKt2/arSSKUQgLidWpZ1XUZAGhoaGhq3BgaOi8DiAMbs1I054R3eN/W1//0RJAG8/PNTmAxWbuq1w8EhELiz1/dWwo7Y5UxNTSGTyYgz3NVCVYLUfjGqGeO9USQNAGRjr2aQsZdJ7RkiaMxBt0CWyKnKCo03du3ahV6vh7W1tTecAzfMp06dEqIUiUTEzIIbebrpqTbw7GHjHKnAsU+K41KdFDlOKnEsB1TLBkmKHMeBaZqiPPF93/fFjVLN9WLPGNUiqmyMEOj3+2i326KgkQhzjqrBCeerKnR08KRixfOToKmZXpy3GuTNNaLySWJFIkIi32q1RmIYSFr4XKh9bTwXx8p+LvU4krtx5VZ1T+S8VKWNnyERVL9YYMku14TKHckXe/ioirEMk/l7tPVPpVIja08lU+0/5PPYbDZH+gq15b6GhobGHYzhAEOvjeAbH3nFCKRTGN5hbW7dYQh/9ZkHsfevl1F7eBKTu28uQSNWHolguv8QQk9pNW2rsSMs+HO5HBKJBJaWlt6wh2scJCTj/UTc5KuglbmKdrstPUYsJ6PSMV7qpeZQsT+LBEE1ZKDCtXfvXkxNTV2VIcr58+dx5swZXLhwAc1mE71eD81mE/V6XdQr9olR1fA8D7FYbIQIsM9s3MkSwIhiQmWIBBDYcAWs1+vodDpS3mhZlvRRMXOs3W5LvxjfS6fTQnZUMxZgvfSS6zo9PY1cLod4PD4SBM2xqi6bXF/eHxIJrlE0GoVlWUgmk7L+vD6vS8LHUklei+WUfH7UtaH6xrWlcslST7WskX10HC/HqP7O8/G6466Tag8bCSHvdSwWE4JOC366KFqWJSSYBLPRaKDRaEgZom3byOVyMmZ+IWGa5ghJU8tNVedIjkE1VOG/C665+u9SQ0NDQ+POgdHr37hesmwawTZwvD55Y863zaj3THxtcS/m/92zMFwPndT2fXGZfNsalt6hTUpuBnaMkhaPx/HSSy9d9WeDwaCU1alqmeo+R8ViPMuLCpTneXBdV0xC1EBj5nOxR8d1XVHSuFmnwqKWNQYCAWSzWczPz6PT6eDVV1+94jl997vfRbFYxP333y+uf+x5a7fbQkRIKFQ1r9vtwjTNEYdDqjA0JFHVwG63O1L+Nzk5KWQQAGZmZiRDi71YzWZTDC5I3CYnJ4UAkRym02kZt+u6WFlZERWHuXI0mAgEAlhZWYE9lvHRbrelP1Alh51OB41GA8eOHcPU1BRM0xS7fZqfqBbzfB5IMFhOqJqHJBIJiS8IBAKIx+NifAJAVCTe836/D9u2RxwwSdJarZY8h4PBQIxVOC7eO35JwPVVx011j6ol1V2W6lIBUwkgsJ5fRjJnWZb0nZHgAkAmk5H7yrEzK0/9N0LVkYRWNTshkWPpLQAxlVFjBzQ0NDQ0bnMMB+i/dg6hvfPwwzdma3rX76yh+sou4J+s3pDzbReGvoG/PrUf+3/sefgAjv/cHhTu1s7GOwE7gqTR+puZYVeDYDAoJVxUt9QeJpbSVatVFAoFUcJUwuK6rrgCclPLTT0JBTfuqlGDan2ulslRoQiFQsjn81dN0gBgbW0NzzzzDBKJBJLJJDqdjhAvIh6PI51OYzgcol6vS08ZCaZqKd9sNkecFVU3y8FgID1q9XpdrkNCRLIxHA5RLpdH7OSpoKnmIel0Go7jSH+ZbdtwXVeID4mdalDS6/WQyWRGYgJc1x3pr+KzQuJFcmDbNrLZrBCFXC4nJXk0BaGqxeeDpXt8HtTMOzUHjOMbV8HUUtpkMjni2sjxkQBPTEygVquNlHAyS0wNTue4SJhIuFUip6pu/Jv3jqDjJu/r8vIySqUS2u02Tp06JeXFNLVRv2Ag4QfWDXRYNkqyqn6BQadH1dmUz5OGhoaGxp2H/rkLCBZyQEqHJhNf+Z1HcPdvv4JhOIJT/+MI0sntKXNUEXugghO/+TAO/uS3t3sodzR2BEmjEra0tHTVn2X/GMvl2GfD/hwqNzSFACAOkACEfMViMXEhVG3h1bwtqho0ouDnWWoI4JJSMMuyMD09jX379uHs2bNXHDMwGAzgui5eeuklHDhwAJlMRkjF2tqakE+WwrG3igrIeJ5ZIpFAo9GQ63Nt1CyseDwux3DNeE7HcVAul6X3izBNU+zeeZ5WqyXZbzRgoWpDl0oAIwYb/X5fiAlJQaPRGOmdAtZdHGmQopqOsJzRMAw4jiPnGy/tA9bJBQO+qVJxnFwX3kOeg8+abdsjCiXPoRIwlkaqMQIcD58Z/k7yqDpQqv1eLHlU89V4/7hOVCJpIGKapqiujUYDtm0jlUqhVCqJycu5c+dQKpVGFC91nXl/OAcAojSTlPMZ36zMUVUNx0uMNTQ0NDRuUwwHGNbqCHjrFS6YzN/xRiCvh7/+xCOY/UoRg2oVRjgC2/LWzTu2GdFwH+mcg/P/21ux9xOn0V+5vdXKWxU7orGDhhVXYrAxDtUJUM274gZSVUVYusXyNtqts9xvfOOpWsCrpXKEmu2mZmKpOVvsbZufnx8hc1eCfr+P8+fP4+zZs6jVakImaFJBdY3kVDWP2MwJkqVoNBQZXy+qbOxFYs8ZSU69Xpf+LZbIce3VUjrXdYVcBAIBCQRPJpOyyVeDqrmOHD9t5FnqSNv8RCKBVquFSqUiroZqyaIaBzDeI6jmoEUiEZimOeL+SVKu3kOeQ+3XU+dNlZHPitqbRpVJNShRyx9VQgaMxieMW/iPB5KrPWK8Hywb5dz4LDPPjmS30WigWq1iYWFBylrHzXQIjp8kjCWuai8fnyPORXWDZP+ehoaGhsadA7/TwaDRwKDRwP+fvT+PsuQ+rwPBG2+P5e35MrMqs1ALCiiCBEGCACkREiVRtGTa0rRstUXLHkuyJY+sY7vdHrv72OM57a1ljzyWPbLlTZIX2u1tJI+6ZWsxj5okuAEECAI0dqBQVaiqzMrt5VsjXrw95o+s++UXrwqoLQtVlfm759RBLu9F/OIXkeR3373f/Sy/BysIYQ2vnWJ9NWSbE/zGNx/FeHpv2ePbIxu/8Y2P4ND/cRaT107f6eVcFdn0GIWPb6L53ceROn70Ti9nX+KaJM2yrH9pWdamZVkvq59VLMv6XcuyTl/+b/nyzy3Lsv6hZVlvWZb1omVZH7mdi79esNC/GZJGlWu2/4qWMSoKwE4YhY4Y73Q6sZAGqkQsSlnU047HonQ2Bp0qkk531MmE0+kUS0tLKJfLUuxfL3q9Hl5++WWcPn0a29vbiKJIVDVG6TOynaSHBTmVv3a7Lb9vt9vodrtCwEiuLMtCo9GQdZOI1et1AIjtiSa17FPToSRhGMZCSIAdIu55nqhCyWQSvV5Pzq0HOrfbbWxsbMisMNd1cejQISwtLcl5c7kcisUiDh8+LAObqWjx3vN5mCXbJDMkPHz+HMe5QgkiweTMNCqtPD57Imn75KBzBntockXbIsFAFgaC6LAOElwgrnpyLAD7IQGILVUrlbwWPfibx2OPGveE/XX63Pw7IFHjPZ/t69NBIwSf+Uwmg3K5fEPPu4GBwb2L/VCPGNwYxmvrGK9eQtRoAdPp7r/rCRiJIrgvruJ9f/4VDKb3jnEsGGfxcv0QHvyZZ3cVqkQSibvUAjr9iTrWv/8wEuZD0z3H9ShpnwXw6Zmf/WUAn4+i6AEAn7/8PQD8PgAPXP730wD+6d4s89ZwK4NvWSiTfLHQpGWMYRIkaLQR1ut19Ho9Ubqy2awU+VQBWDyz6E8kEnBd9wp1g0SH72GBTFsYicFjjz1200XrG2+8gSeffBKnT5+WmV4AUK/XcenSJWxsbODChQs4d+4cLl26JH14HK586dIlnDlzRmx4Fy9eBABRF0lastmspAFyz9LpNPL5PBYWFlCr1eS6GRQyPz+PXC6HdruNtbU1lMtl6d9iTP94PJZh00xO1FZG2jc5ry2Xy8H3fTSbTbz11ls4f/68hI8kEgmUSiUZWs10SdowqVzNxuVzTQzpIDEi0Wy32wB2+wtzuRx6vZ68liApAiBqFlWkfD4vg9OZOMlxA3y2OEx7OByi3+9LwmYul0MYhjHizf7KTCYj95I9fdwrrk1bVBk04routre3hSD2ej1sbe00NFNV63a78reig1Y4eJvrmE53h5Ez6TEIAti2LSMg+v2+kGMOJzcwMDgw+Czu8XrE4OYw6XQweeuc/EPnOgY4b21jvHrjbS53Gk8++Qgq/108Q8F67P3Y+mwV2fTdae9P/nd1vPF3P3inl7HvcM2PFqIo+rJlWcdmfvxDAL7n8tf/GsCTAP7S5Z//m2inqvu6ZVkly7IORVG0tmcrvglQeboZMIyCNjkWvLR7MTFPKxMsIPWgaABim9NqRz6fF6WMVjXHcUQ54Lpp9SM55FBiFrMkcYcPH0YURdjYuDF/cBRF6HQ6eO6559DpdPDggw9KwUzywT3Y2NgQpY09SGEYotfrXdFPRPWn1WqJcqUJlu6PiqII+XxeSA2DURqNRmzmFu2kVGsmkwkajQaq1aqoWrx3LP6n0ym63W6s343r0fZO3msqVdpyqVXRIAiESOs9ZBBMEATSZ0jlh/uibayaaJMEaVJGssT95Lp4boaCALszzWgh5e+oYjHJUyt+/BCA+1QqleSe89r1HqZSKQwGA7iuKwSZfXokTK1WS8ifbdvy90FlkfdejzCgwspr10EnvFdRFCGXy8WCTmbTOg0MDPYv9kM9YrA3mG43gFZ750O7I4ev6FuLVtcxvdzXNg1DXPyZ92P1r0b41H1v3onl3hgiANN4QFZkAcnEne9FeyckrAj5pQ7e/Ccfw/v+wouYXnaYGdwablb/XVD/Q7cOgIMolgBcVK9bufyzK/5H0bKsn8bOp1u3HVqluFGQaOjvaWdkDxjDGYBdu54mhizwdegHgx10QqIeTqz706im6CAOnXxHRadUKglpHI/H2N7evqFrZfz9hQsXMBgMUCqVUCwWUS6XxZrGAtr3dz7FYt8aCZq26elEP51gSHWN62T6JZUSFvjsN9MBFrxukmUW9wyH4fEty5KYeqLT6ci9ZKFP5TMMQwwGA/i+L8ci8WYsPe8NAFkT+/EymYzsDwklv5/tM+RxoigSdYk2SpJxhtOQxJGs0NqoLZHcS6pNJJn6uWSvlx5crvv2aB+k5VAPD+d59fo0gSNhdl1XSHKr1ZL+QP69aGsjjz2rVHOdswmTV+vlBBAj3QYGBgcSe1qP5GAsW/cCovEYGI8RAUg221f8ftrr7VoiowjRC6/A/o2P4ze++0P4oUf+23u72BvAf/nK41h6+t5MMLYzI2SPNrH5E49i8XOrGL994U4v6Z7HLZt0oyiKLMu6YRYURdEvA/hlALiZ998INCm4GZAc6WKS5EETFyoyLGi1QqS/1nPHOMD4aqEKVC+0pY5FOa+JhXo6nUapVJLXMb1Qq3bXi62tLWxvb6NcLuPIkSOxHiLdL8TimaRTW9X0fCutHFFJ5M9IaghaAklMSf54TBIFKi3sZapUKrGgEvak6XvU7/fRbDaFYFEh4twtqoKlUkmi4ako8bW8l1wfCSXJgj6uHiVAQq37wKi8aeWPc8T0fvEa+HNaSrXtT/dJ0tqoh6vz51wbn0XuPfvNuB4dOMJ18jnXtk4SWPYkaqWR18r1keBposav9TXoAdezYTu8Xt7nG+3BNDAw2L/Yi3qkYFVuaz1isPeY1K/vA+nyZ59GcvDt+FLlpPzs/kody07rNq3s+jGNLHz10gnc/2t9WF/71p1ezk0jYUVI/NAWmq3DKA+GGK+t3+kl3dO4WZK2QduAZVmHADCRYxXAEfW65cs/u6Ng8X6zYDFIFYZBCSRtLFqpcukiWicZJhIJ9Ho9sYABEMJC4kFLH1Ucfk8VikEQ2lqoB2rncjlUq1Vks1l0u11cvHjxpojadLozs6zVasH3fTzwwAMoFosyV43ny2QycF1Xeu70zxnUQULG/i5eD5MVgd2Za51OB6VSCUEQANghEr7vx2Z7caA1yRWwo2w1m82YcqTXM5lMUC6XUa/XY5ZMWkYZyMG+QFod9XPDa2H/FokKVSs9hJk/4zXm83lRH0nm0+k0ut0uisWikMpSqSTEHdgh677vx+y67HMEILZFzo2jQsj5Z1Se0uk0giCQdXKterQC7Yp6ZhsthyRzuveOyh2vh71n1WoVr7/+eizZkqodkyCpsulQFqaD6pAQBtCQrPK1DDjR8fwGBgYHEvdUPWJwZ1H4D19H4T/sfv+tv/txLH7P8/J9wpoicXt1g6tiME1h4Y+sYHq59rnXMf2JOs49cAJHftaQtFvBzZK0/wzgJwD83OX//ob6+Z+1LOs/Avg2AO077f8+duyY9AjdCvhpPnuHSAJYuGezWYRhGIt1LxQKElFuXZ6BpdULHas+OzuLhIHhF1p5ACAFKl/PYzGkpFgswnEcPP3001hfXxeCcKOYTCY4e/Yszp8/D9u2USqVsLy8jKWlpdig6eXlZXQ6nVjAhu7XAiDBErSC8r+cBcavOV+LZKTVaomyRXXNsiwZrL25uSnDxfXg44WFBVF4aKtkD5O289H2mEqlUCqVMDc3B2AnHbFQKMD3/VjqIQM1SH5JOBgCAuwMxW42m0JqOI6BBJ0BH1rpyufzaDabQvBI5AqFAoIgEBujThkdjUZi1dQBNJlMRoaU6/5Ani+KInS7XRmUHYYhOp2OpFtqldRxnNhMM9d1xSaqFTSSyVwuh1qtJsSS95bKGP9+eE8IqnRUJkejUez3fNZ57fwbMjAwONC4Z+oRg7sPD/z1F3HW3g2gOv9Li/i+Y2/cwRXtH+Q+tm0GXt8irknSLMv6D9hpyp2zLGsFwF/Dzv8Y/qplWT8F4DyAz1x++W8D+P0A3gLQA/AnbsOabwhzc3MYDoc3TdJoh9OWRapbAGKWLvZnMdyB5I1KWbfblePS4sXBzFqp031LLEp5LpIDFr26V20ymUhfkeu6WF5ehm3bOH36tBCtm90D/hsOh/B9X1IQq9UqFhcXkcvlJPkPQMweSbiui2azCc/zkEql0Gq1JK2S5Jb73Ov1YvbGTqcjBX6v14PjOPB9H61WC71eD3Nzc6I+sojf3NyUlEomCmoViDPPFhYWZP+0MkoFlIocVTJt/wMg/XuaKFHlox2PCZMa/X4/ppLy3Hqemp4RR7LJgBDuFZMeaakkmeE4ApK78XgspA0ACoWCrI2qog4oYYgJANkDz/PQarVkf5g0yXOHYYhut4v5+XkZQcH18Hjj8ViIoV4Dn33utbZHalB143w7AwODg4F7vR4xuPswDQJA1YeL/+QYnis9dl3vDRaS+O6fevaW1/DU+nFk/3kFTri/yEwmNUGp6uP83/w4AGD5iwMkv/j8Nd5loHE96Y5/5B1+9amrvDYC8GdudVF7Cc/zbomk6X4zncxH4gXshiqw70cXu1Qf9JwqYDeGnT+nMkZSMBsSwvfokAeqNFwDVQjHcVAulzE/P4/5+XkcOnQI1WoVYRhia2vrpoNUqO5wDlZ0eeYYkw4rlQpc14XruiiVShK7TsJBdUQHo9A+yK9JkLj3VLpYuOu9Y+x+t9tFNptFuVyW/QYgyhAJBPdV98gBkMh5Tb55nlk1kPvAvQZ27Yt6MDNJGnsTZ9VQxs8zOp/Xyv+SsJHQ6L4skh6dEMnrIsHSqiywaznk8wVAIu5J7AaDAcIwFIWTVk5+SKBTS7mXDDWhQhyGIdLpNKrVqth+9X3jumbvKdeoQ0P4HgCx13G9AExPmoHBAcK9Xo8Y3P1Iff6b120xKx67D//5/XFCF6Ui/F8ef+G6LJO/8cKHYfWTcFaSKP/vT93Eau9+ZNNjZB/fGctzaTCPhexHgShC5ndfuCLB0uBK3DvT/W4S7N/SKtaNgAWzHjitFQuSMx1+AOAKBYz/pR1PJ/pRGWAvGr/mOamc6EKVpGI2BVBH/7/vfe/DdDrFQw89hFOnTiEMQ3zxi1+UGPZbAW19W1tbMheLvXCO42BhYQGZTAaLi4tCKDlwOQxD2dvBYCAEznVdKfR1zDwVHKpG+ue00DGVkT1LHITd6XREcen1enBdV1Sa2SCQbDYrPWHcb33/adMLwxCTyUQUwV6vF+uh431h8iTJyXA4lOui3dDzvFj/GQC5Rj4nJNWci1av12U/rMvjGKhyApD16rAaWhJt2xbyTwJM0jwYDNBut1EqleR5YmIoCSZVOxI9XhtJ8HS6M+qAA7h5jzlLkH8XURSJbZJ/P0yW9DxPlGiScV4HFVHeu1sJBDIwMDAwMLhZjN++gAf+bDzBMFmt4MKvV5Cwrl1jPfR3GpicPnu7lnfXofAdmwi/AxhNElj8b3OI+n1Ew9FOEqfBVbHvSVqxWEQYhpIaeDPQdi3btoX4Mc2QhSiVNfaEsbhkwc3XMbqfhTJ7lkjOaBPUyopW26hm0bKn+7C08gIAx48fRyaTwfvf/34cO3YMR44cwW/91m/JsOm9hI78f+uttwAAtVpN1CzP8/DBD34Q9Xpd9rVcLosixZh93/dRrVYBAN1uF3NzcyiVSmi322g2m5JCCEB6tli8+74vNjrOZmNYCQt+ktR8Po/t7W0JryAJIwEaDAbo9XqoVCrodruYTCYS68/ER94XBmKQPJBYM+1xNBoJQeTPq9Uqut2ukHOSTtog+RzoqPzBYADbtqUfkvefzwtJcqvVEuUNAIIgEJLIc41GIziOI+MMSII454zz0/isaWsm7Y6WZQkR5p6EYYhGo4F8Pi+EWquGwO6HH5podTodIbX8m6H6zL8X9gMSOlDFwMDAwMDgTmKy3UD4Pdf5/0vR1u1dzF2KdHKKrX9RAFBA78s1LP2d/aki7gX2PUk7deoUptOduU03A6osem4TFQkqFFTXGLoAIGZ/01a+XC4Xi1/XhbRO29MqmbajcQ35fF7UESoLVKL6/T4uXbqEl19+WZQs13XxoQ99CD/5kz+JEydO4Omnn8bXv/51rK3d3j7qRqMRGx1w8eJF2LaNbDaLTCaDj370o2i328jlcjIgm8oMC3zf9+E4jgSVkBj1Ln/6wrh527bFflkoFIR0kARoIg0A9XpdIuSp4EynU3Q6HczPz8N1XUkk1PbFQqEgpJphKNVqVfqt2PvFIeMAhFBzNAGPxURFEpLZWH6mNPIaqLqS5AAQEkXyz0HPfKYAiKqlCXwul0On05H1UOnq9Xpynk6ngyiKUCgU4DiOqLV6BlulUkGn05E1TiYTIXq8b/1+H+12W/aBH3DQHsxr5/XzfSSKWj3VKp4JDjEwMDAwuKtwky0lBwm0g04/2sGbv/JRAMD7/5cLGK9v3Mll3XXY9yTNtm0EQXDTStp0OpVkPkaXk3Ro8sHCk6QOgNj8SMoYm0/lC8AVRTkAKcpJCIHdPiyqQSxUdd+btsuNx2NsbW1hdXUVS0tLyOfzcBwHDzzwALLZLJaXl3H48GE8+eSTOH36dGxW2V5idvQB1SkqLy+88AIcx0GhUECpVIJt28jn83L9JG20QDIIhGTBsiwJHqESR3XOtm3ZQ61o2bYttkXO9NLWSc/zxDap77Oek6f7tKgU6Zh5PgN8nZ5rx+eIx9eWRv1Mcb+0qgrsPNOcp8aAED0zT38owOeLSZJ8FoHdZ48KGM9FWy7TGhmDTzVZP3v8e9C9Yzqtk38Xs32APD/ntlGRJlkDdi29DCehKqpHXhiSZmBgYGBgcG+i4PRRcPoAgIt/9H6keifgbk7h/Pozd3hldwf2PUnLZDJYX19Hs9m8qfcz8l1bylgIazsaP/2nGgHEkx9ZMOvwCh2UwQKWaoe26LF4JkGh+kCSokM4dDx7p9PBW2+9BcuycOTIETiOg1KphFKphKWlJfnZ5z73OaytraHb7aLf7+/Brr87tNr0+uuvI5fLSfT90tKSXA9JLgCZlcXCXyttVJFo56NqVCwWJV6fZJHqjp4xx1RJqpxUKTmPjerPLHHk/eJMOI5lIBHk77W1lfeL91wHyxD6NVTJ9LFs2xZ7JwCxL+rAldmQFV6HTqWklZB2yzAMxcY4HA7R6/WEsPFZ4wcAjOXn+bQlkURL93JqZZF/P/zAgemmPJa+Vn2/ScJJ6G42AMfAwMDAwMDg7oLze3ZGHK6cr+ChN04BAKJzFw90z9q+J2nlchmvvfbaTZM0bZmjBYwWQx1gwECGfr8vUeOcTcbByLZtw/d9KfSpuFCZoB2OJIKR5iRtehbabGBIr9eTopoFcxiGeO6551Cv19Fut/HRj35USE+1WsV3fud34ju/8zvx6U9/Gr/5m7+Jz3/+83j55ZdvOVTkRtHv99FoNDCZTCS0AojHsvP7MAxRKpWwvb0Nx3EQRZH0MrHfi3vEvjTa9yqViqRSTiYTFAoFGW5NkjAYDLC9vS1BFjq5UxNtKkAk7o7jiBrJRERteZ1VgID4SIXZodkcpZBIJNBoNOTZ4HPI1/NekUixb41r4J7ouH8qUHxuuR+0d/K4YRgim82K0kjrYalUQrFYlB497g/Pkcvl0Gq1kE6nJSSm2+1iNBpJuAz3hupoKpWSNE7P8xAEAaIokvPrYdrayqkHXxsYGBgYGBjc21g42kDj7+58XfybJ4FnXtr55gB+MLuvSZplWVhcXES/37/pnjT25ZAQ5XI5sVAyeIFhHpubm1JMlstlFAoFKa5J6DivCoCk53U6nVj/WS6XE+WIRSoj27V6N5tMSCWJPXjFYhFBEODVV1/Fm2++iTNnzuAP/aE/FIuZB4Dv/u7vxhNPPIFOp4OXXnoJf/pP/2lcuHDhPVHVCM/zUK1WEUURNjc3hZik02nMz88LUQCAtbU1Iae0ivb7fczPz4syxj62SqWCSqWCwWCAZrMppIS9ZCdOnMDKyorMI2u325ifn0c+n4fneZIKqUFSQbLFhEqud3YsAmfLcZ6ankVG0k6ylkwm0el0MBqNpF+NISLAzvOobYh89vR6qJxR1SUJGw6Hckz9AQGfIx6r3W4jnU6jWCyKiss5dr7vI5PJYGtrSwgn+/q0/XBubk4I42AwgOd5EjDieR5c1409x+yj44cZtKqGYRhLA+31evJ3RIWuXC6j3W6/5x8uGBgYGBgYGNw+rP3lEabT9yOoO3jwp79xp5fznmNfkzRgR8ViKuDNgEU2rXUsjJkyR9uVjuhn8alDRKicaZVM9z0BcVWExXEmk0E+nxfCpOdSMbgEQMyixp4zntf3fdTrdXS7XdRqNTz66KOSnqgj66vVKj7ykY/gn/2zf4ZGo4EvfvGL+Ff/6l9dQVL2CpZlYXl5WRSb6XSKZrMpRJNqFfeUlsPRaIR8Pi9R9KVSCf1+H0EQiHWOPWpUPvWwbdrnuHe1Wk1ICAdRh2GI4XAoimShUBBlc3Z2GpVVWlVJsJj+yJh8EiLG1VMJJJniCAc+I3wWdG+itk7qa6HVUitLWunjbD3ec+4jyRyTSzudjuy/Hi7O9wGQ0JRMJgPHcWIqHm2ovJ96ppm2UGrrJe2R/FvgBxYAxNY7q1xzn5PJJFzXRafTuS3PqIGBgYGBgcGdgZPZcZqF9sGcibrvSRo/ceen8TcKXayySCUJI9ECdgMydEHLHiEdMqFDFnQ/lC5CeQ6+Jp1OC1HSwQ66yNXncl0Xy8vLaLVaknzo+z5arRa++c1vIp1O4/7778fi4qIoHyRE6XQan/zkJzEYDFCpVNDv97GxsYFXX30Vm5ubYuG8WeRyOVlnJpNBsVgUuxztorrPj9dKJJNJOYZOUyQ55WBqkhwdFw9AkjkBoNVqxayIo9EImUxGSDJVUa5B2xW5Jm1f5P0jiWK/GQnSrKWRhI/knq/RipTu0dKqGNU3YHeuGi2EmuRwfSSC2rbJa9KDrrm/7Fcj6dMDty3Lkg8u+DveD75OpzDyXrKfjGuc3Q+SRh5HD4vXf2t8LvjMl0olbG5uXhFSY2BgYGBgYGBwr2Lfk7TZQIIbBYfz6iQ+KmpULvjpvo7SZ/oebWlM4NMFLVUaqg5MaCRxoPLCY1M5oXJD5YXF+XQ6RbFYxPz8PJaWlmKBE5PJBNvb23j55ZcRBAG2t7fxiU98AqVSSdTGTqcjdstMJoNPfvKT+K7v+i6cO3cOv/iLv4ivfvWrOHPmjASpaOJ6NbCg198Xi0UhYY7jIJFISD8fAOlz4ryySqUiA6C5L8vLy1hbW5N10gJIdYbHJDHh/DMqPrwX2WxW7KQAJJiFSYdhGMb6okg0qELpUBDeQxIsKmi8P/pZ5F60Wi3UarVY9D4JEJ9ZkikGd+gkSR6fISI8PnvWZhMfScK1csuf8fnhMRjEwmeQRFI/f3weuX4SZq1Acp+p1OoB13z2dTLk7N8enwkSRBJK/gN2lNAzZ85c19+zgYGBgYHBjSKRyyFRrVz7hZOJiZE32DPse5KmlY6bgU5kBHaTCQFIH9h4PJZeHxa5OtadRTdtkizgWcBqayRfp2P/x+Ox9LKx74wx/HoEQBRFmJubQ7VaxXg8xuHDh0Vhmpubw9mzZ/Hiiy9iZWUF9Xodzz33HD75yU/iO77jO1AsFuF5HrrdLs6fP49SqQTXdZHL5XDy5En8g3/wD7C5uYlvfvOb+IVf+AWcOXMGGxsbCMPwHRWMWq0Gx3FkH13Xlfh9AKJu6nEDpVIJYRjCdV3k83nk83ncd999MkzZtm3Mz8/Lfuv5YeyBIvFIJpNCGCzLQqVSwfb2tsxlO3z4MMIwRD6fl8TEyWSC9fV1jEYjBEGAbreLUqkU6y9j4AaPX6/Xsby8LCSczwcVTh3mou2O2WwWYRjGyBkJFtc8mUxie6j73jj3jfZBWmJd15UeNK3McV4eP1TgnDZ+qNDpdOB5ntwXKr1hGMpzPRwO0el0UCwWhbA2Go3YMHadeMm/FyqYc3NzAHYGn2tiDuz0cY5GI/i+L8SZe6gVSg7G5v6USiV5DgwMDAwMDPYa7T/wYfzVn/1X13zdS/1l/J8fLBzIkAuDvce+J2mz8eY3C/b9cOgysNsfRGsXVQESrl6vh0qlIv0zk8lE5lVxTbpviOdhCAPDMYDdAp2x9NlsVtQ6qhgspNkHxIHCwE7hfuzYMWxsbGB1dRWtVkv61F599VW8//3vx6lTp7C8vAzXdQHgisK3Wq3i2LFjWFpawnA4xOLiIoIggO/7KBQKCIIgZksDIISUShOwQ26ZeDkYDGRoMckBCWO5XMby8jKOHj0a6/UDgPvuu08GLfPY3W4XzWZTQlM8zxPCpFW3WSVG7+1oNEK5XJYetzAMJRyGZJp9UiQg7BHUEfYk1VRaSdY1Kddz13jf8vm8DM7W9ldaO9mryOeP5Iv9eLQL6tRQPZdNPzNa6WQiJoNYqKjygwQd908FlxbZ4XAo18dezcFgcMW+0s7I5EwqmlQM+YEHyTFJnn4WdegKr7darRqSZmBgYGBwW3Dm578dP/p7vnZdrz2VXcPrTx8CADz/bx/B/D966nYuzWCfY9+TNBbFtwKqFiRBs7OygN2ABxIm2sNYVOpeHE0OqKyQhFCN4df8noV+FEUIgkAICAv5brcriggj2HVMP4mg67qo1Wpot9vodruihm1sbODMmTP42Mc+hpMnT8K27SsK30ajgY2NDaTTabiuKwojr40qGIkLVSwAEoih+7qy2az0pJHADAYD5PN5lMtlVKtVVCoVUWFIVHg+kqNsNotSqST9XVQcSZj4cyCelEjFR/cT0lLJvWu321hYWBDSoZ8lkopCoSC9hbzf2p7Hdes+KgBX3GuSfW1RnF0b1aPZHklerw5X4fv4Wu6DTqLUFs1EIoFCoSAfKFAhfqeUTxI0TcxJunjvSfqodHK0ANU+PWaB+0Riqp8VTVhJ5HivHMe5JbXcwMDAwMDgnTCdH+I7vTev67UZa4IfmdtJIXy69Mi7vrb3w9+GoXflB4yFc30kvvLCjS/UYN9h35M0flp/q+j3+6I06DlpBItyqhAsdHUypO4n0oOH9XFY5AJX9nSxkKfdi0RxMplIoEe325UZWQAkEIKFczabxeLiIjzPw/r6Ora3t7G1tYX19XWcO3dOotkXFxfhuq4EbgwGA6yurmJlZSU2q42pgOwvoj2T6hiJTBTtDIvWCk4ymRQVhIU8VblqtYpqtSp9a5xDRvueHt4MQNQ/qju+74sKpMmQnh9GdZTvIRg8MhqNJDWQRETH3fM9ei4YiTkJKX/GUJLZe0sCTtsqCTbPqUmhjrvnuXUQDZ8HfhigyQ33n+mKfB51oAiPT2JLMt5qtWLKr1aQube8L1TUSAr1XnPvbNtGLpeTHkj+HfBDBt2LOft3wueaNlE+HwYGBgYGBncT+ocmsB5/GNFzL8d/kUgieeoE1p6wMCmNrnhfWLOxvH4Ck9Nn36OVGtyt2Pckba9ASxeVFxILQhfLJFqcxUV7H1PqOBOKZI/H4rBmz/PElkbCwIKd52m1Wmi327ImYn19Hb7vY2FhAYlEQhQ3Jvy5rgvHcaR37cUXX5RB0s1mE1/72tcwnU5x8uRJHDlyBPPz87AsC+fOncPGxga63S5s25aeNQBicet2u7IXtGtS4YuiCFtbWzELnOd5MouMhTsDTkgSx+NxTCmxLEvWoEkr7YwMJqHippMA2UPI3i/LsiRCnusiSSVRGo/HaDQaYnEdDAaoVqvSp0Wbo07a5PNCgs57xsAMy7KkB473lveI10hSNKvE8Xo0iaE6qG2IegwByddsuA2JKMlVJpORwd8kY7ZtxwZa0/Lr+76QJPbm6VAS/eEC1b1erxdLk6R1k0Q1mUyi1+uh1WrJhw76ww8ei/eJ7/M8zxA1AwMDA4O7Cv/k934W/+ThT2L4ey4H2I1HQBQhWSnhtb9QAqzxVd/XfWiIN5bmcfJ/Og9MTWrxQca+J2ksOG8VQRBgMBjEAj50XDuLxTAMxWrG2HyqCEwBpLLFIpoFMotsWr56vZ4oeJyHpfubqCIkk0nMzc2h3W7DdV0J1eh2u9ja2oLjOPA8D5lMBkePHkW73UYymUStVkOhUMCXvvQlBEEg5Kper8O2bfi+j4sXL+LUqVP48Ic/jCiK0Gg08Prrr6PdbiMIArFvLiwsxAI1+v0+qtUqms0mfN9Hp9PBm2++ifX1ddTrdSSTSZw4cUIUQVrXqKJRhapWq2LhdBwHxWJRCAvJCZU6ksVSqSRkWNtIdd8V181+KxI3khLO6+p2u0Is+DxtbW0JMdDJnlRMqTIBuySDih0JeK/XQz6fBwBRtHzfR7lcjs2+y2QyomRxThstnlQ5u91uLLxGpyaS0PX7fRQKBSFpeg9IMAHIM6oVXq6THy6QYFuWJUmfWllj4qNWAfmzVqsl89Xy+Tw2NzdlnfV6HWEYot/vixKdzWZjiajAbpgI7xGVNwMDAwMDg7sJf/Lwl7H1YgEA8Nm/8kPINkd48w+nAOvd69JpfoI3//FjeP/fOG/SIg8w9j1Jo4J1q7h06RIOHTqEYrEIx3FQKBREJQJ2SBzJjyYPnIWllRTdIwbszoECduegkbTp2VtAvHeHhToJRD6fl4CIXq8nIRckMdVqVaxoiUQCruvi5MmTeOSRR0R92tzcRLPZRLlcxqFDh3DkyBF4nidWxfn5ebE6knxSpeK1ABDVhT1io9EIjzzyCC5cuIDnn38e29vbQhp0ND7teAwWYWIgj8lACypl3De+joQjmUyi2WxKzH4qlYLv+3BdVwp6z/PQ6XRialK73Y7N8BqPx9je3o7Z9EgGGcRBdYeJhoPBQMiehu7b4rgAKpwkGt1uVyy6JDokfiTRtMzyPblcLtYDyb3S8fjcM65hNkWS9lO+hwPA+XPbtuV55NqpAM4OV59Op3I+PhM6nj+RSGBubg6pVArdbhdBEMgsw06nE+vro/LI50jPv+N7GD7DAeQGBgYGBgZ3AzLWBEvpJgAgSgJR0gLS1yEcWBGQNh8+HnTse5K2F4iiCPV6PVY027aNdrsNALHUPwZcsIDVdi4gbmUjUSPZYFHLol9H9fM4JCCzgQ+0Q5KwUYHTaX6O4wips20bxWIRR44ciVkx6/U6NjY2YNs2KpUKarWa7AMVoUrlOmaFKLAnq1wuo1KpIJPJYHV1FcPhEI1GQwp89pKVSiUhKizASRC4J1QdeW3cN+5FKpUSdZDv1SEaep7ZbO8TlT2Sj36/L6MP9B7rQBiSIhIn3j+C904HemhbrO6104RMpxvyZ7o3i2rTrGqlh2TzWnl9V9sLTdL4X/bMUZUbjUZCEPVzz5AQ3V9HKy+/5nqDIECr1YLjOCiVSvKM8hj6/XrG3NXAns/hcIhyuQzf9w1JMzAwMDAwMNgX2PckbTYU4mbRaDRifUPsA2Ohmk6nxb4IQNQhFuvvVLzryH4d6a5TEakEacskyZoOZqC6BkCKac6vYs+QbdtwXRflchm1Wk2sbFzLoUOHcOjQoVveL2DX6sfz5/N5HDlyBNVqFVtbW3jttddEuWKx3e12UavVpFeLtk6qaiQ0LO5nB0GTePDaORSc6txsIiftfnwvVSI9PFqrhjw2ky1JWIIgiAVypNNp+L4vBKXf74s6NpvKyGsiQZolkTokRRMsgsROE8hkMinhKXx2uBc8tyZ8eiwBCRZBCyufK56D/ZKz4Sr8PSPy+X0qlUKz2RQF8oMf/CByuZz0CrJPj3vCDyN0siPBvw/+zdRqNTSbTVmLgYGBgYHB3YKng5NI9m9eGbNGU2y3PFRL/h6uyuBux74naSyubxWbm5vodDqSkEjLI1UW9gqRyNFu1+v1UKvVhCDMptRZlgXf94UQ2LYthE73CzEkhFZJkhj2pLHQ1T8bDofo9/vyns3NTViWhVOnTqFWq+Ho0aO3vC9XA/chCALk83lkMhnp32PYxH333Yf5+XnkcjmcPn0a586dw9ramvTCeZ4nClin07lC6aJtkz/rdrtCiknItMrIIdokZLqfSkf1t9vtWO8VSZruE2u329Jfxl4zBmNQvWLfGM/FtEcdgsH/cp20UuogGT0wXZN0bbMMwzA270zPlNN2SVptSbho/+R70+m0PMdUPwHERjlo9YvDzre2toRoakLJ62IPWhRF8H0fvV4Pvu/H5qDl83mUSiXcd999CMMQjUYD6+vrkvaoR1hoYk6VuVQqSbiLgYGBgYHB3YJ+lMY3vmceueazGH/qsZs6RvTCK7j/T2Sx8avHkU1fPXDEYP9h35O0vVDRCCo5TLhjIc1P/zlUmQUtCRvtXISORqeVjMdiLxmJHotqkjUSHVrxdOw5h0OTdPT7fVE+bNsGANTrdTz44IPS27OXoHIDQMYCjEYjGQCu0wUty0Iul8Pjjz+OfD4vpAmArD2RSMC2bRQKhZhaxL2nBa/X64l9kH1UzWZT9pkkloSMVkqmIDLIg/vGfyQDvV4P3W5XCApJJAAhECRSJGgcSM77qpUs3rdWq4VCoSBkk2mLuieLCZFU1QaDgVhZbduGZVm4dOkSCoVCTIkLw1AGY3Mgt+61497Spsv1AbuJkiSr2nbI1Eju4XQ6xcLCAlqtVmy0BNfPvw+G35B4M4CFfZTJZFJCZGzbRj6fRz6fx+nTp9Fut9Hr9a4II+HfZKPRgOM4sZRTAwMDAwODfYWp6VE7aNj3JG12gPCtQKsZOrCDv2MRyv/SrhWGIYIgkIKbRTm/phoDQHp7WBzrIdp6tpa+Ph5HB43wHCymtU3PcRw4jrMne6JBEsLQEM5Qo32z3W6jWq3GVETHcXDs2DF0Oh0ZsK17vkiWNEiIaWUlMaPqokM1ODohnU5LwAYteVp1ohI0mUyEOFAl1XbH2SHR+r7wftAqyfvE54CBI8DuTDNeH58NbevkBwK6H41EiBZGHgNATGHSsfZ8LgiSfU1EtVKpbaVU6Wb/hvS4gXw+L+MT+DPuEz9MoI2Xx+HfCf+W+IzzeQYAx3Fk3AGVU030+X0QBCgUCiaG38DAwMBgX8DqJ3Hkv0aYdrp3eikGdxD7vqrR85puFboI1UN/gd3ZaCwydWAC1R4qProPaNaKyQJ/tpjX87H4MxbP2vbFQlXP49I9biRGVNb2ErRVkqjkcrnY+XX/lU6GrFarOHr0KE6ePCmJj/raWPATDGRhkU5CCEDCLcIwlO9pq8tkMjFlazZVUw+4ZrJgv9+H7/uyXpIpvof/tI1ylkQBcVJDkqTJD0c5kDxSbeL95rGoQJIEUu3i/dXPhw4dmQ0QIYHjPdC9bNoizEAPEsXZYdbAToIqia2OydeBObxnVKHz+XyMaOs95bEZUsOkTP496X5N3hP9cwMDAwMDg71C9mwO/3bz4+/Z+RJ+CvkzSeR+81lMLzuMDA4m9r2SpovJWwXjvtnn47ougiCI2Ql1Aauj0zksWvcA8b8c3KvT9ahK0S5oWRY6nU4s8pzkzLIsKXhJajqdDorFIgqFgpBJFvSzKuBegQoVicLscONCoSBElf8YXLK8vIxcLodLly5dEQIC7M6k4zVoMsW9piLZ6/WwtbUlBIg9bOwd47nZX6ZJMwnOYDBAEAQIwxDNZlMCWGgL1PHytEzqqHlNkmgd1BZZKlm0XAK7RJ+22dmQEz1gmn2MWpmioqrHE+jky6tBh41wkDoDV3K5nMzCm7VKkthx7fPz8+h2u0KOZ628eoA54/xbrZYMQM/lcnAcRyyvHOtQLpcxNzcXUzH13EHup4GBgYGBwe3AfX/9KZx9/duBn336ht87gYXWxAGi669DS69ZmPulp676u+E4iXRqgoRlrI8HAfteSdNBCbeKjY0NNBoNIR8kZuwHYx9WGIYyX6vZbEqxX6/Xsbq6KoWsLnKp+Oj5VrZtY2FhAa7rigph2zYcx5HeHR0jn8vlYopQKpWKKRDpdBq1Wg3D4VCI4V6C1wUArVYLuVwO/X4/pur5vi/WTx0jn0qlMDc3h09+8pNCasfjsShZrVZL+tQYyd9qtbCxsYF2ux3rLePw7Lfffhvr6+tCsgiSJhIEBoiQJFD5DIJASM/W1hZ830c6nUav14PneaJ25XI5tFqtWBIhiRbvM+eG8T5Q+aNaR7vrdDpFt9uNpRvOqnU6cIM9eu12W66RARokPFp503ueSCTQarXk3L1eT4iPDikh+eK9YL8dZ9BZloWFhQXUajW4rivH0+R3NBohm82iVqthfn4erutiPB6j3W6j1WohCAJZK487q8QyjIZqI4DYPDmjpBkYGBgY3E34ldXvxv/3IycxabVv+VjRaIgjf/QtbJ+u7sHKDO4F7HslLQzDPUl3BICtrS1sbW3F+mNoXZu1IVKtYujEdDqF4zhot9tCrvRrZ8MjSESoWOhocfbosNDWBSt7lMrlMvr9vpyLRfrx48eFVPR6PZw4ceIdr5dq0vXORSMZAIBCoSA9VdPpVEYC6Kh5nW7Ie3T06FFcvHgRa2trEjwymUxQLpelR4oz0ACIKhMEgRDkTqeD9fV1CWHhfSIxpEI5mUxQLBbFcjcYDFAqlTAajeS11WpV1JpOpyPEbjAYyL1vt9uikgEQFYvJkgBifYVUQElmPM8TuyPDX0jCbNuWn3N4OUM69Nw1EnHuNZ8HbX/VfXi83kKhIFbUbDYrPWgMHxmNRrIGDv1mQAqf+0wmI3bGWbsk1b5isSjqMNVgfnjCnjz9d6r/rrSNlh9QaIWc6Z1U3zgb0MDAwMDAYC9Q/tLb+Ln/4ccBAKf+xsv4g5Xnr+t94ygRsyxmv3UOD/Tvw+k/MVN+Ty2c+pU+rNEEie0O3skfEg0GgBHRDgz2PUmbtY3dChhsQXWDn/jTSshP9akC5XI5SaXjYGU9K8p1XSEZBPvaaEFjcUyrG6HDNVg4AztzrHQMOteVyWRQKBSwtLSE6XSKVqv1jgmPKysrQuKGwyEWFhZw5MgRUU3eCTq4IZPJyLlZ0JO4ALs2VKqQ/Hkmk8GJEydEDSOhospINajX60lEfhRF2N7extbWFoIgQLfbxWAwQBiGomCNx+PY91SvSqVSTGnlnrD3i8O3SaiCIIBt27Eh0FThuDfJZFJSNrUlUSdb8h5TYSLBJgHi60hAdRgJlSPu4eysPT5HHENAdRaI2ybZl8e/ET6/fB33XPej6eAOfX4qpa7rotFoCLHVg9f1PD99TH4YEASB7IUOMqHirIm9fua4Vj7jW1tb7/iMGhgYGBgY3CjGa+vIrq0DAL7y2BN4+5NV/N+P/u4NH2ey3UDyhT6qD30o9nNrCuCF1zAdj/GOxshEEv5//zimpdE7vcJgn2Hfk7S9tEAx/IOFquu6cBxHFAKm5ZEQ2LYN27YRBIGoG9lsFvV6Hb1eD4VCAdVqFbZtx8IedEw549F1gc+iVdu+WPzy9ew9owpj2zaOHDmCw4cP49KlSxLLrjGdTtHr9fDiiy/irbfeElJTLpdh2zbm5uZuaBaVTrIkAaCypi12OoY9iiIcP34cFy5cwNraGiaTCYIgwNbWltg3LcsSyx2wYzddWVnB2toa2u02wjAURYl2wjAMRVEk8dNqJPdVjzhIp9OYn59HoVCQYdW9Xk/WoMkX95r3iMohyREJK4kdyUcQBHI+27ZF+aXipomfVslmA0WoZvF9URTBtm0hx7SD6sRJTdj0BwxUHvlM6q9J+GbJdjKZRLFYRL/fx8bGhqRikkTTQsp91dcwGo3QbrdRLBaFmGoCzGeZH0YwJZJ7y2N4nodKpWJImoGBgYHBbcORn30Kl/pP4Hf/yAeQsCJ8Iv8GMtblsK0oiS913ievfbtRwRLWYu+f9nqo/vMr+9veTU6w0hkklw+h98faWDBz0g4M9j1J0+EVewWqKCRr2t7GfjIW34VCAQCkD4ukicN8af3je1nQ6mKZ59H2QFr5+J5CoYD19fVYSAgL3nw+j6WlJTz22GOYm5tDoVCQ82u0Wi184QtfwMrKCur1uuzbhQsX8PTTT+Pxxx/H8ePHr3ufSGr4NcNWWq2WKIndbheO4wixoAI1NzeHhYUFtNttdDodCU1JJBKy9u3tbfT7fXQ6HZw7d0767HgcKmCNRgNBEGBpaUn62lKpFBzHQbfbjSlOwE70u+u6cp/n5+eRTCZFpdNWRdry+LVOpOT9ASDEi0Se18nnQ/dvsfeNqivDX6iyUX2l1VETdR6DChwtkrqXi4SLs9NItPjM9Pt9IXYkW/zAgQSYA8rZb8d5fPl8Xu4bj0Vixf3gXvu+L887X6dJI/fVcRxRW7WllGvLZrNC/vQ8QgMDAwMDg9uBwz//FF7+eQCWhaWX5/G+7A4Re2NwCC8/HgGXP6Bewit7cr7o0VPY/ushsu9ohDTYj9j3JG2vrI7EcDhEp9ORYAPP8+C6LprNphSQeug0ABkaPJlM0O12peBkEdvtdlGtViWMwnVdALvEplgsIpfLYXt7G77vyywx3ZfD/iIikUigUqng5MmTOHLkCGq1mgQwlEolZLPZWHjISy+9hJdeegmrq6tS3AM7Slc6ncbFixdx6NAhVKvV2HmuBiplWjkisdHzwDzPw+rqKkqlkhCIXC4H3/dRLBZlpprjOHj++efR6XRgWRY2NjZEnSRR2dzcjAWx+L6PUqkk6ovu+ePcNM/zpMeNe8No+FwuJ4Rbh4BQ8aMFlVbSUqmE7e1tWJYlVs9Zax5n31G107a/yWQiRIXjCzKZjCSK8nVcP699MplIPxxJIoeB61h6qlbaVknbp1YEadElSMA0marVanK/EomEHCOTycB1XczNzeH8+fNynOl0ik6nIx+YkIQyjESHi5DUclwDZ9glk8mYMjgb79/v90XZNjAwMDAweE8QRfjdH3gEv2t9SL5HdOGWD5tcmMebf39p9/vUBOVbPqrBvYZ9T9L2KjSEYEGo1S8W0LMWNE0QWcTyGCwwWbxGUYR+vw/P85BMJkWRWFpawtLSkvQ5dTodbGxsAIDYBLVNksU9vy6VSvA8L2ZT5HBtWufOnj2LM2fOYH19XWx4JBH8nmEctVrtmiSNJETH8DOIIpvNyt6QIHW7XViWBc/zAEBIEhUS13Vj1r/t7W0Ji6D9bjKZSNqmtuSRuLCnidY5Emgem/eIqo0ehaBVUsbvM6CCFkOST9opqTZpi+DVRkFQEdO2Sz1/jVZYHTNPRQvYDQzRVkKtvvLZGAwG8mxp5Y17x/PTFsr36UHXJMya+HEN3AfG7OfzebFXkmhpIsr7RIKmVT79DPF12pqpCS5fz//qvkgDAwMDA4PbjfHbt07KiM4f/Xb4ywlMssBceXPPjmtwb2Lfk7RZsnSrYMHLYlpbKals8ZxaKWEB7ziO2LRYfHO2FwCxL7qui8OHD+N973sfDh06JOfo9XrY3NyUYBKdjJjJZKRH7p1CQQCg0Wig2Wyi1+shDEO8+uqrWF1dRb/fFxKjrZa8tmazia2tLZw6deodj60Ld60mskin1W08Hkt8PdU6qlfpdFqGHjOqnT1bWlnRe65treyRYuFOokfFjEomLX8EyYLumaLCo1Uqfq3vse6R06oWCTvvkw7qoNKl94Y/o/1QB2jomXP6tXwu9QgGPp+zYSR8vw4O4e9J0vgavQfcHz0KgIqxJkZU4wqFglgedb8bP+DQATncc31O3fOn/8a0LZQEWAfn7NW4DQMDAwMDg/cS0cc/hPVPTLFwtH6nl2Jwl2Dfk7SrqRe3ejxGk+vBzfydJoTD4VCIBwMwyuUycrmcWMeYDqnVhEKhgOXlZRw9ejRG0ICd/pxjx47d0jVcunQJ586dw9bWFt544w2JSM/n81J8UyViIInruhgMBmi1Wu96bB2brokUh0ADu0SOFjZG26fTaZRKJQCQ/rVyuYx2u41kMikqV7/fR6lUkt4+pjT6vi9Fvx6qTIWMQ8ep5jHchSrXbGgGgNgMMwZ8ZDIZOS+VRto7aYfUfWuaTOVyOQkzoS2S1kNtX2WwyXg8Rrlcjo1p8H0ftm0L0Z1Op3LvNFkhEWN/2exwbap4msRRSUun03AcR8Y4cEwBr59Eknugg0lSqRRKpZJ8EMB16rEFJJOcg8ZURyqDBFXQXC4XC2fRQTO6n+52DGk3MDAwMDC4HbBSKVipFJBO49JfGmIhE9zpJRncRTAVzQ0iDENsbm7GQi5s246l8JHY0Iamh0mPx2OJttd9TgCEpFQqFdx3333XPZ/sZsA1LiwswPd9KXTL5TKazaZY16j86DCTdwOJK5P9gLj66DiOJP1xthUVKj02gAPB8/m82DKpuARBIDbFdDqNra0tKfA5gsC2bbF4ZrNZUXh0Ic/4/9nhzSQTAESl63a7EjEPQPr5SMx8349ZJz3PE/JFMsq1kThxNhl79Xj/wzAUpYn9cLNkJAxD2bPhcAjHcWL2QpIyPYuPr0kkEvA8TwiaVu8KhYKog3oswmxvH0kVsNuDych+Wh6r1Z2Bm61WC4lEAp1OR5JCU6kUarWa3JfZ8Q48pk5ypOqnX0uyR0JsSJqBgYGBwb2CjZ/5GKzv3wYA2CkTrW8Qx76vaPba7kjLFu1YtAeyYNV9Mnwtk+20tQ3Y7bvR9jESFKp1s9je3sarr74qASCLi4uYm5u7oWuYm5tDJpPBww8/jEqlgul0ihdffBGnT5/GeDyG4zjS78VeuH6/j2q1imKxeMN7NhwOEQQBcrmckFNNqnK5XIwA6blYVGGOHDmC4XCIfr8vayPR0/1JVGw4t8vzPHieJ8ob9zuZTEr4hk4cZLFP1Y+piFQ5tWVPD8pOpVJCknTSIJ+T2blmVBqpOgK7gSs6eITKG89PEkI1UFsWdbKnfr7Y70gVis9ZJpOR8QRMcNQjAvicksjp0JLZ8QCz/ZcknMPhUJ5lTfrm5uZQKpViBIyz3bhvjuOISsd/DEnR6qQO7Nlr5dzAwMDAwOB24MJffwLjB3uopvY2O8Fg/2Dfk7S9BsM7ZiP49aBfFs16ThZ/TwULQKwgJqi0cZi0/vnq6iouXLiAM2fOYDKZIJ/PY3NzE5VKBVEUiXWyUChgcXHxHa+hUqlIiATVDs6jOn36tMx9A3YIE9P4KpUKarXaNfeIZIOEQUfDU70BIBHvmjgBu31QlmXFRhlQiUomk2g2mxIVT9VPkyeSWPbpua4bSwRkkqQOr+C9oPJGuyeJEMk4lT3dX0YFTgfVUB3jNWkSqJ8TEhWqUY7jCOnU5JFfk/TpyHqSKD5LHPXAvdDKJsmN7q3kvZkNwSEB5fdcE6+BryUhJJHWASj8miQrl8vJNWoCzLXpPk7+zVBF5bOj/+Y0SdvrNFcDAwMDA4O9RCKXQ/O//zCmD/mo5nvXfoPBgcW+J2n6E/e9AOdzEZqksYBlsUgrJImZVti0KqIVM6YWbm1tYXFxUXqH2u02XnrpJZw9e1aiy5PJJN5++22k02lMJhMUCgXkcjksLy9LiAR74nS4Q6lUEsWJOH78OAqFAlqtFoIgiCUOJhIJ5PN5HD58+F3JH4BYKAj799iPRlVOF9rsNdLWOU1O2u02MpmMkCUqUHpgMY/DhEP2nlUqFSEn2WxWZq1xPY7jxOZu8VnJ5XKxwc4MOOE6B4OBhLYQJJya5JCo6/AOPRuP95vPEW2MtGlqksP1ELPhJlrVsixL+vj0c8UeNRKeRqMhJJYDpZneSFumTlPkXuhQGpJJnbKp59VpSy+VTYa4aFKlh2Pzmvk1LafsodN9a7q/jYTVwMDAwMDgboXlucD/tW4i9Q2uiX1P0maL6VvFcDiUIbwMiKC1LggC6TVinL7jOJhMJkJ6tB1LF+iz6YeNRgMXLlyAbdt48MEH8eyzz+LixYtyDl008x/T9M6ePYuvf/3rKJfL+PjHP35FfxvHCHCNRLVaxWc+8xn823/7b2VodLfbxdLSEo4ePYr77rsP5fK7/88K15XP5zEajeD7PjKZDPL5vNgeSbR6vR5c142pJiRAvV4Pvu8LSWs0GkKCWLCzB4rFO+9JLpfDiRMnJAZeJ00y6EKrNSRAHH1AEsl19vt95PN5ZDIZuX/si+Jr9RBlJnby/GEYxhRTEnSmVPK9jNYnOdKkSg/D1uSJ18D3UT1zXVf61aIogm3bsVED+tr5YcFgMECxWBTljqEheqwAn9XZhEcdnc/jcp9JziuVCubm5sTWSoumJrG0mQK7KZ2837lcDsViUf4GeK/Z06dVUmN7NDAwMDC4K2GZUTEG14d9T9JoidsraMsasBscwcAEqgyMjacCpIcps1+JRTCLa6peURTB932cPXsW9XodL7/8MlZWVqQ/p9vtIp/PxxSMKIqwvb0tc9Jotfud3/kdmbl28uRJsUb6vo9WqyUz1HS8+4/8yI/gtddew/nz57G5uYlqtYpPfOIT15yPBkBmwUVRhFarBdd1ZV5YEASSAqiDSHSUOgnDhQsXcPbsWSFIHOpMBZDph1TNut0ukskkKpUKKpUKyuUywjBEp9ORoeO0dpK4azWG6Pf7mJubw2AwkP4wrhnYDabQoxZIhkja9bwxbZVkOiTJnbYXkpDpQeZ8v+u6qNfrEuwxGo3gui6m06mkT3JwN589rU7qOH9NZhhcotM32+22jHAgAaXlkXumewqZdMm+NT6jOpTEcRzYti1/G1QrSSBpl6QKq0NB+DXPpdUyflgRBIHMk8tkMlhYWMDa2to1n1UDAwMDA4P3EsNPfxSrf3yACozN8XrR/co8Tv3KGziInXv7nqTNRnrvBfTMKg0WmSz6WVCymKQiwqKY5M51XZRKJZTLZTiOIwV8v98XsqNVCSYH8nuSNNoUgZ10wV6vFyND0+kUtVoN5XJZCNHKygqOHTsWs7DlcjkcP34ctVpNgk/y+fx1DQrmHvT7fSEcOu2S18/kwsFggEKhIOSOCtvGxgYuXryI1dVVFItFNBoNOU6z2RQrIe2H7G2iekniUyqVhHTQfsprpXVPjwogAdSKGYM1uNdMTuS1UMHTx+LQZpIbJlnqn+lh2yRuHEmgB3DzOaHaRHutTn+cVY40QdQpm3x+aeHUvW36fQCkz4/Ek2vVNs7ZfjCukz1xnEPHVEZ+rUmdtiTTzqhTG7k/et8I/Tt+XywWZTC7gYGBgYHB3YD2H/t2bH4UmC8YgnYjSA6BSX372i/ch9j3JO12Drflp/96uPFsj5EuZC3LEhWO7wEA13XlH3un9BwqzltjYa77n1iksreIEeu0kFHJGwwGaDQaYhGsVCqiLF0NpVJJZpbdKLgPnK1GEqBtbSQYtIpqlaVer2Nrawubm5tYW1tDt9uVPQF2bHl6f7k/jN3XyYmO48jP2LtGksIB0CQ9vDfaEjlrESRB0RY/km+tktHOSCLU6/ViYTO8FpIUPickJ7o3j2vh86IDS9jrpQen6+Pp6PrZc3EdtChSHdNpjvp54vu5DpJ//oxkkR8MUB0GdvsuNZHT0f+8Pq1qzgaB6PvB7/XfEdfCcxoYGBgYGNwtaHzAwvwDW9d+oYHBZex7kqaL270Ei0dt1WPxrOeezcaw6+ADxq97nhebO5VMJoXA8DVMNuRgZq2esNilikRLIEmbLmQ3NjZgWRaKxSIeeOCBPd8XkkgAQhD5s06nE7N+RlGE5eXl2Pun0ynOnj2LS5cuCVHb3NwUEkRLaTabFcWGZEjvI3u9qHBls1mxXvq+H5szRgKs75lWuniPGGVPggTsDt1utVpi8xsOh/A8T6ya7GsLggCFQuEKMqIJC/u9eA20KOo1cu+0arq6ugrXdYXs8rnja0kcta1V9/hpRU33rPG8er25XE76LHVEPtVDpkbqnjmOLpgNzNHPMF+v/151IiWJo35WaMWkOscY/9nUVAMDAwMDAwODewn7nqTdrkKN/Ty04Ok5WCwYm82mEAMmHbKABXYIDe147K1hhD7VHRa3DOJgQMjsTKhZOxjVHe4BycN4PEa73Y7F++81JpMJgiCIqVMs/BcWFpDL5TAajVCv16/Y0zNnzuDcuXMyt833fRngnM1m4XkearUazp8/L2EZuhdNDzouFArI5/NSxDMafjawhdH57PXq9XoS2OE4Dnzfl9h9jkjQJIp7TwWR5EUHljBwhXZBklUSMCpRtAnyviWTSUlL1EocrbAkVwBEubIsC51OJ2Zh5H912mQYhkLSeK1a8RwOhzHVkj11qVRK9oDnZe8ng1x4fL6PymK32xUC6XlezBo8mUyQy+Xk2rSFkedmnydJG5VL7isA5PP5PX2eDQwMDAwMDAzea+x7kjY7u2qvwKHPehAyo9tp4QJ2gzRIsHRPTy6Xw+LiIsrlshSyJBks9Kls8FhUQHRkPPuxWCzzXHp2F8mSDmi4HWCxTQKh+6L0fLhMJhNLihyNRtje3sZTTz2Fl156CZubmzGVrF6vo9frSaDHmTNnhLTxn57npa18wA5xbLVaYj/UIRg8f6fTged5Qiwsa2dOG8mRVp04TJrKD2fLkaT0+30h8XrvE4lE7P7w/mm7J9VP/o73X4fUaFLOoBCei0RUExcmZ5Kk8VnknjEYJZPJyIBwrVJxX33fF3sqSeashZMfVnCvZtepLZxcN4kWR0vweFTfuHckwiSWei9IkHVCpIGBgYGBwd2MjZUyTv1SDxt/dYJs+vbUZgb3JvY9SdvrdEdCh2DoQb/8HQMmSqUSxuMxut1ujHBls1mUSiUUi0Wx3bFYp+WNx5qNb2egBAtdvo9FL1/H4pg9QhwkTDXkdoDFNMkFlR0qNZlMRvZA9w5tbW3hlVdewSuvvIL19XUMh0OZGTccDtHtdgEAQRCg1WphMBjI/pGk6T4xqpMs/rmvLOr1/C3dk8XB3SQEswOeaRvUKZZhGEp8vu6v0rZWnkPvE3+mexkZeU+yzXPqfjeuWQfT6IHgWsXjGnXfpA6+oTLFNfD9+hnSJE2H3nAf+aEDr08TZVpNJ5OJDLDWyZA6wp/XqfsNZ59TroHH0bH7JLGanBsYGBgYGNwNWHxmgu7KPMYu8OM//jn5+X8rHcEzP/IQfuL4F5FNjGLv+a1LH0R/vO9LdYN3wL6/8zpyfK/AIpbFpQ5MIFgwFgoFmfk128dUKpVg23ZMkeHMK62sUJFi8cnZa3rYNAtWFuu0nJGcjEYj5PN5uK6LQqEQO+9eQtviAEhUvA44IWiBC8MQp0+fxjPPPINz584hCAKJlu90OpJcqfvGHMdBrVZDtVpFNpsVksRwEN0jlclk5Hr1veF+6f4rWgC5LwwX4f3WPYUkQt1uF4VCAWEYxkYt6OPoUJDZNEJN0gBIPD0/YNDJiKPRKLaHJFDpdFqIHPv39HrYz0VlTfe96d41/TOqfvwd16vvH0kkz5FKpcSuSJLlOA76/b4QMxK72Z5NXqdWP7WlWFuX9b5qwksCa2BgYGBgcLfAeuwDcFZDOKtAfz6H/7lyZveXlTPAsSev+r7X/UM4260iHKWv+vv9jq1GHuX2we0v3/ck7XZhPB6jUCjIAF6qNQyXCMMQw+EQm5ubUshqm6Nt2zJ/S1vv+BoAcsx0Oo16vS5R8uxJIsnQRay2xwVBgFKpJMX54cOHsbS0hEKhgLm5OSmU9xqMugcQIwmziKIIq6ur+MpXvoJvfOMbePXVVzEej1Gv19FutyXVUSOTyeDw4cMoFovyXxKLzc1NzM3NYW5uDvl8HrlcDtlsVvaNvX5aoWHE/nQ6hW3bcF03NvdL9xjq6/F9X5IhSTay2azY+hjvTwJGKyKfF867I3njfWi322J/pf1PK0YAYn2N/HCAvXRUsBiwonvwHMeRtFCul+mNVGg5BkAPYOd1k9QPh0M4jiPPNRVHrRZyThxTR2kBpXrG1E8Sq+l0il6vhzAMAUBIJ4/LfXccRyypXBvvQyaTERuqgYGBgYHB3QArncE//fVfwvG0d8Pv/Rf3fRU/W38ffnPl4duwsrsfJ//hGHj26Tu9jDuGfU/SGAKx12BCH/vIaBmj8sD5Vb7vxxSJXC4ntq8wDOH7PlzXFQWM9jISL9rrbNuOJUCmUikJArEsS2xyw+FQbHIkCXzv/Py8zESbTqe4cOEC7r///j3fG42rkbPxeIz19XU8++yz+OpXv4qtrS2srq7i4sWL2NraEtVsttgmkSgWi3j/+98vcfqDwQC9Xg+lUglzc3OoVCoxqx2VSJIMPg+e58n9ofJFCyRVUm0N1fbEWcsh+7h4zm63Gxs5QMJClYlBKLRzAjtWRypovKcARH2iYjscDmHbthAdx3GkT06HfehB6lynju6ncsaxDExtJPFjUiNJJb/nOvmBA7DzPOo5fQwh4QcWtDh6nhcjZvo5zufzaDQaABCzVOrRCRwLwJ60VCol4xlc14XjODKfz8DAwMDA4E7h7P/74/i1H/kFAMB9Keemj/M/V1/Cd7pv4i+/8cN7tDKDewX7nqTpmO+9wnQ6xdbWllj5gN1YcgBiQ2TABwtencSXyWTgeZ6EMLAg1a8jWdNDsLXqwQQ/27aloAcgRTCLVr7edV153dUGIBP1eh0XLlwAANRqNZRKJVErNOkiCWDxTrWP15xIJOC6rpDUVquFra0tbG1t4cKFC3jhhRdw+vRpNBoNtFottNttIR5XQz6fx8LCAg4dOiRkl3ZJWiDz+bwolNrKyN48qkm8J7TU6eh67qHul9JqF9UmhmZwHbNhHHo2GYBYSEYQBHAcRwgM7z+JtT6OjuXntfFYtMESuh+L6+W9DsMwdt20n/IDBx0uAkBIEUHVkNfF55x7wvPz+S0Wi0KuOJtvbm5OSCBVUj2/TSdM6rCQwWAgCZvaLkqlDdhRGNmraGBgYGBgcKfw5i9/FP/Td/4mPrwHQVZZK40PZXz8lQd+G3/nzKcxme79WCmDuxP7mqSxuNtrJW08HuPVV1/FwsKC9AFls1mxpul5ZwBihb7uf2KQBlU0kjQWoFpJYqHOYlarNCSiOhaeNj8OcuagZ22dowrBYde9Xg/tdhsbGxuo1+soFArI5XIoFouyj1QHmbJIorC9vY3V1VWxZ7Lo18Oi19fXRTFbWVnBW2+9hU6ng263K0EZ7wTXdVEulzE/Py+pkDo0gmoh0xv17Dru22zSJC2jhCZJWlHSippWo/iP90oHjHANPC7/6Qh+rl3fU16XTqgkSFh4HKpRfI/uT+P91b1kJHx8VqjC8jmdTaHk1/r8qVRKiNxsSqgOuWFojH6+K5UKFhcXZR16TIDuqdR/r7xvJLezASy6n4/7ZnrSDAwMDAzuBKxsFtt/9CP4p9/7L/FpZ7Bnxy0nHfyA4+MLtQt4qXkYwTBz7TcZ3PPY1yQN2LE76k/q9wLD4RBf/OIX8eEPfxi1Wk1IEPtpgF11QBfRtB4ywEPHvWtCp1UVHRGvE++o2miyplMl9QBmz/NE1eC8NBK5M2fOwPd9bG5u4sKFC3jppZcwGo3w0EMP4aGHHkImk0GhUECj0UAYhgiCAGfOnEG9XpdQDwA4e/Ystre3YzPNfN+H7/vIZDISo7+9vY319XU0Gg2Z/3U9WF5eRq1Ww9zcHAqFAoIgEDWIg6w1IdP/pVJJFYn3gioWiUYikYDnedILpi1/tEOSVPE6GdPPgBF+MMB0SR53dlg0sJM8SmWK5JwEiKSF95L2RR23T8sfe9FoQaRCp5MReb+73a5YKRlmQzI6Go0kYIUKbhAEsSRLPm/swwvDUIJbeL3T6VTGU9AO7Hke5ufnMT8/L5ZQjlngPvJ+cn/4bF2NPPKa+MGGtnTeDnuzgYGBgYHBu8HKZmE9cBxf/9l/jKS192pX0krgFw49hx8ffhfeHM7v+fEN7j7se5LWaDRg2zZKpRKazeaeHltbxlzXRbvdFnWBPT6Mj6f1kb1DVJgAiCqhY9J1MiBtYPyaKgsDHgDIUGba/YrFIo4ePSpqRjabheM4WFlZkb4l3/fx9a9/Hdvb29je3obv+5hOpygUCjLHbHl5GQBQqVQA7Ngfjx07BgD4T//pP+H06dNoNpu4ePEi1tbWRO1Ip9NoNpsy0DiKIrzxxhs3vdcLCwtYXFyUfqNisYhut4tUKoXFxUUUCgUp5qMoQqFQwObmJmzbjlkFtXLV6/VE+aO6UygUsL6+LvvLXidgh1Rx/plt2xJHD+wSCA6B5vBxKnskfbqXbTgcolqtSg+ejrPns9FoNKRnkYqkHg0A7ISYFIvF2H7xeLzXfIZKpZKonHpA+mQykeeJASckPCR/JLau64r10LbtWC8a93lW8aM9l9c3Go1QqVTQbDbl/TquH4AQWFpUOUON94v2RpJjPX7AwMDAwMDgvcTWT3wEz/6120PQDA4m9n010+/3USgUUCwW95ykETqGnKoJ1QgWryyyWXADu4OltfUN2LXxsVin0kGSB0DmnvF33W4XQRBgYWEBy8vLmJubw+HDh2Ox6ySqW1tbWFtbw9raGi5duoRGoyG9Wa7r4nu/93vx7d/+7VhcXIwRxVkb2fd93/chn8/jd37nd0RpIvnrdrtoNBq33A9oWRaOHj2KSqUiBBSAqHi5XE5IJfeM0ISG5CAIAiG6AMSiyr6+TqeDwWAgYRwkADpNk6opCQFTEgHEAi1IHDKZTMyKRzUzn89LSIhlWRLK4bpuLLSDJJPPSK/XQz6fl2eMM+JIYHQPI581DgYnYaMaq19LCyiJpO6F4wcPBN/PCH2tGHNMhLbt6lEDk8lEIvmr1aokXfL9vJdUC3k+kmz20GlizD3nPhoYGBgYGLynsPCeELS/vfTb+NX8I/jVCx+57ecyuLPY9yQtCALpq9orpFIpnDp1SpQWHSmuCYAGFTcW7exlm50/RbLEwlYXtyR/+hw6uIRjAarVKiqVigwqJglkz9nm5ibW1tawvr4ulsNarYalpSU89NBD+NCHPoTl5WWxR1I9mgXj79fX19Fut3Hx4kUEQYDBYIAwDGP2z5tFMpnE3NxcTBEjMbVtG47jyDo1gSV54vckmXrwMkM4SJSjKILv+2J5pD2VdkbeR+4738N9JlmY7SXj/dU9YVQb+TsqfLT/8Z7zmZkNzdDPAK2QOsSDSh/XwWPqDxN4LAagELrXUffz8Zmk+sY+Rb3v/BvQg7N5ft03R+WMhDCKIrTbbbnPvH59DK2G8jpoBSVBm7VEGhgYGBgY3G7U/9THkfkDm+/JuZZTHhbS7ffkXAZ3FvuepLVaLeRyOeTz+T07ZjqdxqOPPio2OPbmUFUAdu1vukhliMdsrDv/q3uBWGDPhkmQaGiFC4AoP8ViEYVCAfl8XmL+GfDR6/WwtraGzc1NsTemUiksLCzg+PHjeN/73ocnnnhCetlomWPfEQA5ThAEaLVaeO2116Sv7e23337HZMabAfur5ubmRFEieaAN0XVd6Y8igeNQZCpMVI60BW82BESnVObzeVGcdJCIVknZP6bnio3HY3nNLKj+6IHcDJvRvXOcnaYJGABZTy6XEwV1FnxeaLOkSqh/pwNVuC4+RyQ5OgFSv18TzV6vJyox7wuvj2RwlqhRWeZ+clYgn+lWqyX3SZNqfX2zVkbdc8jrMHZHAwMDA4P3Erk/uIGvPfLrd3oZBvsM16xmLMv6lwB+EMBmFEUPX/7ZXwfwfwOwdfllfyWKot++/Lv/B4CfAjAB8OeiKPrcbVj3deP8+fM4duwY5ubm9uyYiUQCy8vLSKfT6PV6EoLBQAUSKE3SOOvJdV0hPOz1AXZ7zfg+TdqYlEfFrtfroVqtxsIkJpMJyuWyzLRyXVcIX7fblb4zhn7QbvahD30IP/RDP4TDhw9L3xuwo0CyL0vHsLfbbTz11FP48pe/jF//9V/HhQsX3jWV8Vb2OJ/P47777kO1WpXrmk6nePvtt7G0tATP8yQhM5FIoNlsxsI2crkcNjc3ZdizVm5IJqhA6v4prTB1Oh0hZsCuauV5nqiG+h4SJJI6aASAWPl4jna7LYO3uc+8bwAktIR7Qsusnv/H/jSqbdoWy/eRJPIYJJ9U0Uha9Yw+rkMPWO/1ehJWQhVNh6GQxAVBgKWlpZg1UauPJItcTzablQ8WZnvitHrJ8Qk6NITgTDYTHGJgsD9xr9cjBgZ7hUlkYRpZSFh7E4hncHfiej5y/iyAfwTg38z8/P8TRdHP6x9YlvV+AD8K4AMADgP4Py3LejCKojtWNW1tbeEDH/gAqtXqnh2TBXav10Ov15N+JQYxAPH+Mh0UwkKSpI1FKlUgqgia+DAwgUoF53IBu/0+wE6xvrm5KWRQF73NZhNnzpzBuXPnAAAPPfQQvv/7vx9LS0vST6b3zPO8mMXx9ddfx5e+9CU89dRTeP755/HWW29JMX07UKvVcOjQIRw/fhydTgelUkkGGeteJR2VXygUEIZhrH8qn89LYc99plLZ7/cxHo/huq4QPY4bIGnp9XoxksZ7PxgMZM5dr9eD4zgSRJJIJNBqtWSOHnukOKNOjzLg0HPOMGO/FmfcbW9vi41Sx/fThsgo+0qlIvZbPmdadXUcB61WK5Z0ats2Wq2WWGJpGyRBpBrGNZNg6YAPholQvePeJhIJIaC2bV+hKvNctDB6nif9gHy29XOpZ94Bu9H77IubTCay31ez5hoYGOwLfBb3cD1iYLBX+JvP/yAe/F8DNP7enV6Jwe3ENUlaFEVftizr2HUe74cA/McoigYAzlmW9RaAjwF4+uaXeGtYWVmRInUvQbsWP91nsiCLXao+AK6Y1cW+HMayaxWCRaieQaUHKfN4mUxGVBAmRk4mE2xtbYkdjVH1q6urWFlZwdbWFqbTKR5//HE88sgjOHz48BUFLdMddRDIF77wBfzWb/0WvvWtb+Htt99Gq9US1eZ2YHFxEdVqFcViUWyODKro9/uSUEjyy33mPWAf13Q6hed5EqxCAkGS1O/3hShz/xmeQbKSy+UAIDY8meEcJB5M22S/liYTJNdMcux2u3LPSCh4r8fjsQSC8N6zHy2bzQqRHw6Hom5x3AGvn6pTNpuFZVlCYNhfxnWSOFIt47PItfHYJFz6A4QwDK/4QIF2U36tbYzsTeSaeA94DdxDHpPQyZl8DdeqPyCg+qk/GNFJmAYGBvsD93o9YmCwV/j0A6/hc3/8I6iKgGywH3ErzRt/1rKsHwfwHIC/GEVRE8ASgK+r16xc/tkVsCzrpwH89C2c/7rQbDbFLrhX4Kf4BItsFoYshK8WnHC1Acc8JlWxWYvXbBEKIBauwNdZloV+v4/t7W2JJu/3+9KHFkUR7r//fjz88MM4fvz4VfeERX+9Xsfm5ibOnz+PX/u1X8PXvvY1XLp06YZmm90sFhcX4XmeqDDdblcIAJUvWjEZ/qH3jYU8VTMAopTxNdxvklHuN0MxuN/aPqmHkusQDJ5L2+9oE9T9Uoym5z7z3pKAjEYjFItFISA8N8mXJvDALonRx+A1kLTr54ZqFLD7DOs+PT5T/F6Pg+CaAYhyTKJFIkjiRxsviSjVMfZwcg+4Bh5Lh5nMkitN4rR6y78hvTe8r+x/MzAw2PfYs3okh739QNfA4Hbg44W38NwHj9zpZRjcZtwsSfunAP5XANHl//49AD95IweIouiXAfwyAFjW7TPVsqdnNm3vVsFingWjJlk63U+TORaimhSwcNbF+2g0koJ+OBxKjL4mAiQLLJT7/b6Eh4zHY2xu7qQMDYdD1Ot1+L6PQ4cO4fu+7/tw3333wfO8q16XZVkIwxDf/OY38ZWvfAWf//zn8fWvf/2qr70dSCaTOHTokMwi8zwP9XpdCAUAeJ4XG2dAW50mG1Qf9cy68XgcU7o4Q07fLx3Yooc1ayXTtm20223pjdJzwYAdm2Amk0G/35eQGP1+PiuanMz2JPKf7pnj88V/+noACJHUvWokkRyMrnvr9HH4/qsF3+j4f5JlHp92Xyp+vE9UzbiPvu8jn8+LAglA+v70hxz675WKog7PoeLI9egPPmYTN/ciXdTAwOCux57WIwWrYj7ZMbircWbk4/zg1J1ehsF7gJsiaVEUbfBry7J+BcBvXv52FYCm9suXf3ZHoQvDvYKe2cQZToxiB3ZTEDOZjCg+vV5P7GE6nIHHIFng4GttHdPFPLCTWkkLJxUZztni62g5y+fzKJfL+NSnPoVTp05dNX2QmE6n+Pmf/3n81//6X/Hmm2+iXq/v6b69GzKZDE6cOCHjA2hnzOVyWFtbQ7FYRKVSiSUXUrGhbU/HsM9G0/N93FMOktZBG7ynJGjsG2O/FYdD8zyTyQRBEEiq5mAwiFn/2u02XNdFoVBAq9WKDasm0XccR45HMq57wTTBJGni6xkMQ4XRcRxMp1P0ej2EYSgkkgROB4fo9Ec+i5lMJhZyou2SPH86nYbv+7HesVarFVO0OMqA94RjDPSHGVoh1MqeTtKkHVSPE+D18dpI9DKZjHxgkcvl4Pu+CRExMNjnuNfqEQODW8Uf//N/EeOcBfzYe1efGdwZ3BRJsyzrUBRFa5e//YMAXr789X8G8O8ty/r72GnUfQDAs7e8yluE4zi3ZXYSC0kWvSw49bl0QMVkMhELn04r1JbGMAwlAIL/aC8DIP1N7IcDdnve2F9F5YWq3OHDh3Hs2DE89thj70jQzp07hxdffBFPPvkkfu3Xfg3NZvM9HQpcLBZRq9UwNzcXSyPkAHKqKuxF04mCAFAoFIQEMPBjVl0EdpMPU6mUEDDdK8hiX6cO6vRDkjfeO+vyaAWmQFqWJQQT2O2PY/8g+6VImnq9nqhHnMlGWyCvMZvNCrGneqRHIpDE8VnSdkHd16WTJsMwRKFQENWWShhJGPeKfYDaYqhVXU2e+Dvua6lUEnLb7XbR7/clzIbXQ4WPJJD3AYDsJ59D3n/eI/5tca38MCaRSAgpZp+igYHB/sS9Vo8Y3Pt4+2c/juGhETLraRz7f773LY7jn65ju+Wh8p6f2eC9xvVE8P8HAN8DYM6yrBUAfw3A91iW9WHs2AveBvCnACCKolcsy/pVAK8CGAP4M3dDktJeWx15zNlP6amEkAQAu4ETLDA5J01bvmaH9mpLGtUBAFLEarWIhf+79ckdOXIEKL8PvwAAoIdJREFUDz/8sKQLzuKFF17AN7/5TTz33HP40pe+hPX19fdUgXAcBwsLC7jvvvuEgJEY0C7nOA4KhQJKpRIsy4Jt27GeMP6XoRo6/VGrVOxB4zgDKkeahIRhKPeJIRkkBzpUQwdVzAaI+L4v6ipJiSYSet0kVzpFUhNw3/evIKXs9yIp1X12fN6pwDJsZPbvgM8kz8l1kjhRQSN0kqbuoyNh03PQtDUxinaGhDNlkx9QkOzp/eUzTsWZ+6V71XSPnLY6EiRxt+Pv3sDA4M5hP9QjBvcurGwWGz/1GKKTPhYKPTQ9B5t/+gks/ItvYvirC3h08KN44aP/8badfxCN8IF//+eQPuajUgxu23kM7h5cT7rjH7nKj//Fu7z+bwH4W7eyqHsBuqeIBIkBDlp5YWHJ12lr46yFi4W7jkPXYRdUS/RA5dlkPBbLXFs+n8fhw4dx5MiVDabT6RSXLl3CF77wBTzzzDN46aWX8Prrr79ne8ieq0qlgkOHDmFpaQn9fj8W+MDrpGWzXC7LEG5NnKiwUZkiIdV7zGPpwdOz5BbYHXnAe6j7tjQB15ZHEhsSCCYyUrkcDofIZrOxIdo8Hu/h7DgGqmKtVkvInt4XnbCor4X/SGRpm7VtO/Z6nSDKvdGjHfgM6d4xEloquiRpJFXakkmSBUBsiJy7R4VSEykdJKItjiRcui+P+071Ws9f00qfgYHB/oGpRwzuFJKFAibvP4b0D2yhcjlGoZzvYfoDIaYvvA+1X38V7e778LePn8JfmXtjz8/fmw7xwjCFk7/q442ftFF0Tc/1QcCtpDveM9DBEHuFWYLFIpHqlS5UdX8Uf09Fh/OsuMbZwcnpdBphGMZCI4DduHX25TCsggmIPA+HQc+C6sYv/uIv4gtf+ALOnTuH7e3tPd2jdwPtmadOnYLneTh8+DAqlQqazSbq9Try+bwMoC4UCrBtG6VSCY7jYDAYYDAYiJWRgRUkQYPBAMViEdvb22IL9H1f+rX4+sOHD4s1jySj0+lIIiH31PM86UPTVkIdSmLbNgaDgRAn13URBIHco36/j3K5LK8hGDrCc+o4/el0ikqlEiOLtCt2u91YPx5JIpVEfWz2yHHf+bzati1z5UiYmUrJ1yaTSbiuK7bHIAgkzIWEmgEdWlFkr146nYbjOJhMJhIeogNBSOq0VZFr5N8OiVev1xPip4k8yRsHvbPP7916Lw0MDAwMDK6JxM7/j4QffxDh/9jCbONMworQ+V8C5H7xQXj/6Vl87avH4X/jv8FL7F2iOAC8PLLwF1//DPA3QizAELSDggNB0nzf31OSRoLjOA5yuRwcx5G+Hhbk7B9ioa+HBHNYMnuXZuehsWimUkFkMhkAkEQ/DhDW4Qm6d6lSqeChhx5CrVaLrb/X6+HZZ5/Fv/t3/w5PP/00zpw5c1vnns2iWCyiXC5jfn4e999/P2q1GiaTiQxbXlxcRDabheM4MjAagChTnAFHS6Lnedjc3BQbHdUyDVr/SCgymQx8349ZFqfTnblq6+vr0gvGPjSGuTC10XVdIXLsJdRhH4zAJ1HiM0N1LJlMipUS2LG9UvnS56Uix77GXC4Hy9qZ+8f4e03KtN2y3W4DQEyNYt8Z16fnr5F0ktgxnZLWUyp9DAnRg7i5P/xHVY3HzOfz8DxPhnXbti0KH8/NQBf273EGnlaguVdaneY5tS3Y2B0NDAwMDG4Vpz/7Ibj5PhJWgKs3jOyg+ad8jD7ycRz5W0/hMx/8ffgDT72Bny5e2pM1/J3tB/BfVj+4J8cyuLdwIEjaXkdxa3Uim80Keej3+7FByABE0QJwRc8NlRkei79nwQtAVDem7BFadRuPx6Is6cI0mUyiUCjE5qGdOXMGX/va1/DUU0/hq1/9Ki5evPieBoTcf//9Yl3k0GrXdeH7vhDapaUlUYj6/T4OHz4sJGU0GolljgrTeDyG53mx0QTsSyJhJVnSQ52pTmn1kemYWrnSPW69Xk8IFACZjec4jjxnVLVIgBjGQSsfsBvDr5Ut9sjx9/p50YE0w+EQnufJ2kjwaEXUZJHXwPRPkkSdNsprY9qjtg1y3VS+MpkM2u02giCQfUskEgjDUHooSSC1JZNhHlRHgZ2/S37YEASBBL1QGda9mjwWAHneScp4bp1cSaXawMDAwMDgRjD89Eex+j0pRABKpQYyqWu3MtqZEXof6OHcz30cAPDLP/8+/PMfruPZR3/tltbyVzYewbdayxhPzf+fHUQciLvu+77Ecu8FptOpDIfW6Y5aEdP2RT2TioXm7Nwr3V9EMka7oy5QWdCTxOgCmvYvvlf3P/X7faysrOAb3/gGnnzySTzzzDNYWVlBEASx990OpNNp5PN5CQc5dOgQqtUq8vk8isWivC6TyaBQKMDzPLHe0S5HdYxWNpIjEh/uExUckg32MunXspjn66mizQ53BiD3eLZ3CthNiqTaqcM7dMCFJtc8lg4Z4X+v1nuo36fXBOwqrySWOpqfKpfuYZvtW9NhIjp0hOvVvXE62IPz0dj/RzLH66cVWA/bzmazcF1XCDYA2Xt+MDEYDK4YSK6DTXT/pU7yBCAkW/fyGSXNwMDAwOBGMPnkR7D28TQqj2yh+sjWdRE0olryUXlkC5VHtmBFQOebc/i9r/3gDa/hs515PPbNz+D/CDw8s30MjdAMWD+oOBAkrd1uIwxDVCp7E1g6mUxw5syZWO+OJhV6QLX+RH/2k30d9gHsFtKzs9PCMIwV+oPBQApYWiKpSFDZ0P1s3W4X9XodzzzzDD73uc/hG9/4Bs6dOwff9/dkP94JLLYLhYIkTC4tLWF+fl6GaTMG3rIsFAoFHD58WMgJyZ1OG7Quz30DIMoa946q1GAwELso9z8MQ1GKeI/Y58R7oBMFNXGiIkclifZB2gv1UOnZHjASENoi+Zzw+viM6GHXVDZ5Lq6FzwnXTnLKPaSSlU6n4XlejMzRSsi+NdpbeY20GAK7SYm8bj5X7D8bDAbyM1pB2Teokzb5HLuuK8SbRJH7pNMg9agK7jGPpQNYqGKSkNIWyfUwodKQNAMDAwODa8FKZ5Cs1ZCs1XDmD6dQ+ujmLR8z+Ye2YK9b6P/CYTQnvet+32vDHv7mMz+Ihb8E/NzpTyMYZm55LfcyRoUMEpdrvoOIA2F3vHjxIhYWFvDggw/i0qVb9wiPx2O8+uqr+NSnPoX5+Xm4rouFhQVRpbTqwaKY6g4/8WfxTwLG97Bo1eEJk8lEQi+0dY5FNIthphqyTyiKIrz11luwLAv1eh2//du/jd/93d9FvV5/TyyOjuPggQcewMLCAkqlEkqlkvR9sW+q0+kgl8uJgra0tISVlRUAEPWl2+0CQEyB0aERAMQqSAXJ931Rebi3nU4H+Xxe5qjV63W4rivkgeoPh2mT6AK7apgmv5qE8t5RHaJFVQ9SJzmkKpVKpSRJkuewbRvdbleIXL/fR6lUkiHcfA1JVTKZxGAwiA3l5hDodDoN13WRyWTQarXEigvsqnA6ZTKVSqHdbsuesueR5Knf78P3fbkfAOSDCpJAEk7Ob2N/XblcllAWkm+qiuPxGGEYot1uo1wux9IdObCd16+TTYHdWXK5XE6II3vpDAwMDAwMroXJEx9A6y/t/P/aAhp7dtzsD25iAOD3v/xj+PwH/wOcxLX/f+mn/tJfwOFxhMbPT6/52oOA3l9oo/75D+LQ33vqTi/ljuBAkLS1tTUUCgU8/PDDePLJJ/fsuCzIqcQwAY+kIJvNSgof+9RI2hhtTnWHKsx4PEa5XI7ZxrSSRLWEA5UBxJQ4qgxRFGF1dRWtVguDwQBra2t4/vnnsbm5GQsj2Wsw6TCXy+H48eNYXl5GPp9HJpMRxQOAFOAc3JzNZmHbtiQG0sIXBAHy+byQk1wuh8FggE6nI99TKWPSIsmG53kSnpFKpbC9vR2Lfyep0P1qJGNU49jrxXuqB5BTMaVyw/vM54C2SRKiWWsq0xxJtvlMaXus53lCuPkeEj/HcZBOp9HpdCTNUMf5U8VLJpNwHAe9Xu8KNZfJkOl0Wq5D2w51aAc/COAMOKY9skePxAmAhOqk02nkcjm4rhub79fv9yUxMwxD6XHjefX69D5qSyitm+zL417yPpmeNAMDAwODd8Lpf/RtQH6EZHqKKzOw9w5RZOEPnf4DSFgRXnvhKI7+zhhf+Ow/j71mEk3xI2d+L7Z/eEd1K9/G9RjcOzgQJI0qwsLCwp4el71uVF0AxKxpVDW0nY2KgCYrLG5J9nQhr8kZsDtQGNi1vVHNYK8TVSYqH41GAxcuXMDKyoq89nbAcRxRzCqVChYXF1EsFqUPiaSTxTQJkw67sG1biKkeTs0hyFQYXdeVIp0/Y1AHCQ/3LpPJwHXdK8iy7gUEIEonCYu+R3p4uO6Z0rPESCQ0oeAzQCueJlLD4RClUkmuj+vhuWgZ1BbM2Wvlcfh88NnjhwMkn3w2dc+dDlHh88dr18T/ajPitLLHgdv8noScZIk2SO4LP0wg8db7yvNo1Zl7q/dudqaansk2e18NDAwMDAyIZLWC1T/2PnjLTTiZ2/ehtQb7yqb5MTYey+Bv10/Ffj+FhY1eHuX89VsjDfY/DgRJ63a76Pf7mJubE3VgLxAEgfQ9kSCxmNYzzLRSNhgMhHjoUARCDzrWfVg6NOFqvW2a7AGQ4rzRaGBlZQXnzp1Ds9m8LYUriVelUsHc3Bzm5uawsLAAz/NisfiDwUCUFapgVBN5HJJZHaih+5a4Z4VCAZ1OR5RDABIrr4NVSJZzuRw8zxPLIYM59H7r/i/aEvXxdD8Ve930bDCqbLSfzqZx8nsqSlSfmLpItY3Xz/fwdZokzYaA6JCRVCqFXq8nCuHViDltkiRTvA79gQGw25+m7wkAsfCSIPm+D9/3xea4tLQkoSYk3pxjR3sjsEPKqca9kyqpw1hmn1+tGnLde/k3bmBgYGCwf5BaXEDvw/fB/r233nd2M1g43AIOA/95xUTqG1wbB4KkBUGA8Xi8Z8EhRBiGUkTSqqWLd1qwaB9jQT4YDGJ2LxaaJAmO40hAQhTtDkam2pJOp9HtdmNkjeoHLZBUbra3t3Hx4kVcvHgxVnDvFUislpeXsbCwgFqthlKpBNu2Re0CINfheV5MdXIcB0EQyHVxf8IwlALf931RXGgPzeVyaLfbsVh5nZ7JfaAlcTAY4MiRI2Id5F67riuWOVowSbiAncj9WeUqCAKUSqWYDY/9YSRFmihG0c7ML/bHcT94j2mNpXWRRIlqEwmXnoPGIBIAQnxJDhl3r4dps1+O183nkQPQ+V4+O9rKq+fD6TlrJGedTic2Z6/X62E8HsN1XRQKBekDLJVK0s/HHrpKpSLW0tFohFarhel0Ctu25Xp1cAn3iwSWc+U0OeMHJwYGBgYGBgBgXf7/1fr3nYD1R7bu8GoMDK4PB4KkATu9Unttd9SkhwUtC3Um6LGw5OBgqghaPdHv1cUwbXVUK7TljsW+7vXq9/vSYwUAKysruHTpEtbX19FsNvf02gHIsOlSqYQTJ05gYWFB1DEWzAyPYFKh4zhCYDjcmH1fnU5HhlMTDA7hXpPMMaSF4SIcsUCy1ev1ZD85xJkjGFjk6zlrOhKfipa2K+r0xGw2C2C3f6xUKmE4HMo5bdvG9va2zC+j3TWZTCIIAvmaayGJJzlkMAiDYDzPkxl4JEO+78uMvkwmI71qHNjNc2jSpodb83pJ9klAs9msEGmmOHKGW6/Xw2g0EhWTZHp2EPp0OsWrr76KkydP4tSpU3jiiSfw0ksvyd+Bju3v9XpCvD3PQ6FQEKU5l8shn8+LMkiyphVjBoWQlPFvSg8KNzAwMDA4uLCyWVz89yeRSk6RSm7DfIRncK/gwJA0Kgh7Cdd15VP8brd7hWqhe3loddSz0Wiro5LA/iLaAgFITxJ7t6issDeJ56BiQ0IXhiHCMMTGxsZtidqv1WooFosoFAooFouo1WqoVCpinxsOhyiXy9JfNxqNZE7WrKWPag6Lfib1kSCQFBHapkelh6RY9wByj2h35NBjrimXy8n3ACQZk2B8fKPRkHtFdZRKE/umqMjpqHzed93rxf4s9p6RrFNN4wDv2V5FkhoSTQ7VJnGkykqbYKvVQrVajX0owCHRXJseAk3w3vBZSqfTMTsq+93a7TZ6vV7MbqrR7/fheR7y+Txs2wawo2jncrnYhxnb29toNptyLBJPPuvstxsOhwiCQJRNXivVNpJu9h7OzrYzMDAwMDh4SDz8Prz5kyVUc9tIWMYGb3Bv4UDFn+11nwqLfBKp2Rh2FvbAbu+MTrjj19o2R5JFdW12KLKOb6cFTCfgkeixoG21WtL/c6ugtbFQKKBcLmNpaUkGU1MVy2QyMmzadV04jiOBHyzCaQklcZ0N4tB2SB0owd67WZCg6pj82RAO/Y/qJueUAbsz7HjPdP/fbE8aj00iwXvCDwH0UHNgl/CQVFFl1PH8fB2THvn6WTVVx/nroBgSOa6Dz4XeLyp/HORNskrwmkhyuVckpySJJKf6g4hZ6OAOPjc6MITnYEIkSScAIWok0PpYXLceXk3w3rE30MDAwMDg4ML66Adx6VMV1N5XNwTtHkawNMXgBz56p5dxR3BgSJou4vcKmoCRKNEixt/Tekg1RRfmLN5ZOGulj8U/AFGWdHqeJofsL2IxzwHW3W73inCNW0Emk0E+nxfVbHl5GYcPH0a5XEYul5M5WFS+SNioOIZhKOukzdD3fbH5MUZeq2EkRnq/Zi2fVB71OAQGc1BxYZ+TJmlcA1U3Wul0KiL/kVxdLSaevXIkTZxBp4m37pPSg7EZLMPX0J7JawV2FD4dEkIVkMSHRInXpVUoAGKBpdLIZ0c/uyRofE5Jknq9ntwLWlF10uK7/U31+31R8UjOSfBSqRQcxxGizpRU/j3wb4X7pUNd2IvGnr5ZwsYZdYaoGRgYGBxMJBfmceHT+TsWEGKwd6i9r46LP3r7ksnvZhwYu+NoNIoN4d0LkGywwB2PxzGVZTweS/gFi+FqtSrFaBiGEnxAsM+KJIR2SABSwFuWBd/3Y+EIJG4c4gwAW1tbe0pMa7Ua5ubmEEURlpaWJIil3++jUqmg3+8LIaNNjUOfWeC3Wi1JfdzY2LjCqseerHa7LXvBfSaBYngFiQODRBjkwmANbavkOmktZZ/TcDgUssDCvtfryf73+330ej3UajVkMhmxo5K4kcDoOWEkR7Tt8b5qtZA2R903xnWS0M6qjFTMbNuW95PAsL8L2CH13HOS0sFggHw+LySHJDGVSkn/VqFQkGAVPpu5XE4U0Fwuh5WVFfn6WuEcw+EQm5ubOHfunNwb/r1o8qevm+vhtfL8nD3HNWlbJkkblb3xeCz7Y2BgYGBw8PDm31/CXNkQNIN7GwdGSev3+9ja2sLRo0f3LPlN29qYTEgL2awVLp1Oo1arifrBn1MVYiFNckO1IwgCUYq0Eqej2HWwBcMltra2sLq6umeDq7PZLEqlEg4dOoSjR4/i0KFDcBwnljq4uLgIYIcQHzt2TEI0eC3sR+p2uwjDEAsLC2Lt5H4wkl0X5loVY4CGVh5JerjvvAckgyRCtm1LTxMDTQjuI+eL2baNbDaLXq8niY/a0qjJJxU5ErVCoRCzruoh2cCuSko1Titu2s6qA0RoN2RPIq+R/XZUIamuFovF2LGoFPLZooLFY5P46L8Nrfgy3MRxHCFB15q55zgOBoMBNjY2ROkqlUqo1WoyxFwTaT77nKE3a98NgkD2iOeeHTtBMny18RYGBgYGBvsbqcUFvPW/PYpyMbjTSzHYQ1TKAd763x5Fsla700t5T3FgSFoQBFhZWcH9998vFsRbhS6+tVVRJweycHddV2aC6bhw3/el4Kali2SA6o4mf8BuMap7c1hwkxSw12ev1ARaHD3Pg+u6sCwLYRhiMpmIesbgi9FoJEX/7P7onizaNNl/RRVFD7smieU904qV7gVjrxgJiZ6bxXVwBhf7ubh3HHbNVMPZ4l4TcK6JBEaTNxIckiEqanq2GPu6qMrxfpPIk2xSrZrtJ7QsKzaygPZRnf5J8k9FUdtBtV2TaZhcIwNYgHgC5mxfJa/tWh8A6N41qn18LkkwtW2Se8q/A/Zd6hmBWqXmGAKGueh7Y9IdDQwMDA4YPvZBnPuT96NW7SKZMP/7v5+QTExRq3ZhpfY2APBux4Ehad1uF2+99RaOHz++ZySt2+2KJY1WLWBXlWHxyR4fxoXrvikm+QF41zQ62vf4Wh2WQSJEksbz7KWKNjc3h0KhANu2USgUhEACOymXVNR0X9PV1Awd6sHEQd2DxNleLOCpAJEA0XpHRYX7TZKmAz44S4uvI8EjgaZ6qdUbkkadlKnDOfT91WSTpIYEgqoWSQNJoZ47pgMxtBqrSdZsKIgORCHS6bRYbfl7ff383exxZpVEvp7H1/sBIBZ4o+2k7wQ98Jr3hzPSdJCOvg79vHDNvIarKXfaRqpTHbvd7jWVPgMDAwOD/YHkB05h/TvyKDxhLI77Gb0PHUFyYf5OL+M9w4EhafV6Hc8++yweeOCBPSNp58+fR6vVQiqVQqlUQjabjYVTsEeICgFDDljsAogVwNPpVOLOGWzhOI7Y7Vj0zoZpRNHOoGgW4Ixx3wskk0nMz8+jWq2KBbBSqcT6xzzPg23bohClUink83mxL14tRXE8HqPX68G2bUmApErUbrdjig4VJxJb27ZlH6lS6SAMrpuKGu2N/B3JAWd/cQ9p6dMWVN/3ZZ2pVEpGLZDsJZNJ6QHs9/uybj14Wg/appWTwRnsv2J/Hd9PBYr7yH66VCqFYrEoJJeqJa+L18ZB6Jq46WvXSYhU88IwlJlrlmXB8zzkcjkEwY5thB9EaPL5buAMNKp5vV5P9linZAIQy+Ussea4g16vF4vf556STDNUh+R0e3t7zz6kMDAwMDC4e2Fls3jjT5Zhf58haPsd4Z9rovOdx+/0Mt4zHBiS1uv1sLq6ilOnTl0xd+tm8corr6DRaFwRgMCiUYdMWJaFQqEgKheL1bm5ObGP9Xo9hGEYK4CpJNACSIWGSgeTCalQMOii0+nsyTUmEgnUajUsLS3JkGna9vL5vISEkIDYto1SqSQkkQU67WdaAWN/Ga2I0+kUc3NzUogDO/eN5JcEeGtrKzZPq9PpIJ/PCznp9/vSi8b/UqEqFAoyS2s6naJarQJAjPA0Gg0AkBlfVLn4j/eFFlMOmw7DUAgY7w/3QIdbsPeQgSNUuJrNpvSzMRkxm83KPmui2ul0JESk1WrJ7DySKD5rAGTdtFLyGdSDvIFdKyr72/h6fqhBSybJ6bXmkLmui3w+D8dxAOwEmpDIJxIJIW2j0QhRFMnr+Mzzb4TPT6/XE4JIAsw18Fo5jLvdbhslzcDAwGCfw0pncPHfn0T1we07vRQDgz3HgUl3BHbUqmq1GgsbuBXoWVOZTAZzc3Not9uijuheM2BH7SHRYU8Q+6NIUqj+kPTNqg4s9lkks1Blr1EQBNje3sb6+votX18mk5GgB6oXVDBIkrTqpBP/qGxQ+aOiRvsn7X7tdhvFYlGCNyqVCnzflz3jNZLU2LZ9xaw5KpckRlSZSIC0/Q+A/C6Xy0mfIK2LVNZm0w2DIJBh3J1OR+4NSQ2DUXjubDaLdrstM+OGw6GsnT1UYRjKfdbWTm2dJJnUgTPD4RCu68p1k+DoPkUAVySNkvzpAdiM95+1dHJfe73eFWmZvI/X6vki2eYQ7X6/L3ume+lor6SFVpPLMAyxvb0tfXjcV66R+0O7Kkc5GBgYGBgcDKSSUzMHzWBf4sAoacBOUcfQi70CyQXVHioNLLJZ7EZRJP1o2vrHf3wfwzf07Ccdv06iRpKm+6aAHcud7/tiUbtZJJNJOI6DcrksKg4tadrWqJMDrzY3DrhymLHuyePXtN5xDhywm4I42/+lZ4lxH7k/3G8dyMF9y2Qy8jVfR5slXzsYDIRsUf2j8gMgZiHk9yRBs9f0TsRHz4mjEsrjkBTxNXwW9LWxl4+v4T7r3i99P/RsNU32OZuMa9Y9YHrtWrGaVd3eDRyEXSwWReHT/0gqAcTSHHlPqZJ2u93YzDdaXLlWHRqje+gMDAwMDPYvUosLaP3hjyCV3NsZuAZ3N7YfTmL0ex6708t4T3CgSBpwfcXljYA2LB1+QSsWgFix3e/3Y0EhwK4aQJWIRCUMQwn/0EWpLsQBxMJIqJQEQXDL89E4uLpcLgtBo3rmuq4UyyRXLKJ1mAltpZqoamsjr5sWNtu20e/3Ydu2XJvu1+IeamWSr9H/uO+z3zNxkqSQCqi+R+yj09ZJ3huSUb5fK15UDHXgB3/H6yPh5Nr12ARaZGl35ev0QG2GZFBJ1emOeqRDOp2OqbF8ZjgyAdi1d+pETE2EdeKmHrau1bRrKdLscavVajIrkOA1kaSRCJM487lotVryd8BxAlRA9Xp5j3mPDAwMDAz2NyZH5oE/Wkc6aT6YO0goftsmLvzevcmWuNtxoEhaFEVoNBp7WsT1ej20Wi0AECsji3LOiWLPlg6noFLGZEQW5SQPLFhZgHueJ+SP87sAxKyDVGIYZHEryOVyQtJKpZKQNYZwaIKWy+XEoshifzgcCrFicV0ul2V/2E8F7BJdWg2pMjH4gcObSSzY+0WlcTwex3rweAySRw4MZ6w91RY9tBqAWCBJpqiyNZtNAJDrp02QvWgM29DDtUmeSBR18qCeqRYEQSyZsVqtyjm4RzpMI5PJyLo06eUxqfxxn/ncjUYjdDodubZkMolisSh9b4VCQSyzjuPIWAVN0rTaeT3BIVtbW0gmkzhx4gSWl5flvpBgUrnUaZJUbvUHKezv1JbL2ZROne6pUzMNDAwMDAwMDO5FHKietOl0ikuXLklxTaJzK6jX61hfX8fDDz8MAELO0uk0ut2uzEij3U/3S43HY5k1xqKUg4pLpZIoF8lkEpubm1KYk/TwGFRWGo0Gms0mfN+/5evigGQm/LXbbQndCIIAtVpNgiVIPAFI8U11T8fkMzSDaYT9fl/uA4dVl0olmQ3GwAvaSWntY18byUYmk4nFulNNInHTSlU6nZbzFQoF6dHSkfMk13qGWrfbFatno9GIxeAPBgMhnCQ1vK/sOeR5ptOp2Ed7vZ6Mb9BDsXlcXgvnyUVRhE6nI3umAzTy+XwsEZNJlFRE2W/H/jYSfSZqAhDyxGeKz+VsQqe+znfDiy++iAsXLghJ5CBsXtdgMIiRPX64oANm+Kxwf9gfqC2zHG/BGXR78XdtYGBgYHD3YutnPo7e9/oo3emFGNwR5B9s4vS//gge/KkXEe1j98yBUtKm0ylWVlbwwAMPYHFxcU+OWa/XcfHiRfi+H+vJmj0v1TMWwbNK0NXWSoKje3hY0JIEAZBer16vh3a7fcv9aIzWp+oE7JI2Fs1cO6+VihIJBtUlqjokjjoJk/tByyEJCUlELpeTuH3d10X1hCRFB6loqyUJAQmFjpfnehl0onuwqF7qMBTeJxKmXC4ntjxCExtaPblPem4c10iiSZVQp0Fyb3lNJHJ6L2i3pCVUk2ISLpJojjHQ4SK0mJIA8nq1ssv9v1rv2LV6v7Sdkfum1TcSd62O8WttI6Vyx/vADyp4nVSXqZDuxYcUBgYGBgZ3L8aOhZIX3ullGNwhZNNjeMX9f/8PHEk7e/YsTpw4gfn5vRmG1+v10Gw2EQRBrL9plkBRJWCxTgvXLGjTI9nRvVZaOdGWLhbY7KG62nFvBCRHnF/GEBGdRKitdCySdU8WgJiyRdI5O99Kz73SyZbaTqlDVGgdJBnisbXKpu1vOtwknU4LudXXMrvX+l7oJENaMPUAbq6V16ftnVppmu0rJEkDdtMmdZon90bvse5DZL8YSb8+t1Y1+fpZG+RslD2fIe4j90QTVB3Bfz0kbRaz0fmzc9J02IsOwtFplXovuE4+J7xXa2trxu5oYGBgsE9hffSDGFTN/8Yb7H8cKJI2Ho/xwgsv4L777tszJQ2AWBhnBzMzqZEhIPy0n8SNs8RmC9d8Pi9JfLqg1srPbHACi/O9KE49z4PrujLniioa1S8SFIKFMov8RCIB3/el6E6lUvA8L2b5Y9HPIpxKCQc28zV8vZ6nNdu3RwKkY9y73a7cCxIAkkDeDyo5OhWQxIFR7ul0Gr7vy/5yfZp0k8gAiN0PKm0kjyRJujdLWwtJlPTwa622klwyfIaK43g8RrvdljXwQ4HZABvd68W9Yn8fsPP3EQSB2EFnUzFpP+Rw9esNpyHZ5QcQOiREB6BwbfygIZFIoNvtSuAOLajcV31tOhn1lVdeMQmPBgYGBvsUb/4PaZQf3brTyzC4w7AAJBwH2MMwwLsNB4qkTSYTfOMb30CpVEKxWNyz447HY7RaLUwmE1QqFVQqFbHrAbtDfBk4oe1lAKS4ZLGsY/jZk8ZZYZrYaQJAu+OtpjoCQKlUguu6kiSowx6SyWRsZhVteCRDHKQ9a79jP1GlUpGifH5+Ht1uV6xttEcyFTCdTqPdbiOfzwuRy2azaLVaseRIkh+d+sh+NyqWtBVqW+SsDZGEh6SCwSMM0eA1MGAEgAxaJmHQ6YqafCWTSemFZI8aiT3DQXzfR6FQkDAR/fxwf6nAaSshnw8quSRztL26rovxeCzhLiRK2qpaLBZRKpUkdZHrpd1VB6Fwrh2f32shnU7jxIkTWFhYgG3bMnBaD/LWyjL3utPpyEw6krgwDGUOIbBDSFutFlKplHygwKAXAwMDAwMDg/0JOzPCxr9bhPXh99/ppdw2HCiSBkDmRe3VQGsACIIA3/zmNyXggnZFJh/qqHRgt6jWPUp8PQfx9no9iZtnoASLahIMrULNzku7FdCWSBIG7Foqx+OxBGtwbYzTp+KjBzJTNWG0favVklANJkXqOVx66DNJKQkL9yyfzwOAqGIkSSRX2WxWbIJUXUajkZA7kh3OCXNdF7lcTt4fBIGcdzweI5/Pi6LJwBaqWeyPK5fLsbluJGYkMwzy0CSax+PXmiCSlCSTSQyHQ7GzDgYDeJ4XC6BxHOcKOyXtp/yZti9qpZKBKkybpOLI+80xBXpNeubZtUDrou7f0+mm0+kUjuOgVCrJa3U/Hee9RVEke06lUVtpdfKjgYGBgYGBwf7Hfh+/cOBIGoCYDWwvMBgMcOnSJQCQhDkORNZhDSRQs6EILIb5Oz3LS1v7SNaoCOleJJKBWyVp+rz6n/5dNpuVNdHux2vR10jCwq/1zDJgJ0ykWCzG9kfPMNMWQq0akczwNTp4hb1sLOhne7lmo9p1fx/vlZ5lx/XrZ4bXw69nhz3roBe+hmRr9jppESXx4D3U/XE6cZH7zSh8Jmxq4qI/gNCWRdpVaRukisjgEu6FDiQhwZxV666XFDGN0bIsIcPc+0KhEHu2qBZTWdPJp3zGZ4kZ1VCqjter7hkYGBgYGBgY3M04kCRNx3fvBfTQXUbwF4vFWIHJ4pOFN4tlraCQJNDSxoKYisvs4GRglxQMh8M96UlLJBKxhD8W5sBuX5MmJVQ9OKML2E1gZC+djsBnyiL71gqFghBVALEwFK6HvWQA5Fw8j/6vDiGZtQPqgdW8Fh2YwbVqhUfP89IBIToQZJZsMU2R6yH54PearOveNu41ybZOeGS8P1Ut2j4ty4qFjnC/aGXU/XF6nAF7vvg6Whx5DqppAEQx5bEZJHO9wSFBEKBer2MymaBYLMLzPBnaXqvVRNXj+TmcnbZZTcC1IqhVW870m06nt5xsamBgYGBwlyKRROr4USSSJjTE4GDgQM1JI5rNpgyI7na7e3LMKIrw4osv4iMf+Qg8z8Pc3Bw2Njbkd3oWFAt62tfm5ubEbkeLo445p21Pqzvs2+Ixfd+PhSncDFiEMwCF56EKwuK83++jUqnIIGIqPVw3j6MDMtjLplUZz/MQhqHMTnNdF6urqxIYQvJTqVREZfF9X1QTkj7ODev3+zLTq1Qqwfd9IQS2bccSMKfTKfL5vFgeaVsEIAX/eDwWayEthySgjuMIASKJpBLW7/fhuq4MiuZMNBJfroOkl6QznU6jUChIz5UmyWEYXkGkSMJI7LjXWnnSqiV70ahUMZFUDyDXJI8qnSb/uhfyenDp0iV861vfQrFYRKVSEQUtkUjIPDM+591uV/at0+lga2tLgkJIyPSHA3wvB3O3Wi1cuHDhpp9/AwMDA4O7F6n5OWz/4zTmsDd1m4HB3Y4DqaS98sor8DwPTzzxxJ4dM4oifPnLX0aj0RACwmHJtJexCOaMrvF4jEajIaoaCQsQVwq06gPsqDe0jfHc2op3K5i1C7quC8/zYkmTXDuLdR0AwSJ+MpnIMGmqS+xvoyrl+z4Gg4EEW9Tr9ZhqxD3TtjqttFEd6/V60ifHXjKujQTW930hoCRlmgSzv822bXieJ9fYarVE8eR8tFQqhTAMRSmjRVKrfZZloVAoSFAMe+X0KAPdAwbsphryfUxf5L5ymDXDY3hMjiRgHxf3Vc8l089FMpmMhZGQWOo95wcDqVRK9oyqF4mtTvh8J2xubuKll17CW2+9hY2NDVn/aDTC5uamPOd8/tfX17G1tSX9e1QMdQInX09rqOd5KJfLqNfreOONN27p+TcwMDAwMDAwuBtwIEnaN77xDQyHQzz00EN7etzRaITt7W34vi+2QRbSeraUnglFAqeTHKlI0fKoBzazICeZASDqlbbS3Qx0fxXJCAtlrRrl83lRg9hPVy6XY2MEaJnUFj+qOrxWkgqmQ2oVUK+FypEmVlrd0TY8PeeL5Iqv1yqQTpMkCdSJjUw4dF031iPIXkPa8NgnOKsuacKo7a562DSwG+TBe0vFjUQykUiIYknCxGPo+65B26cmtwzv0PZaKrHcHx1Sw/VTEeX6uP+61+/dQJJ94sQJFItFSW18++23cebMGbRaLQRBgE6ng3q9Lh9UcL9o0eTfCQCZi8ZnQ/89mfAQAwMDAwMDg/2AA0nSTp8+jel0imPHju35sS9duoTNzU1YliUKBIBYZLkOl2ABTtKhLWu6b46qDYvlWVJDpW0vetIAxKx5AGIqHlUm/TOqQ1SmSGh43To0hOtk4IhWfHhsADFSqodXszjnuUlUdAIkSRX3UPeK6aHKmtySRFKh4XVpQgYg1mvHvdGKFYmOHjBNkq2vTZNMHo8WQ93LpufDZbNZCY7RYxw0Qee+8ZnQRFnv22yPnO5r06EhJGW6J0/fs3cDz2HbNgaDAbrdLra3t7G6uorz589LzH673RabJ4kk91aHrei94n+5J7Nk1cDAwMDAwMDgXsWBJGlbW1tIp9NYWlra82O//vrrePvtt8XymMlkMBgMZPYVAEnT42yz2cANEjsGalAdoKJBy5yOt6cd7FZ60kgCNZmkFU4Xx/w5SQOti8Bu8APj9am28Xick0ULIPeF4RWO44idj9ekBy7zfCQNJLSdTgf5fF766BKJBIrFothCLcvC9vZ2LNylUCjEyBTJ1tXSNLl+WvV07P1s2iHJDG2fXDtVMB5fk3StEnLenQ4yAXbJJdW20WgkvYh6HAMtllTBqE5xJh3JaBAEQgq5RwDEyqgTNjWptW1bVOJrgfPZLly4gNdeew2nT5/GhQsXsL29jbW1NXS7XbRaLWxtbcn91fH6DALRIyF4T/iMeZ6HRqMhPW4GBgYGBgYGBvc6DiRJA3aj8vcaVAV6vZ4Uv8CuBY2BH/l8Hvl8HolEAu12Wwrx0WgkJE7bH0liLMtCqVQSK2G/30cYhnAcR5SjmwXVEx1SQisiz02CxaKdxLLX60mCINUXFtjc63K5HOsHY8IgyQjJGBMVqdoEQSDqG3vHOp2OKIrj8RiFQkH6ndgnx72mpTGXy4lKMxgMZM0kZlQHGQziuq6QAdd1Zbj1bJgIh3iHYQjf9xGGYSw2f/YfSdlkMoHjOHKtpVIJURQJcZ2Nu2e/WLlcjg2qJnkjyU6n07GAGu5BKpWS66C6VS6X5f76vi9DphlswnvQ7/fR7/fFvjs3Nyfz6t4Nw+EQzWYTb7zxBl577TU0Gg1ZW7FYlD1Mp9NwXVcCWTTp7PV6EgzDvwvaZCuVCo4dO4bTp08L0TMwMDAwMDAwuNdxINMdAUjBe7uO3ev1ZF4akxFJHvQcK51qxzAHbT8jEWKhzSCPfD4vAR+tVkv6x27V7gjsEMVut4tUKgXP8yR1UCsYDPugDY+Du6kQep4nipJW30g2+HMOmLZtO9Z3RhUqDEPk83mZ48V5YwwhIYnQg6RpBSSpI3HlIGqqMlReOBZgPB4jCAIhYTyGttmR2OtQF9d10el0RMlMJpMolUpYX18XVWswGKDdbsO27RgBpfpGUus4TkzZiqIo1oeWy+XQ6/ViKmK/35cAEGBHjcrlcmKTZLCJbdsSTkK1ksor18J18X7NWgxJ4HTgybvBcRxUKhUhnwxC4bPENEnXdTE/Pw9gN2qfzwD3J5FIwHVd+QCE3xcKBWxsbOxZUquBgYGBgYGBwZ3GgVXSGo0G1tfXsbi4uGfz0ghay6j0aIWLqhiLTBbnVA5o/WPRTHWJ1rNEIoFCoYC5uTmxU7LQZk/WzYLnYBHOFEZ+DeySW23ZI0nQxbwmcFSNeD268Navpao1m0RIYsXXkshxP0moqOQxCl/3NrEfTM/IY5gGlUGqfADkZ7wXfI8m2yRbnCXmOI6QslkLJdVBbV0EELOy8jq51yRd7OPj8fg9r0sTcyqCOlCEpJTfp9PpGFnkemZVX/YU8nvuyWQyEavktUDljT1nVIn5vJPk5vN5UQX19dCeyeecKZs6kIcK6q2E5hgYGBgYGBgY3E04sCRtfX0dFy5cwEMPPXRLxOZq6Pf72N7eRjKZlLldAIRM6AKVPUO0cjHVjgSu3+/D932xvLHXqlKpSF8Qgxxc142pdDcD9sFRBSSpYeHMkBOtkNFiqEMlrkZGGN6ho+szmUwsQINf6/4uKkckMgSPTXKpSRrXyD1nSqJ+PaPoCR3IQnLKwA+unzZPphZOJhP0ej1JmCRxoFLFNdAGquP89TOhQ1/0gGcSDx2oQcKq7aaaeHKPuAfsweNzxT6u2XtJos3gFZ3qSYJK8tftdq+rB4ykt9lsyqw3/uOHAZ7noVKpyHXzOvkM0KpJtZDXQEJ88eJFExpiYGBgYGBgsK9wYEnaysoKVlZW8MM//MPXFYBwI+h0Orh48aIUw7QysoDWs76oDvm+L7OhGApBErG5uSm9YoVCAaVSSchfLpdDPp8Xq+StEM7pdIowDGOqEgmkHrgcRZEMkGaMfb/fF7sZ+7my2Sx830e/349Fzmu1KpPJiH3Ttm0p/lmoM/GPpIgq1vz8vKh4tMVp0pLNZoXAAjuqFEkKLXulUkl6rXid7BEjWSaR7nQ6MnSbBGN2yHO/35fRAuyT04pWpVLBaDQS2x8AeQ2DZrgW7hWVR5JG9iySxLMnjuekSsvz02bIPSLhIqHj8UnCtILIYeBUIKfTKXzfRxRFqNfr12UvZJBMr9eD7/uyb+xVi6JIlGHOeGP/mVY+tUI4HA6Ry+VQqVRQrVbx5S9/WUixgYGBgYGBgcF+wIHtSTt9+jQsy8LP/MzPXFdvzY2g1WrhzJkzMudJExNdiLN4Zy9Xv9+H53kIw1AUGdr7aDvLZrOo1Wo4cuQINjc3kc/nkc1msb29vWd2r0QigWazCQDSJ8TCOJlMYjAYCFmjEpJIJNDtdlEoFJBMJtFsNpFKpVAul0VBY58eyQ37q0h6OMg5nU6j2+0KMc3n85JimMlkEIahqFfcG2BXreM+6Z40x3FQKpXgOI4QQV4HFRwgHvAynU4laZMkodlsSg8dlblKpRJT2tLpNPL5vJAgHns4HMJxHLlebUdkb5keGk4MBgN4ngcAct3amkjlTpMsrRoSJFxM4mRICPeMKq/uSdM2WypYN9L7SAWM90croZ1OB8ViUT7E4H7ogBVev04e5YcV5XIZ+XweW1tbe9KLaWBgYGBgYGBwt+DAKmksPJeWlva8J43kYH19HQCQz+cluY+qAL9mr1C73cbm5iYajUZsZtdgMIgFXnieh7m5OSSTSczPz0vKHodP79W16MHPVJR4fBIuhpnQ0mfbtlgW+Xod404CQmugbduijHGuHHu8SMj6/b7sFa2Y7O3itZKIafKjQzB0kAhJAm2U/DlVO85bo20SQCxunkqn67pit2MiJbA7RoHR9nweGLfP13A/uLd69hzJlg5koYJGmyNJnN5nbRPkveKx+EzqGWc8Bo+pPxRgzxf3XX8AoMdGXM9zxPTLWbtoLpeTUQZUOrVNlqEi/JvRaaee58GyLKytrRmCZmBgYGBgYLDvcGBJ2u3GaDTC6dOnMZlM4LouisWipPSx6GXYRb/fRxAECIIAvu9LIcv+HFoHSSjYR1WtVrG4uIi5uTkUCoUryMXNYnZ48+zxOMtLkzSSSQAxWx3niukZa7qHigOttQ2UfWBUuPRgZxJAqj2z4SN6jSQFfC1tmwzyoFI028ulxwvoAc8koP1+X66dP9NEkn13tOdxX/T8Ne6t7pubVXSpmDGIRJN33bul7YuaPPH8tEJqEk+LI8Fr1b1xOixl9pzXQ4xoOQUgRJzEfzAYCFHn9TF1UoeVALtKo+7v4zy9N99885rrMDAwMDAwMDC413CgSdpwOMTq6mpskO9eHvvZZ5+V6PNyuYxSqSREgL06iURCCtPJZALf9yVMg4VquVxGuVzGaDRCu92W1L9qtYojR47gxIkTmJ+fR6FQiAWV3Az08GMW47RjkhBwXhitnAzP0IEgVMDYh6QDQnge9ngNh0N0u120220hSACkqCfxJGmxbRuu60oARS6XQxiG0puliRDvLW1+VJmYhMk9JsGijRPYJUl6uHYikUC9Xr8ihZJqHqP2s9ms9I5xXIC2IwK76Zbcb76ORJ6qI/dCp2zyGWMf3Ow1067I9/T7fRQKBQnfYC+fDi0hCWSCp4705/6Q+F0PKpWKBJZwn9knSEWSdl1eD+fN8YMCnRxK4kvluNFo4Pnnn7+Rx9vAwMDAwMDA4J7AgSZpo9EIFy9exMc+9jFUKpU9PXYURWi1WlKMMp49mUwin8/DcRxZAwdR9/t9bG1tIQiCmJ3NcRwUCgXpU9P9SuVyGQ888AAee+wxHDt2DPPz83Bd95bW3mg0hFDQ2qgVHBIn9lz1ej3UajV5DXuX2O/F69Skj1ZHDnSmXZOhI+xRy+fzMsxYz/3icGoqMcViEZlMRkgZRyAwhCKbzaJYLIrSNZlMZCi0JnP6H68FgPQMjkYjmXvHddDqSesmwy+oMJLkJZNJCeqgojSdTuE4DtLptMyMu1pPGMkN92SWIGp7ou/7yOfzMgi81WrJM0nFjiRVK3gMR6HKxjXSlptMJuVY1/OhxokTJ1CpVCScRYeyTKdTHDt2TAac8+edTkeek9kB4MCO9XRxcVGstQYGBgYGBgYG+xEHusrpdrv4/Oc/j+/6ru9CrVa7Lef42te+hjfffBODwQClUgn5fF5+l0gk0Ol0hKCwIN/a2kKv15NilWSAxezKyoocw/M83HfffXjsscfw6KOPYnl5WUImbhbsk5tOpxLooIkDCZruXWIvGAt6ElIqW0yEZO+Vnv9GdUmPJ6C1jtdNlYrpk9qKSCJExSwIAtkzqmHsWSMh5u/L5bKomJlMRqLiqRjqOWv6Hy2NnEHWaDSQyWRQKBSQz+eRTqelB41KUBRFQna4p9y7yWSCYrGIVqsl6pHjOHJNWhFkeIi2AVKF4qDvIAhk3Tw/VU0Asr5cLidEjGmXVLdIrnitwA7RazQaorK9G0qlEsrlMorFIkqlUozEM81TK4EMNAF2rZa0iPJ6i8UiFhYW8Prrr+O11167pefcwMDAwODewOj7H8erf/vInV6GgcF7igNN0oIgwNNPP42jR4/eMrF5J5w/fz4WNU7lQAdMUGViX8729nYsGp2gitZoNORntJB5nodHHnkEp06dkkTGm4Xuq8pms6LkUPkKwzA2+FnPHQMgUfDsQ9LkRgdEMCBEz74i0SBJ4egCHShBwsXj6H4tnmc2gp9ER6szwO6Qa60Q6nuig0v4j/eBBIjr1wOsda+Z7qvTkf1cm0675H2mkqSDQrgOvp9BKiRwPJ++j1zvbF8f7YdaQeTPuDcklDwWyVS9Xr+uuWQk8Jxz1uv1EAQBxuOxDK7Ww845eoH7xKAU3neOtMhkMtja2kK9Xr+p59vAwMDA4N7CyEtiYbF1p5dhcBdhNElg/L/XkFzfvtNLuW040CRtMBjg9ddfR6lUkqj1vUaz2RR1xHVd1Gq1GMHRs61YmG9vb6Pb7cYSAgFIMUvL2SxOnDiBD3zgAzh27NgtrZmKFcMmgN0wDSYkctA1rZc6Fp7khkqZJhYkAyQdTEdkiArnlaVSKdi2Laoez6GTCfU/Hpsz3mjn0zH9mqRp0sO9Z5ogFUMStXcK2OA52Z9HiyYJmu5b4xp5bm2p5Jqo3ukeLV4390yHe8yuWxMqEklNMGcDOLRSpfeI5+XcPU2KU6mUzO27FvQ1RlGEdrstRMy2bURRJOErnB+nP5jQoxVI0orFIsIwRKvVErXQwMDAwMDA4GBhNEli7l99A+O19Tu9lNuGA03SCN3zcjuwsrKCs2fPIpFICEkDIEUqACEXnDe2vr6OVqslxTH/y2CKd0rXe/TRR/Hoo4/e0nqpeNCSSEWKyhEVDRIZDqSmmkZ1SROTIAiQSCRw6NAh+VmpVJJrZ/oig0vY/0UyoZMUqbBRGaLip2ebpdNpNJtNIWws9HXfGnvmeOzBYHBFgIfneUIMCSZEMhafhI1fk4CwZ47DyfXIAJJgWiqDIEC73RZiSvWJSZIkd+l0WvrdOEMOAObm5uC6rqizjOanCslroLWSah17/YBd5Uz3H/IeUF3MZrNYXV2NzTt7J9DWGgSB9Fvy/lGxY0DJcDhEvV6XDyJobWTyqWVZKBQKWFhYwIsvvohOp3NLz7iBgYGBgYGBwd0MQ9LeA5w9exYvvfSSFP21Wg2lUknsdbpXiHOjms2m9Eexr2s0GkmIxjvh5MmT+NjHPobHHnvsptfr+z5835ehz7lcDkEQoNvtIgxDADv9fADEYmjbNsIwlNeQzJEA8BjNZlPIBrBDeCqVCmzbln4q2uR08IUmBb7vo9PpCNliPxhVoVQqhW63K4olUwkZHEJ7ZBiGsWHTiURCUiOpPOlByiQ6hUIBg8EAtm2jVCqJYscAEBLFKIrQ6XTQarUQRRGazabMBmOMPI+dzWYl8KXb7coe+b4viiV72kiuqFYCOwPUObCbxJ9ENJlMot1uS+AJ7awM8+DrqegCkD43Kp0ArrDavhMSiQQefvhhHD58WGzEmhBzTQxN4fmLxaI8c+xVW15eRi6XQy6Xk1EUzzzzjAxbNzAwMDAwMDDYjzjwJC2KIrz99ts4efIkPvCBD9yWc0ynU3Q6Hbz88stS+OqoeN1z5XmeFPo62Y5fU4VpNBqxlEcikUjgyJEj+IEf+IFbXjOPx/60dDotSo8erMzADypvtKxppSuKIuTzeVFP+DoNEiFGw2ubG4dbZzIZIV3ADvmrVCpoNBoxeyAtglSDeD6ShXQ6jWKxKK/XoSU6cp69UcAuIWVwB22hvIckObpnzbZtId4k2tyffr8vyZmJRAKtVgue54l1lGqhtkcy1ZIqrE7ypOqYzWaFyOvQjTAMpYeR52QfHPeTaiBVP92DFwQB1tfXrzkjzbIsHDlyRNY+Go3QaDQkuZNEk/P2SOx5z3Xvo+/78DwP8/PzKJVKWFtbu65+OAMDAwMDAwODexkHnqQBwLe+9S0sLS3hwQcfvG3nCIIAr776qihjuVxObHa6T4jFPPu6er2ezCjTtsxLly69Y19QrVbDJz7xCVSr1SsGJF8v9PBmBoCQ3FztNSQAhE7y0yEZfD2vg/Y8Xjt702jtBOIBFuzD0sfnkHDuj/5HXM3SqgdW6zRBvpbrpLKm75UmKrPBHuzr4p7xftJGyWORVLJXjDZF3W+Wy+UkuZLH0cOxuUYSeD0EWpN7Eh8dFsJr4H/1cWf3j++/nj6wRCKBarUaC0rhXDbuh97v4XCIIAhifXzc0zAM4TgOyuUycrkc3njjjdtqTTYwMDAwuPvgXArRfP72pHAbGNytOPAkLYoiPPnkk5ibm8OpU6du23mCIMDLL7+MdrstPV20u7E4JtFgkd/tdtFsNmV2FIlAMpnE6dOnhcTMolKp4GMf+xgeeuihmw5E0bH1OmUQ2J3LpXu1giCQQdoc4KznbWUyGfi+H7O8sSBnXxSvW8f00zqoiSvnkgG7CpDjOIiiKJaUyb3S8fXcYyqSvC49Z0wPU9ZJiPzetm35Pa+DyhlJLGejcU0kGyRJ2hapB1ZzlAHnzzmOgyAI5D0kjCR1HNDNVExeP1M5GaPP48+mPGqCzNCSWaLK9wK4rl40ALFZbvzggRZH3heS/tFoBN/3pUePzxUVvHw+j3w+j8lkgueee86QNAMDA4ODhq+/iJP/6OydXoXBXYLhOImgm7vTy7jtuCZJsyzriGVZX7Qs61XLsl6xLOt/vPzzimVZv2tZ1unL/y1f/rllWdY/tCzrLcuyXrQs6yO3+yJuFRcuXIDrupibm7ut54miCKdPn5bCs1gsykwvFtAM1HBdF1EUodvtSmw5AyV838fGxsa72r6y2Sw+85nPiKXvRrG5uYn19XUhF1S3qOyxkE+lUnBdF0EQiHpEotPr9ZDP56U3jQES7MECIHY29p8xqp0DwElEqbj0+/1Y2iH7y3hMEpBCoQBgd/i267pYX18XyyTJcb/fh23bsG07NjCZNsVSqSTEgqSCQ8rZ10UF0LZtjEYjNJtNlEolsXPyWjhgG9ghl+VyGe12W9SpfD6PbrcraZe0PAKQIBDuey6Xk+eBhIjrJPmdHUnAdXL+XqFQkB5D9vHpGXIkZCR6vu9fl91R7zvVsnw+j1qtJtbPZDIJx3EkUGR1dRVra2uxBE4qcouLi/B9H2fOnLmpZ9nAwGB/4CDUIwYGBtdG540KTv7YC4j2efvD9ShpYwB/MYqi9wP4dgB/xrKs9wP4ywA+H0XRAwA+f/l7APh9AB64/O+nAfzTPV/1bUC73b7tiXHT6RRf/OIXJeRhcXERCwsLsG1bBloPBgNJQmRRzch9Fs5MVazX67EBwFr5yGQy+LEf+zHcd999QohuBN1uF51OB0EQxFQahlfo1L8gCKRXTPczMdiCISJUdBhKEQSBKE9UTqIowvz8PGzbjs3R0umOHLBNa2ShUJCERCpNJDBcE/uher2evKfb7cqwZgCwbRu9Xk8IJlMueUzaVEejkSh5VL6o+gCQOH6SJT2qgOSKqiHJn06CJLkhEeV+c76YtsfSjqnj/WmlJCHkWADeM9okOd+NSiPvDQkeB22ToAdBEHveroZ8Po8PfOADkj7Ja+V1aNsqAHnGhsMhXNeVhE9gR707duwYqtUqLly4gKeffvqGn2MDA4N9hQNRjxgYGBgA10HSoihai6Lo+ctfdwG8BmAJwA8B+NeXX/avAfyBy1//EIB/E+3g6wBKlmUd2uuF7zUajQYGgwEOHz58W88TBAE2NzcRhiFc15U0PiY9UkkCIBY/kgk9gBjYGZS9vb07xG92BlapVMIDDzyAWu3Gfdw6uIFKHsNBSJoIKjU6XdGyLFEDGW3Pn/PabNu+YnaYDtegstXtdmMKEYt4fk+yowmiTkXUgRokPjy2HvRM2yn7ymhjJEkiSdWqG1Midc8drYgkcjoshEPKGY7Cvaal1Lbt2HyxIAhQKBTkOkkK2Q8IIDbvTa8LgKQiMsCD10cVUp+fCiWfMW1vDcNQkirfDbZt48iRI7E9IinWQTEcX9DpdIS8D4dDDAYDURA9z0OlUsH29ja2t7ffNdXUwMBg/+Og1CMGBgYGwA32pFmWdQzAowCeAbAQRdHa5V+tA1i4/PUSgIvqbSuXfzZ7rJ+2LOs5y7Keu9FF3w5cunQJg8HgtiU8arz99tvY3NxEOp1GqVSS9EL26cz2EFHVob2RCtva2lqsaGbxrW2Qjz/+OJaXl294jbQXdjodSW3Us89oGdThGFwblZlcLhfri2LPFwCZWUZSRgJh27YEpZB40V6p1UVNBvXQbN0jp9UmrcYxYXI2xILrpk2Qx6IiR0WKc9d4PhKLWeWMJI/XkkgkZP6bPh8Jy2g0Elsjf9fr9WJpk1yDJnL8WqthJEi0Ws4GoOgkRe4z91YfC9ghVO12G9vb29dU0tLpNMrlslg/ST6Z7Ml7P5lMRK2l6qjDXRzHwdzcHDzPw9tvvx37MMLAwMDgdtUjI1y919vAwODuwGa9AGfNuvYL9wGum6RZluUB+P8B+PNRFMV8gdGOJHDtRpX4e345iqLHoyh6/Ebed7vwwgsvoNls3nJ0/fXgW9/6Fl577TUMh0PMzc2hVCrFCnFGxqdSKXieh+FwCN/3YxH3nJlGgkLooAgA+MxnPoPHH7+5LR4MBmg0Guh2uzJ3rFqtwvM8mQMGQAZwkwCw8GfPFu14VNtINjgfjr1qURTh0KFDMdvkdDqVQdkAhGjNBpBwADIVJj0rTId10ArIXjTG7muFjdZF3WOn182+LgaOMChFK4XZbBbtdjs2XFsfm+SLg79prfR9H/l8PhYq4/t+jGTyWFS9ZlMqmYw5m85I6yKPo6+PpFH3hHFPAGB1dRXnz5+/5jOj7ZN8TknseSzeO5I+qsSckcfwmJMnTyKTyeBb3/oWVldXb+oZNjAw2H+4nfVIGjfeHmBgYPDe4fBvpHHo7z91p5fxnuC6SJplWWns/A/iv4ui6Ncv/3iDtoHL/928/PNVAEfU25cv/+yuxmuvvYbt7W08+uij78n5Go0Gzp49i0qlglKphFQqJamDxWJRvgZ2iMn6+jqazSYGg0EsFfH06dN47rldMXJWBTl06BBOnTp1U+MFOp0Ozpw5g3q9HksVHI/HaLfbYoFsNBqx3rAwDMUimM1mkc/nYdt2bDh0NptFLpdDu91GPp9HqVSC4ziiJnmeJ3ZJpi3qVEQqMe12G/1+X2aKcU0kwJGaf0ZCxOO0Wi0hLAwCoWWQRIJEMQgCsXpyKLZWyBqNBjqdjpBK3k8mV+bzeQAQxSqTyUgfFnu1GExCcjOdTlEqlURF5bUBELsinxPaMLnHei7baDSC53liWyQ5o8rJcwGQDwiSyaRYFLvdLlZWVq5JlB555BF827d9G4rFIqIoguu68DwP2WwWrVYL+XxeSBj3stlsIooiCY+ZTCYoFosol8tIJBL43Oc+d12x/wYGBgcDB6EeMTAwMACuL93RAvAvALwWRdHfV7/6zwB+4vLXPwHgN9TPf/xyqtK3A2grG8JdCx3u8F5gZWUFTz/9NAaDATzPk9AIAKKWMVGPYRW+70t/GpWPTqeDS5cuyXGZxNjtdgHshi985CM3HmpFixyteDweiSKJDINPZgdUz/ZKdbtduK6LXC4Hy7JigSNUjba3tzGZTORaSch4TgZnkFiRUAG7YwNIODi7jTa+wWAgAR0ke1QDqUyxT47nTSaT0t9G4qO/5n2gvZFWRQBCDEkO9RwwqnEMIaFaqJVS9p+5riuEkUEj+jxUBXUoh14TAzyoumlLqt4D3geSN+7/22+/jVardc3o+2q1ilqtJqSR9yeRSAhJ5PnCMMSFCxckLIaKWiqVwvz8PKrVKobDId58883rjv03MDDY3zgo9YjB1THZbiL//8qj0Xbv9FIMDN4TXI+S9h0AfgzA91qW9a3L/34/gJ8D8H2WZZ0G8Hsufw8Avw3gLIC3APwKgD+998u+PRgOh2i1WjKI93ai1+thfX0dm5ub8DwPxWJR5m+xyKYqpfusut2u9Pmw2G00Gmg0GvI+pgMSx44dw+OPPw7XvfH/YZtOp1JAMwBkMBiIzY/EhAqQnk9G4sOeJx0JDyCWgEhCyL4lhonQnqePPTtoGdjthdPWP76G+6d7/YAdUqIHipNw8TX6ddwLrpH7TMKk+9/0zDQd5jHb70XiyH2kysg105LI9WuroyZ77FHTEfw8vt5/WlT5GqpamljrPj/2i62trV1TzapWq9JfOTu8XKt6vP7BYIB2uy37yj7BYrGIarWKyWSCV155RdRJAwMDAxygesTgSkSjIRJfeQHjUfJOL8XgDqH97DzciwcnRCx1rRdEUfRVAO/Uofepq7w+AvBnbnFddwTdbhfnz5/HqVOn8Pzzz8d6u24HRqMR3njjDXziE59AuVxGEAQyh4oDnaliJRIJ+L6PdDqNxcVFALvR6p1OB2fPnsWjjz4qBbLneXKeEydO4IknnsChQ4dw9uzZGyp6p9MpWq0WOp2OWNKo7LDvSZMbHptKDu2FqVQKxWIxRlhIyLTVkxY7HT5C1W00Gol1cTbtkeQF2CVlPC5JDcklicp4PIbnebEeM63M6cHLvF+cyUbyONt7pVMVqUJGl8cGsG8O2B3CrddE5Y33nwRZB3pks1lRzqiApVKpWG8ij0fixv5GHpN9crwmrgnYDUkJwxDdbheDwQBra2vXTFY8efKkJJVyj7g2Kok6qISknyMYSOTm5+dRKpVw/vx5PPnkk9f9nBoYGOx/HKR6xMDA4Eqc+KWzGK+t3+llvGe4vXLRPYbV1VV88YtfxI/+6I/KMOTbiV6vh//yX/4L2u02stksKpUKDh06JIOA2avEZMR+v49Wq4VGoxEjBsPhEC+88MI72jUdx8Hhw4fx8Y9/XIrn68VkMsFbb72F1dVVtNttjEYjZDIZCcVgBH0YhkilUjHbJsklyU8QBKKiMAilUCjEeslos8tmszLsmMO/GU+vkySpzmkiwHWTqJI8ZbNZFAoFmVHGwBOdOsi1aELGYddUCzmAm+SGgSiz6ZE8B5VNWiW1xZLDsieTCcrlcuyDAZ1oqSP0AcgzUalUkMlkYomR+jw63EWrbrRx/v/be/MoOa/y3Pf31TxXdVVP6kGjhTxibIyv8ACxDSYJEMz1WVlOfALnxAmQnCRwgJtFIOvesy4ra3EIh3C5N2QgCQlDjHEciGwmY7CDB9mKhCVZsqZuqdWDeqjumufpu390v1u7WrOtHiTt31q11F1dVXt/X1Xb39PP+z5vJBJRc9DEuRWB12g0mJiYYGxs7JxO2pYtW+jq6lLDwnO5nDrvIk4lBEUi9eWYxJns7+9nYGCAbDbLzMzMWdczGAwGg8FguJwxIk1jamqKF198kbe85S34fL5lW3doaEgJlp6eHpxOpwp8kN406ZcCmJmZIZPJqJ4jmHcBR0ZGzuh4BINB3vrWt9Lb29s24+x8mZ2dVYJGeo6k3E8vL5QL73w+r1wwcZSkj8ztdqs+tkAgwNzcnOrNq1QqBAKBtiCPfD7fdsEvvVp68qI8Vh4XDofVeAIpzdSdNrfbTSAQoFQqKfdMnDcdPcJf3DPdXZRQD7fbrY5RBL6UEoozpUfN632Fi0s45T3VXcdIJKKGccugb0nGlFJSeW/0ZEp7YWi1/jxZQ9w96X+T90bmlVUqFfbv398mDhcTCAR45zvfSTweV8cCqGAYOV4p3xWRJiJMRLXH46G/v5+Ojg5efvnltjAcg8FgMBiEa/5kitzz3Su9DcMyMpsJkfi9Ko3p5EpvZVkxIk1DBJGEUywXO3fuVHPTOjo66OrqUo6IXOjqrozE14t4gHnXat++farPZzGBQIA3v/nNbNiwgUAgcMF7nJ2dVQmGImz0MAxxYqSXDNqdLofDQTgcVv1HsncRVyL0xIkRcSFzyaTvS8on9REAUnopryXryh5arZbq95K+NkAJJTg50FmEjB67r4d0AG3z1QBVzihr6T1rem/j4t41EVL6awFtASRybqrVqkpdFIEqa8uxiVsm78Fi90ycNd3BkvMtfxjQ+8OKxSIjIyNnDdRxuVysW7dODQoXkStBJnqZpQxmF9Et7wFAOBwmHo/zyiuvMD09TbVqZhUZDAaD4VQaEydwlVd6F4blxG5aNI6PQWv5Av5WA0akLcK2bYrFouq7Wg6Gh4eZm5tT8e1dXV2q9E6GXEv/lzhChUJBuSbiHo2MjDA9PX3agcNer5fNmzezdu1aotGo6gU6X2SYcSqVolgsKkGm92zpw7b1PicRB3o/mJTySf+ZiBk9eVHEhj4sW+8Jk2PXBY24T9LbpYtY3U2T15Kf671aen+dPEbOsS7IdOGjP0/eFxGXsob0ewFqL/I+6+El8nM5h9JzJvPDFgeCLH4d/RgXizS5T0/G9Hg8yonTk0UzmYz6XJ6OQCBAd3c34XC4LaxEEiX1JEkRpYVCQbl/sp9AIKCCc1555ZUz/qHBYDAYDAbDlcVcJoT7qH+lt7EiGJG2iEajwb59+7jqqqvo6elZtnWz2azqTUskEqoPSb/AFvdDouqLxaK6OZ1OSqUSBw4cYGho6JTXdzqdxGIxBgcHGRwcJJFIXND+yuUyMzMzjI+Pk81mVVmllAmKm+bxeNqGXefzeYrFItVqlXQ6DcyXH4pIFIEgTlA0GlXHKbdisaiEiKxl2zalUkm5el1dXfh8PiVAZA6Z3tclQR9Suihz3fQRALobqLuEekmg9NFJWaAIL93dKhaL+Hw+JVakl08epztYukgFlCCTveslil6vl2AwqM6RvJZesqkHn+giXh9YLe+DlDLKc0SA53K5s85FsyyLDRs2cOedd6rjlz5K2Y8MaJeAm1KppJxYeU5nZydr1qyhs7NTxfKfK6TEYDAYDAbDlYF/V4B1/9eVMbx6MUakLaJWq/Gv//qv3H333Vx33XXLtu6zzz7LD37wA8rlMj6fTyXlSZ+TnoInYQ+5XE5F9It4Gx8fZ3h4WAmixfT09HDDDTewcePGC95joVAgmUwyMzOjyguljykSiSjRJKWZIpYCgQCBQOCUeHspyZMeMyndkxASeY6U++nJi/IYcadmZ2dVcEe1WqXRaOD1elV/VrVaVcJNBIveS5bNZkmn08oNFEHjdrupVCo4HA78fj+BQICZmRksyyIUCqlwExmgLeWUwWBQOVPST6jvR9+fiE4pYdWDR+SY9Ll0MltN+v1EnEpwi6RwVqtVdR4ljEaSFSXERUpBa7UaXq8Xl8ulhqQ/99xzZ/wsbNy4kbVr1+Lz+dRnUPooxdkUISnHNDMzQzKZJJVKUSgUCIfDDA4O0tfXR7PZ5C//8i/NTDSDwWAwGAwGjEg7hXq9zksvvUQwGHxNc8VeK5J+ODExgdfrpaenh3g8rhLxRIQ1m02SySSzs7NkMhkqlYpy2USMjI2N8bOf/ey0a8TjccLhMF1dXVx11VUXtMdcLsfk5CRjY2NqhpbE4udyOfx+vyoztG1bhYMASiQBqrSvXq+TSqXU99VqlVKppFItpXdKBIm4YfosOHGLBJnZJsJG3CgRTeLuiCiUcBO32004HG4ry5QeLb1MU1wtPV1ThKR8L711UpIpojGTybQ5erqL53a78fl8lEolJcjELZTgGL1HThdAchORJMcp6zcaDcLhsDpH8vrNZpNgMKgEVblcJpfLMTU1dc7h1Rs2bKCzs1ONMZAAFXH7pORV3oNcLqcSIvUexkQiQT6f58CBA8s6TN5gMBgMly6D353A+mbnSm/DsNR8q5PB755Y6V2sGEakLcK2bVKpVFtZ2HKRz+fZvXu3Kk0Mh8Nt87/kIl1mp0kCX7lcbhsCncvlGB0dZd++fSpSXQIlRDhJ6eOF9N1J8MPY2BiZTIZisagcJHF9pDdMRJoubsS90csCReSI0BFHZ/FQ6sXDokUEASqAQxwyl8ul5nAtHtwMJ0WijtPpVGmHIpr0kky9D0zEmx7WofeowUknTl9HH96tn1PZs/QfioCRfei9cyL4BP0cydqyZ3kvRKjp8+UW9xHCvDtZqVQ4fvw409PTp/0MWJbFpk2b6OrqUmEhXq8X4JT3SW7yudHntcncPJ/PRzqd5tixY+f9OTQYDAbDlU3j2HGiR84+GsZw6RM9UqRxdGSlt7FiGJF2BiRFUC5Al4NsNsuuXbuo1+sEg0FCoVBbaqAe/ADzF8WVSkU5bVJiKMl8Tz/9NFNTU6onrFKp4HK51Lyw3t5eotHoBYnRWq3GxMQEyWSSTCajHBSgTWyJSyUlkS6XS7lmgogTvVdK72WSfiuJ/Be3SAShCBEZiiwulpTsSXKgrKEPVNb7tuDkjDQ9El7KLXVHSgIw5HjlOXqKpQhTOSa5ydqyrrhpsh+ZySbCVxIx5bXEgapWq21Ole4YyuvL+yDiNZ/Pq/4zud+yLMrlsnq8fEaGhoY4ceLUv1zJcPFbb71VfW7091PQe/jkXxnfIOWXMm6iXC4zOTl51v43g8FgMBgMVxaz6TCO2pVdYeM690OuTHbv3k04HOa2227j6aefXrZ1bdtmbGyMtWvXEg6H6e3tJZlMnuKmSI+Wz+dTc7ekzEzcnomJCZ544gkGBgaUIIN5AeLz+YhEIlx//fW8+OKL6mL9fGg2mxw8eJBgMKh6oEKhkCoVFLFVqVTUkGXLslRfllzEL3Yrpa+qUqmocAm/369KDqW00ufzUSgUlNPVbDZJJBLMzs4CKGHS29vbVhYorpKgl0zK86SUUE9AlOeLGAoGgypxUcJS9LlkUsYopZRCvV5Xg6PL5bISmeI42bat3Ck92bNcLqtIe3k9KYcUZ1VKKPXX0uewiZDUQ1FkJIDOyMiIGlq+mHA4zDXXXMPatWvVOAC9Zy8UCqmSV3nvi8Ui2WyWXC5HoVBQQrS/v59EIsE3vvEN9b4ZDAaDwWAwVOsurvov+2nVr+w+deOknYHHHnsM27a56667lnXdRqPBww8/zMzMDKFQiN7eXpVcKO6R9HlJaePs7CzZbFY5SdLfBfMCZnJykn379rFv3z4VTuFyufD5fKxfv54tW7YQj8cvaJ8TExOMjY0xOTlJKpVSJY8idGq1miqfE4Eg+5ZSTZi/8JfHSuhEMpnE4/EQDAapVCpK3IngELEiJYIAmUxGOVMS8S4DpGFe1IRCobYh4UIsFsPv96vySBE6TqdTCVFdtImYlPATKa0UR1HEV6VSUW6snAM9CVFcMcuy1PBnPcBDhG4oFFLvt4hAcev0eW56kqOIMSnj9Hg8ZLNZ5bjpzykUCmQyGTKZDDt27CCXy53yfofDYdavX8+dd97JunXruPrqq9myZQvr1q1Tpbmy53K5TDAYpNlskk6nmZiYULH6brebzs5OtmzZwsGDB808NIPBYDC8NnYfIvF7Vap14zcYLk/MJ/sMpFIpXC4X3d3LP9W+WCySTCaJRqMEAgHi8bgqG5MLcxFCUgJoWRa9vb3qAlxck0KhoERDsVhUM61kHlmpVGJgYIByuaxu50Oz2WRsbAyv10sgEFACUh8GLbPJ9BlnIgxELCwezAwnB1DrQqRUKin3Tdwuceik7E7OUaPRIBqNUi6X2/ql9LEGEi4iZYyyp0ajgc/na3Pg/H6/Ent6yak4SXIsuttZq9XUz/ReOX0empQYynG53W7y+XzbgG1diIm7pp9TKW/1+/2qnNLj8ahB3bK2Pvxb3gfZswhbSV7U+9QAQqEQ73jHO9i6dSvr1q2jq6uLWq2mSm2lHFT2I+JSgmByuRypVIpwOExHRwfRaJRKpcIrr7xyQQ6uwWAwGAyCXa/RPDFNy9600lsxXESmp6P0P+HCNmFiq0OkScy49BCtBuRCVYIOljsa/NixY0SjUTZs2EAikSCXy6kSMimDE1HSarVU2qKIB3E1Fg/BlvvFhbIsi0QiQaFQUHOqzpe5uTn8fj/xeByfz0cikWiLjdd7rfQ5b+IEibOkixk9TVHv29IdMXkcnAwB0csGAfU8mdkl50UfAC1piiIs5HnymiKAA4GACvBYvLYIRBHHOnrpoaD3ccm/IhLluHV3TPakB67oJaLi5sl64tzpIk3KUKW/T5+9J2MPMpkMR48eVWmhgsfj4Zd+6Ze47777uOmmm3A6nQQCATXcXNIoFwu7UqmkPlMyxLqzs1Mllg4NDTE+Pn7enzWDwWAwGE7BblE6GKN+VZ5IYPVcQxpeO46Mm+BjL670NlYFq0KkyUX+6OjoWWO/l5t0Os3s7Czd3d1MTEycchG+lOzYsQOPx6Oci1QqpeaiLXacWq0W2WyWVCqlZnfpJXO66yQpe7qz4/f72bBhA36/n4mJifOOQpdY+WPHjhEIBEgkEqqUUYYa6+KoUqkQCATaerakHA9ODnH2er1KSIlrpg99zuVybUJFHCQ9Pj+fz6v19dAROWeAcqVkf9I3J/1Yem+dnEsRjKFQqE2wAWoPrVaLSCSiygvlfl3oy3uoCzVxxKREVNId5TEyJFv2LOdDXFIRYtLXJoOtZW2Px0O5XFZCWFzHYrHI+Pg4L774Yttn3OVy0dfXx9/93d/R3d2tjqXRaBAIBFS6pwxS9/l8qoxTynAl0dO2baLRKB0dHUxNTfHEE09c2C+EwWAwGAyLsBsNNn5qO0Nf3ApXG5F2qVNrOHHUrXM/8AphVfSkTU1N8c53vpNf/dVfXemttLFjxw62b9/Ohz/8YdX7tJzMzMywb98+3G43XV1ddHR0qBJFiTEXsVIqlRgZGWF2dlb1OUniI5wM5RDXSG7Sj+Xz+ejp6eHd7373BaU9ZjIZDh06xL59+5idnSWfz6teNBnuLGJDLtZl3zI6QJIgw+Ew5XJZpTVKyIhelicCQS9vBKhUKkqkSFljKBRqSx+UGWXS31apVJTbJvfJcGtJpRSRqYdxyMw1vXRSBKL0A1arVXw+n9q7lBeKcyfiS8JFJExEjk2cThFvsh6cdMxkuLYuMuXnIh71/rpSqaTWlfcpl8tx4sQJJiYm2gSn3+/nxhtv5PDhw/T09KgyVTgpbiuVCul0uu3ctFotSqUSo6OjzM3NqVTHdevW0dfXx+joKAcPHnytvxIGg8FgMBguUzyPdLDxUztWehurhlXhpAFs27aNzZs388ADD/Dtb397pbcDzIdjhEIh/tN/+k+nzNVaDkZHR2k0GmzcuFFFlotTUyqV2lwpufAWkSKCSFIBpSRNBIUgLpYMIQ6Hw7ztbW9j9+7dpNPp89pno9Hg6NGjrF27VgmXzs5O/H4/xWJRDX7WZ38BqrzV4XCoQdx6+IWccympkzJPCT+RY5XyPinnk74yQPXm6cmTInpqtRqBQIBCoQCA1+sln8+3DeQWN0rEoDiS4nbJ/iX0RMQznIzZl943v9+vzpcIIpkdpottvbRRnwknQnXx68v3EsLh9XpVKqb0sumplbLnfD7P3Nwc+/fvZ//+/er97Orq4r777uMP/uAP1B8F3G63CkuB+VLXTCajzksoFFJljclkkrm5OXWevV4vV199NY1Gg8OHD19QSa3BYDAYDOdiy18naYV91BJ+Sh/LrPR2DK8RRxNomV40YVU4aQDT09OMjY2RTqfZunXrijhXi5HgA7ngX24qlQpTU1Ps2DH/V4VoNEo8HldODqD6z+DksOl0Ok2hUFAX6NA+D0y+t21bBX7I8OZQKMTmzZtZt27deSc+SpLi4cOHGRsbU71KuhgB1AX+YsdHT02UJES9T0zfr34Muruj91/JeVlctinP02eyiSCTn4vrpDuQsg8RbnoZpX6T1xAxI+JQP2a9904epwekiHOoz23T96AHh0gJJ5zs65TwE11oSn+b/FwfMj06OsrU1BT5fB6YT7q8//77uf/++3njG9+ogk10pw7mXd5MJqPeB5fLpXrbpqenlUvp9XqJx+N0d3dz6NAhpqenKZVK5/W5MhgMBoPhfGgeHsbetR/fjiOUn+ymWPWs9JYMF0j+2W4iw2ZAuc6qEWkAx48fZ/v27WzdupXu7u5lHSR9JqS/JhwOX1AZ4MWiVCrx9NNPk8vl8Hq9xGIxEolEW5qgfkFeKpXIZrNkMhklOOQ49GRCSTUUcSYX4F6vl56eHrZs2cLatWsvSKAODw9z9OhRJiYmmJubU0O2F8/sgpOiSvYB8wLkdHH3cgwiyiRBUn+u9IzpA7LlZ3qohwhXeU1xy+QxUvKozxzTy0alvHGxuJJ19DX0kA4RjbpoEpdOnqP/Kz+T8yfP1dMk9fult093/QQ5z/L64iLWajWOHj1KOp3G6XQSjUZ585vfzO///u9z77330mg0lKDSBVq1WmVyclKNPZDPV6lUIpPJMDc3R6PRUK/Z1dVFs9m8IHfWYDAYDIYLpZnJ0vulFyiORihUVv4a0nBumi0HM8kI6782jP0fr6z0dlYVq0qkAeRyOb70pS/x4IMPcv3116/0dsjn8zz11FO8973vpbe3d0X20Gg02L59O5lMhlgsxuDgILFYTAkREQJSVihhHqlUSrkgkgBZKBTUhb2U6AFKjMD8HLM1a9Zw/fXXc/PNN1/QXoeHh3nppZfYt28fc3NzbVHsTqdTXbz7fD78fr8SVXoK5OJIfulRk5I7r9erSvukn0yElPSlibMjzlar1VJJmCJaZG96r5ceTS8JkFJWCSdDSmq1mhoxEI1G21w5SVYUQaaPPNDj92Ut6ZULh8NqOLS4YfpsOP04RFx6PPN/LZQwEymBlYRKOX49BKVSqVAsFpmammJoaIhsNktXVxe//uu/zo9+9CNuuOEGKpUKqVSKarWq9iXnaM+ePczMzFCpVNR5mp6eZmpqSrlrpVKJRCKhxkL8/d//vZmJZjAYDIZlYfNHX8R6MbrS2zCcB+lskM0f/AWNqemV3sqqY9WJNOErX/kKkUiEd73rXSu6j1QqxbZt27j77rvp7OxcsX3s2rWLo0ePUi6X6ezs5JprriESiQDzIiYWi7UNiy6Xy6RSKRXcsDjKXoYOy89l2LTeO9bV1cWNN97ILbfcosTAuRCX5ec//zkHDhxQrlqpVGpLSBSx43A41JgAl8ulyjQrlQqlUkmJNhlALU6Q/KuPE9CdORF2cp8EeoiYbTabhMNhFUYiISSyR4fD0RZCIq6YnpSoizK9tFPKEaV0U44PUD10fr+fQqGghI44trpr12w28Xq9ql9PRKHf76dUKqmUTkAdsy7K4aRzJj10tVqNYrGo0hd9Ph8PPPAAX/rSl/jCF76gjr9YLJJKpejq6moLX2m1Wmzfvp1isajEXzKZ5Pjx4+p9bjabxGIx+vv7qdfrHDp0SCVdGgwGg8GwHAz81Sv4/t+Old6G4Szktnfzho+aPvUzsSpEWkdHxykCqFAocPToUTKZDG9/+9tXJLgDaIuOF9dpJajX67z88su89NJLyqXo6ekhFoupOPXF889yuRzJZJJsNquGK+tlgDL/TRcXi3sBfT4fmzZtYsOGDYTD4fPaa7PZpFAosHPnTo4cOcKJEyfIZDIqrl2Eg5Q06gEcsi84WZbp8XjU+Zc+LBFIIrDg5Gw7vR9Njk1i/fX3UNIRxTnTXTzpS6tWq6eUuYpIk0RIic7X++H0UBQpnRQhJ66YuGyLSynlZ3Lsshdx4aRUUu/DE4dNxi0sHiQuoxHksY1Gg1gsxu/8zu/w4IMPcvvttxOJRLBtm1wuB0BPT0/bsafTaZ599lnm5uaoVqtUKhWy2ay6iaj2+/1s2rQJl8ul3Doj0AwGg8GwnLTyeQK/OI7rHxI0W6victegUflhN2t/kKU5O7fSW1m1rIp0x3A4zMDAgJq7JUxNTWFZFrfccouaFbZ4aO5yIBeuXq+XQCCwYsEHJ06cwLZtEokEN998M11dXcpBkV4fPdlRStZgXgBEo1F1YW9ZlgrA0AWwuFvyWi6Xi3g8zvr167Esq62k7Wy0Wi3Gx8dV7L84S36/XzlielCG3qcle9CDQ6RcT4SOBGjoISTyPHk93RUD2kJC5HvdydNFjR6dDydDQ6RkU4I69MRFPUlR9inrybHI64n4WTwEW0oh5Tn6PLTFIk0ec7rXWLy2DKnWY/Q3b97Mvffey+bNm0kkEgBqbILH4yEUCqnXrFQqTE9Ps3v3buU+NptN1Ycmz3O5XHR0dBCPxxkfH2dsbIzZ2dlzfl4MBoPBYLjYNKdnCP1bhpmb34zrDWbg9UqTK/loDM3/wf+qn83R3H9ohXe0ulkVIs3r9XLXXXfhcDjYvn27upCsVqskk0kOHDjATTfdxI4dO8hkMsv+V3nbttm7dy/xeJy+vj6GhoaWdX2dyclJHn/8cd74xjfS1dUFzJ+nubk5JTrESQGUi1atVpVI04ch6wOYRZSIAJLUx1KpxLp164hEIsRiMf7jP/7jvAdej4yMqJI+GbAtPWmRSIRqtYrH41GleMFgUI0SEEetXC6r0kU5LnGx5HjldfQB3z6fT/WdtVotOjo6VOqkiB1dDIq4kfMiQkhKFmU2mQyElptt26r/TcogLcuiUCio8lLpO5NzoQeo6M6mHIe8J7JHed8WC1vLslSPnrxvtVpNjSqQz4I+ANyyLDo6Orjrrru46aabVECPDMsOBoNtQ7Rh/g8mR44cYWxsjEgkogar5/N5ksmkei8CgQDd3d3UajVeeOEF5ubMX8gMBoPBsHLY9RobPr2d4S9spbiphWXZBDz1ld7WFUet4aQ2EmLTp7YDYIL2z82qEGnJZJKbbrqJ/v5+QqEQP/rRj9TPisUiBw8eJBqNcu211zIxMcHIyMiy7s+2bf7mb/6G3/qt3yIYDK6oSIP50rVHHnmE++67j1AoRH9/Px6Ph/HxcarVqnJ2JIRCyuD8fj/RaFTNTpML/3K5rEodxRESF0pGENRqNTo6OvD7/cRiMZ588sk2p+tsTE1NMTs7y8jICHfddRddXV1qf9KDJi5RvV5vmzUmpYXQXo5ZKpWIxWJq3letViMcDisny+FwUCwW1fPkeEXEiQskwlVElKQeBgIB1eMnLpEIOHHP5DVFiBWLRSzLwu/3q34xj8ejShEDgYASYADBYLDt9fXSTV2YSQ9ds9lU+5SSTxkqLedPAmFEoAFqr8ViEa/Xy5o1a7j22mu544471Hsk4xvONHZh586dPPvss6TTaRwOB3Nzc0xPTzM3N4fT6SSTybB27Vp6e3sJBoP81V/91XkLeYPBYDAYlppN/8dLYDlwrR9k7v9bFZe/VxTuf4mz6RsvrvQ2LilWRZFuPp/nxz/+MQAPPvggb3rTm9p6o5rNJnv37gXmh+bKUODlpFwuE4/Hz3t22FLSarUYHh7mpz/9KUePHiUYDLJhwwZ6enraXBh9UHKtVmN2dpbZ2VnVL+XxeNqcNRFs0m8kKYJer5dIJEIgECAYDNLb28u73vUuBgYGzntMQqPRYGpqipdffpkjR44wOjrK+Pi4CuyQ/egphuIi6Z8F3d2RPUsZop5m6Ha7VW+VCCk5D+L4SO+YOHuyjh6yIqWZfr9fHas4ZbVaTQVyyB6azSbFYrHNZVvsVMprVyoV6vW6El7iFsp7LIEnXq+3LYxEL/EUF05CQ/SAlWq1qnoO5dh6enq45ppruOmmm9reH5fLpYJoFvO1r32NnTt3qpCZ6elppqenKZfLKkWzu7ub3t5eyuUyjz32mBFoBoPBYFhd2Da0mjTHTtDxJ24yheW/lrxSCXwpRuKHQ/PvgeG8WRV/Smg2m7zwwgu4XC5uu+027rvvPvWXermALxQKjI+PEwqFWLduHQcPHlzWPcoMMkmuO5++rKWkUqkwPDyM1+slFAoxMDBAV1cX5XJZCQVBhII4ThJnLxf8euS93ncFKJEnce/ymO7ubjZv3kx3dze5XO683MVarcb4+DjNZlOVDA4MDLT1f0np4uJBz3oZot6XKGWJIpYANaNMXk9PldRLBWVNvWxSetvgZL+X3iOn94DpTpecN3Gy9HEC+kw4vd9Mn+mmH6NeCimz0SRdUt+T/n7JSAERj5KAKSMGJFVyYGBAOdaL0UNCJOFx165dHD58mJmZGYrFohJ+MsoB5h3B/v5+crkcY2NjTE1Nnf8H2WAwGAyGZcSu17D3HCD8o7cyvdVHz4CZ37lUlGpunD/pIP4fB2iaOakXzKoQaQD79+/H6XQSiUR497vfza5du9i5cydTU1PqovX48eOsX7+etWvXMjk5SS6XW9b+tJGREQKBABs3buQXv/jFsq17JpLJJCMjI4RCITo7O4nFYpRKJdW7JSV64s6Uy2UKhYJyrNxuN4FAQL2eiAopp5OLcCk9FMdHHK6rrrpKRccnk8nzej9SqZSKgbdtm3g83uaISRliKBRSoSP6UGsRRXoIiN47JvfpUfYOhwOfz6d600TkiXjRhao+AFy+X1x2KKJI37ceUHK6gdeyJxHGksKoizZxNBeLO+kPk1EA+n71FMtyuazuk+MXVzQQCBAKhVizZs053eBKpUKhUODEiRM88cQTFItFKpWK+vyUy2W1D5/PRyKRoKOjg6GhIQ4fPmySHA0Gg8Gw6on/w3YavtuYdsbAgp7ezEpv6bJhNh2mWXFiFVxs/ssXTP/Za2TViDSAvXv3MjExwcDAAH/8x3/Ml770JZ544om2IbgjIyOkUinuv/9+Hn74YVW6txz88Ic/5NZbb+X+++9fFSINUPOpWq0WN954IwMDA6o3S5xIcV7k4j2ZTKp5aNJDJI6ZiASZzyU9UiKEpPzQ7XbjdDoJBoPE43He97738S//8i/nlXxZKBQolUokk0kikQhr164lFAqpxET5F2gTinIcMgRbLyGUgdjyGBFfEugRDofJ5XJtiZBSFinHDSfTFXXBpc9Wk/1ImaTb7VaJlfJ4/TwVi0XlrOlBJeLMwsmyTRF8Pp9PuX5ut5tms0kgEFABJpIiKcme5XKZSqWiXDlAOdBS7uh0OtXn4lwiav/+/bzyyiscPHiQfD7f5kYmk0lgfjSDz+cjHA6zbt06jhw5wvHjx1fcYTYYDAaD4Xzp/soLdAOOYJCZb/fjdp5fr73h7Kz/qoXzmZ0rvY1LHms1/NXbsixb+5pAIMAXvvAF+vv72bdvH5/+9KcXPx6Px8O9997Lnj17GB1dvkF473jHO/jABz7ABz7wgWVb83zweDzcf//9XHXVVXg8HrLZLGNjY0xMTFAoFFRwSKFQwOfzKVHW19fH4OAg0WhUOWh+v7+tNyqTySihLCLH7XaTTqeVUPN6vRw7doyhoSFGR0fVxfz57Hvjxo1s2rSJtWvXqpI8cY/EvROxJK6W7qxZlkUqlWLNmjVKUDmdTrLZLC6XC6/Xq0JB9CHM9XqdYDCoBI3X620T/SJO3G636iMTkeT1elU/nThk4oBJaaE+YFr6yaSUVJyo3t5e5fBJj1wwGFTriRgV0SjPlZu4guVyWfUYNptNZmdn1bnyer3E43F6enro7e1lYGCAwcFBFYwC88JZ+hwzmQyFQkGlU5ZKJSqVikpylFEUXV1ddHd3k8lk+P73v6/c0UsF27ZXZviiwWBYtUSsuP2/Wfes9DYMK4AzEefAn11Fz9rUSm/lkqP2eBe93zkZp9/K5rBXYGTWpchL9k/J2anTXo+sKicNTsaAf/WrX+WBBx5g8+bNfOITn+Av/uIvlHCQXpuXX36Zvr4+Ojo62LNnz7Lsr1AoMD09TW9vL8lkctUEJNRqNZ5//nkVlx+NRhkYGFACoFAo4HK51IW59CylUikCgQCtVktFr4uAkdAKCWqRi3UJ2ZCkw0qlgsfjoa+vj0gkwqZNm0gmk6RSKY4cOXJWd61WqzE2NkahUGBycpJ169apUke/3088HqdUKqk9SI9WpVJRwsW2bZUIKSJGHCtxsXQRpc9p01ncEydloXpfnLhhsq7cL4JIhkZ7vV4lrhYPCBfHT8ShiE/dwdOHTgNt/W7ynuq/DzAfHiL7kHJLcT1t26ZcLlMul5mdnaVer+N2uxkfHyeVSqmB1MeOHVPlmBJuksvllNgMBoN4PB56enqIRCIqal8GWRsMBoPBcCnSnEtx1Tdr1CMdVOJO+E0z4/NsFCpeOv8+ADbEj0zTnDPi9mKz6kSa8Itf/EI5Kr/0S7/E008/zcGDB9su+MfHx4nFYqpkbjkctdnZWfbv38+NN97Iz3/+c9UHtBoYHR3F4/HQaDS45ppriMVidHZ2tg1r1oMpJGAklUqpNMhYLKbEipQMSqqhuDf6TC49kCMWi9Hd3U2r1SKXy5HL5YjFYoyPjzM9PU2hUDjtviWQQtIDA4EAgUCAeDxOJBKhUqko0QEnB3br881ECMmx6YJMBKmUFMqgaQlCEREjZYj6aAERYnpvGbT3qsn9eomkXvao3y8unj5YWl9LnDbZoyRsSg+evAfyevrr6KEieiiJvHaxWCSbzVKpVJicnMTj8TA0NMTMzAz5fF59HuS8NBoNSqWSel8cDgd+v1+9N5VKhdHRUcbGxi7wk2owGAwGw+rD8dxuvEBwTS9D12wgeEMKj2t1/DF+tTBzLIE748BZsvB+/wXAzDxbKlatSAPYtm0bo6OjfPazn+X3f//3+cIXvsDw8LC6SAXYt28fAwMD3HnnnaoHayn/oj82NsbTTz/NJz/5SXbu3LmqRBrA0NAQpVIJj8fDW97yFnp6eoB59yadTqv+LTgpJlKplAqGENcpHA63zS3THaXZ2VklAOT+Wq1GKBQiFovh8/lwOByEQiHuvvtudu3axVNPPcWRI0faYuF1Go0G2WyWvXv3UigU6O7uVgO09ZAOWVfElhzP4kRKEUQSPCKzwlwulyoT9Pl81Ot1AoGAcppkxpz0qImrBrT1jUm/mGVZykUSMSUhH1ImCajh0hKnL/s8nWMmyZfi+Ml8M0lulHOnC1DZrwhxeYyUY/r9fiW2JJ0xEAiQTCapVCpKGOqBMbVajXQ6rWaueb1efD4fnZ2d+P1+Dh8+zL//+78vwafYYDAYDIaVozE5xfrPTHHkn27GH6ridLau2AHY1bqLWuNkoNqGxxq4frZrBXd05bDqetJOh9/v5wc/+AHHjh1j27ZtfO973zvlMW63m9/8zd/k8ccfJ5VaWss1Ho/z8MMP88EPfnDVxo273W7uueceNXOuUqkwOzvLgQMH1IW99PZJdLvb7SYWi7Fu3Tq6u7tV+aCkRsLJHrV6vU6xWFQCanBwkI6ODgYHB+ns7MTn87Fp0yZisZgSELZt89nPfpZHH32U/fv3n/MYOjs7ufHGG+nv71cBJ36/n0QiQalUUoEf4upJ4IiIKF3Mw3zYhdPppFQqUa/XiUQiSsyJEJuZmVHOodPpbAvskOCQYrFIOBwGUOchHA6rMkvpLYvFYm1CUlIrZYxApVIhFAq1lU3KMG2h0WhQKBQIBoNKpMn7stjdEwHbbDbJ5XJtx+ZyuZTAkzJG+VyIcJY/cIj4q1QqzM3NtYn1devW0dnZyWOPPbZqP/vni+lJMxgMizE9aYY2FqpR6u98M6WPZVZ2LytE/d+66PxbbQj1KtANlxOXVE/a6ahWq3ziE5/gv//3/8573vMeIpEIX//619seU6/X2bZtG29605uYnJxc0jlq9Xqdffv20d3dTbFYJJ/PL9lar5V6vc7zzz9PJpPhuuuuY+3atUQiEUqlkipvq9VqeDweAoGActLq9Tq2bVMoFAiFQgQCAaLRqJrVBRCNRgHU8Gtx0kqlEg6Hg3g8Tn9/P9FotC3O3rIsPvKRj/Abv/Ebalj2V77yFX74wx8yMzNziruWTqd56aWX8Pv93HLLLXR3dytnTcSQ1+tVgR9S0rg4Tl9602SAtUTj1+t1NYy51WqpXjtxw/R5aIASOzKsWtwtKRWUfenx+HqvmpRkigAScSzInkT46n1oekKkOJt6GefidEePx0M+n1ciTpI6RcBKKaPuoEk4jIhwfTZcPB6no6ODQCDAd77zHdJm3onBYDAYLncW/h/q3X4I3/gaAA58MnLFxPX7v9xBfNcQTSPMVoRLQqS1Wi0OHDjAs88+y3XXXcfWrVs5dOgQu3btahtsnE6nmZiYwO/3s2XLFg4dOnSWV33t1Go1nnrqKa677jpqtdqyD9Y+X/L5PMPDw0pE3Xjjjar80e12k81mlTMjoqfZbJLNZmm1WpRKJRKJhHoMnJwt5nK5iEajeL1eqtWqCpZIJpP09PQQDofbBJrQ09Oj9tBsNnnwwQe54YYbyOfzpNNpHn30USYnJ1W5n8T1HzhwgLm5Obq7u6nX64RCIXw+H6FQSDlX+nw3STrUB0XrSYki3qRUUBwoQUSazBoTpOxSnitriKgRkSQlj/K6MqtMxJLuaun9cA6HQ82y01MtZZRCs9lUqY+Le+WkpFN+JmmUemCMvE+SdAknHTj5vlgsqh40SceMxWIA7Nmzh9nZWRMSYjAYDIYrhlY+D6/O/0G+70dbqcS6qEUtgnfNrPDOLpxaw4n1/Ticx//G47uGaJ5nWrfh4nNJiDSYjxh/8skncTgc3H333fzyL/8yo6OjKqlOOHz4MFdddRWDg4OMj4+rC8+LSa1W47nnnuMzn/kMMzMzq1akwfzA62q1SrFYpKenh2AwSE9PjxIxc3Nz6uJeZnI1Gg3S6bRyf/L5POFwuG2Yc6vVUrPN9HCJqakpOjo62LJlC4lE4qx7czqd3HPPPdxzz3xpyeTkJNPT0xw4cIBCoUA2m2V2dpZWq8XIyAi5XE7Fwnd1dSlXDVDCTVwhPaJfSvhECMna4nzJuajX68o5WizABHGvqtVqWxmnuHZ6UiOgnDSXy0UgEGgTP/J6tVpNDaqW81mr1fD7/ao/Tu8/k9ATPTBERJkIUHHfarWaem/r9TrlclkJNXHyRLyKsyalnSLEQ6EQDoeDmZmZVTMf0GAwGAyGlSD0nRcJAa4N63h1czcAgUSJsL/a9rhUNkg97znj63T05i5KKMn0ZAya51+9b9UcbP7qDmide20TCLKyXDIiDeYHWT/99NMAPPTQQ5w4cYKf/OQnjIyMtD1uaGiIyclJbr31Vp599tk2t+1iYNs2+XyewcFBIpHIRX3tpSCXy7Fv3z4OHTrE7/zO79DZ2an6uxwOh+otE2Gil8Mlk0l8Ph8dHR243W41c0xKBMVpEfFWKpXYs2cP09PT/N7v/d5p3bQzsWbNGv75n/8ZgJ07d/K9732PL37xi2qtVCpFPp8nl8uxefNmJVpE+Hi9XlVSmE6nCQaDSlA2m00ymYwK+9CFlXw+RKTlcjk1+61arapSQhE5hUKhLVa/XC6rhEjp7RMnzOfzqZLM3t5eFVIiLtuxY8fIZDKkUimmpqZUTL5t26qvTU+DlCCQYDCoHL1ms0mpVFJ7kscA6me6QIN5p0+GZudyOSqVihqnIHPgQqEQa9asweFw8MILLzA0NPT6P4wGg8FgMFwGNI4d5w0fOg7A8f/7rbTe3B7ZH30yQPxr28/4/MP/cAtdPdmzruE4e2QDANf+6TiNyUu7R9xwei6J4BAdh8NBV1cXH/nIR7jtttt49NFH+dnPfsbRo0dPeayk/C3udbpYfOhDHyKbzXLo0CF27969JGtcbDweD+9973vZvHkzbrebVqvF2NiYSvETh0VPSnS5XMRiMTo6OkgkEvT09FCv1/H5fG2hGpFIhHK5rMrxuru7ef/7309/f79yjhYj4u5098tekskkX/7ylzl06BD5fB6Xy0VXVxexWEzNP+vu7lbumQgvma0GJ8sXxYFqNBpYlqXCRMRpc7vdamCzz+drE16yTynpFGeyXq9z3XXX0dHRoVw5Qf9aL4eU3ztxsCYmJti7dy/Hjx8nlUqp/TmdTrq7u1XJpDiJuotWrVaVGNSHec/NzZHJZFToi6RKyuw0mX8mpZciWr1eL9FolFgsRqvV4oknniCfzy/Z79FKYYJDDAbDYkxwiOG14AgEsBZd57SqVexq9QzPAMdCivbZOPSVTXTFz557kHioYETaJcwlHxyiI47Kt7/9beLxOLfeeivhcJh/+qd/OiXVUY81By764Okf//jH3HHHHdx6662XjEiT4cOTk5Ns2bKFjRs30t/fTyAQIJPJUCgUVMS+lOI1Gg1mZ2eV2+Z2u1UoiIgNt9uN3+9v663K5/M8+eSTDAwMsHHjRrZs2XLKfkQ4iIMUCASUcHI6nfh8Pvx+Px/60IfI5/OkUimef/75tplr4v7pM8MkAl8Eis/nw+VyUalUVFlio9FQc+Wkt0t6y+RrET0iXKLRKGvXruUNb3gDwWBQ/SEgGo2eItDOht67Js6bCGJJlISTf2iQPUrZouxtscsGqPEAEs4iPYWSglmv18nn8yooRu6v1WoEg0E6OztxOp2k02n27NlDoVC47ASawWAwGAwXi5Y2w/e8n3MeoXMb/waa/thZH9NMmVmllyuXnEiD+bK0Q4cO8cwzz3Dvvfdy7bXXcuedd7Jt27ZlDTQ4fvw4b3vb2+jt7V22NS8Gk5OTyj1xOBz09/ercka/398WLiHldJVKhVwup8oKfT4fHs98rbX0QYlbJV/D/IDtfD6v+qw2bNhAIBBQPz+T66Tj9Xq54YYbAFSp4ZEjR5iZmaFcLqs19bJWKf0Th0zCMUSUiEiToBFxzCqVStuQbnlsIBAgkUiwZs0aQqEQg4ODF1TKeTZKpRK5XI5yuazmkUkfnMx6kyh/EWaS9iiIqJPkSBmtIOdET7SsVCoUi0XVwyafA0nytCyLTCbDiRMnzKBqg8FgMBhWCMezL+M4x2NWvh7OsFRckiJN+O53v0swGOTuu+/mt3/7t3nmmWdOW5Z1sR00HXGOpH/pUmFubo50Os2JEyd497vfTSKRIJFIEI1Gqdfraj6Wy+Uil8upHqlsNqsu6KV8UMSCLgjkubZtk81m2bNnDy+//DIPPvigmnkmIg/mZ+Hp5YlnIhQK8a53vYvu7m527tzJ5OQktm2rwA2YF2jBYFCFZoggE8GlD4OWtEXZs8wkE+Ei/WTxeJw1a9awfv16QqHQRXsfGo0Gk5OTjI6OkslkVMmh7E/KEKVfTpCSUjkmEZrlcplMJtM2lkBKNU83E00SMS3Loqenh1gsxtGjRzl27Bijo6MX7TgNBoPBYDAYDOfPJdeTdjpuueUWPv3pT+NwOPj4xz9+2v60peJNb3oTV199NT6fj2984xtLKgiXkjvuuIPbbruN7u5uyuUyY2NjzM3NkcvlVHKjDIsWZ6qjo4NYLKb61UR0iSNWrVbVfRJQUa/XSSQSXH/99bz//e9/XXvevn07Bw4cYGxsTM15k3LBzs5OMplMW7hGOBxWYlO+h3nBKomJ4ha6XC48Hg/BYJC+vj42bNjAhg0bGBwcfF17Xsxzzz3HyMgIqVRKiUxB+tVkcLeURc7MzJBOp1XCZCwWo1wuU6lUKJVKypHTRWmpVCKTyShXMBwOq+TTUChEZ2cnmzZt4ic/+QmvvvpqWznp5YrpSTMYDIsxPWkGg2E5OVtP2mUh0nw+HwMDA3z5y1/myJEjfP/73+fJJ5887WPFObhYx+31etmyZQsf/vCH+djHPtYW134p4ff76e7u5qabbuL2229X4iybzTI6OqrmbcF8yaEkOno8HiKRCFdffTWBQECFUjgcDubm5ujo6MDhcKhod0lClPs2bNjAHXfcwcDAgBJN50Oz2eTw4cMMDw+rUQwyYNnlciknLRAIAFCpVPB6vTidTlXGKbPcZOiz9H9VKhX1On6/n87OTq655ho2btxIX1/f6z7X2WyWEydO8Morr5DJZNp6x6ScUcotxU0TWq0WMzMzahxBvV4nEomQy+XUcyQARcJX9DAYcTClX2/NmjXEYjEqlQo7d+5kYmJCpWle7hiRZjAYFmNEmsFgWE4uq+CQ01GpVBgZGeEf//Efee9738vb3/52Wq0WTz311CmPbTabhEIhNej39VKtVsnlcpRKpUt6wG+5XGZychLLsshms1x99dVEo1F1sV8sFqnX61SrVQqFgkpErFarZDIZJiYmiEQixONxNZZAyvREkElIhtPppFqtMjU1pcRbLBZTz7/uuusIh8NnTISE+YTFbDarHCQ9LVF66GT+G6DWl94zh8OhYuv1ni8puZSZcJKyKDH5gUBADXa+EGzbZmJighMnTpBKpUin00xPTysxpI8EkH0CbX1nEm4iJZwwX/aYzWZVIAqcDFKpVCrqPMioBHkdh8NBPB6no6ODqakpjh07xsTEBOVy+YKPzWAwGAwGg8FwcbksRBrMX4R/5zvf4ZprrmHTpk28+93v5tVXX1U9S4K4P3JBezGEWrlc5tVXX6Wnp4dkMqlS9i41arUaIyMjjIyM4HQ62bRpE5FIhL6+PjWwWlIBJb5fXJrJyUmKxaIK8ZAyRxE6IhTgZBy+JEDu3r1b9X/19fXhcDjo6OhQiYxut1s9V3q0Dh48yNTUFNPT0ySTSRqNhpovJiEi8lgpeZTSP3mtYrGIx+NRCZbiQEnEvYiiZrPJ1NSUCuzo7e1VYkoPTNEHYOv9ZDLL7ciRIxw8eJBsNqsEpD7YevFAbIfDoQZuy88B9XgRkvJ5k9eT8kaZ8dZqtVRipuzR4/EQi8Wo1WoMDw9z+PDhZfqUGQwGg8FgMBjOxWVR7riYhx56iA9+8IM888wzfO5znzul1wegq6uLcDjMsWPHLooD5nQ6+eQnP8kjjzxyynDtS5Xe3l7e+MY3cttttxGLxcjlcqRSKaanp5VYEyEAqHlq0WiUcDhMV1eXipOvVqsqXVAEsoghEU4Oh0P1i+UXomnj8Ti9vb10dXUBJ53L8fFx5Zhls1k8Hg/FYlH1bjmdToLBoCqtdDqdRCIRFUcvSY62bas0S9u2CYfDKmRExJDP51Pfi4MYDocJh8NqKHgoFFLCy+Vykc/nqVarqkesUqmodEkRhVLaKAErtVpNpVKKs1YulwmHwyrgpF6vUyqVGB0dJZ1OK8Gsl5k2m03GxsaUSHS5XESjUZrNJoFAgI6ODjo7OykWi2zbtu20vx9XAqbc0WAwLMaUOxoMhuXksi93XMzXv/51XnrpJb75zW+SzWb53ve+x/DwcNtjZmdnqVarvOMd7+BnP/vZ6w78sCyLTZs24fP5XtfrrCamp6f56U9/yjPPPMN//a//le7ubhKJBL29vUxPT5PJZEin02r4sggo6ZXq6+tTYtjv96tesVKppBIFJejD6/WquHmPx6PSJ1OpFCMjI1SrVeUwibCSfjOJ99dTERuNBvF4HLfbjc/nIxgMKmEk7pw4ThKSIcO5xUGT+/x+v3KrRAxms9m2cA2Hw9FWKgm0Rf2XSiVisZj6mbh9Iu7q9TqNRkOlMoqTJqWi4oDBfE9gsVjEsixCoZByA0XgyuN9Pp8SqDK4u6enB6/Xy9zcHE888UTb2AKDwWAwGAwGw+rgshRp9Xqdo0eP8kd/9Ef84R/+IalUCqfT2VbSZds2pVKJPXv2cMcdd7B//35mZ2df85qtVot9+/YRi8Xo6uoimUxejENZUaTUr9ls8oMf/IB4PM7g4CC333478Xgcn8+nhk+XSiVmZmZUSaGUQGYyGeLxOF1dXXR0dOD1ettcKhmULNHwEn8vTpgIrEKhoBypQqGgZoaJgySumMw9c7lcqpxRAjT8fj8ej4dAIKBK/qQMU/q04OSAbb3PDVDOndvtVj1u8lhx4uTYW62Wiu+X16pWqzgcDlUOKc8TQQao15fzKOWLgBJ7Mt9M0jabzSbZbLZtsHUsFsPn86no/nA4TH9/PxMTE0xPTzM9PW0EmsFgMBgMBsMq5bIUaTDvrvz85z/ntttuY3BwkLe85S2kUqk2IdZoNJiZmWH9+vV0dnbicDiYmZl5TevZts2OHTtYs2YNjUbjshBpOmNjY8zOzpLNZvH5fPT19RGNRonFYng8HgqFgpq9VavVlEMl7o6UPIp40PsCBRE5kpApIkSEYLFYpFQq0Wg01FgAicoXx0l6wEQU6UOcAeWmicADlBASUSZCB1DiT1yyer1OIBBQryOiS8SQOHFyHOJiyR70MBRx2RaLQ33Qt3wtTmWxWGwTiuIGShmn7EdcN0lyjEajFItFTpw4wcTEBJlM5uJ+QAwGg8FgMBgMF41zDTK/5Pnc5z5Hs9lk69at3HDDDacIA4AdO3YQj8fZuHHjay5XtG2bF198kU2bNnH11Ve/3m2vSsrlMiMjIzz22GPs2LGDZDKpUgL7+/tZs2aNKokMhUKqJ6pQKHDixAmOHTum3DXpYxM3yu124/V6Vc+VuGlut5tQKERHRwc9PT1qDRE7lmXh9Xrx+/1t881EpDQaDVUKWC6XyWazZDIZdatWq6rcUJIQq9WqKhvM5/MUi0W8Xm/b4Gc46bSJKBIXT+6T3jO5Tw8zkf41EbAyAsDr9SrBqIeQlMtlisWiEqpy/OVyWZVd6sEiMrja5/OpEJY9e/Zw9OhRI9AMBoPBYDAYVjmXrZOm82d/9mfcdddd/Omf/ikPPfQQExMTp8wze+GFF9i0aRP3338/3/rWt17zWvpF+eWKCFIRpVu3buXGG29k8+bNlEolstksMD97TZy1SqVCMpkklUoRiUTo6upicHBQCQ5xyxbP6BJXy+l04nQ68fv9xGIx3G43yWRSiZvOzk7V81Wv11XwSKVSUULwxIkTqjdL79eS2WnVapW+vj7cbjeBQEDNULMsi3K5jMfjwev1kkqliEajqiTRtm0V5S8pkuKCyfcOh4NCoUBnZ2eb0yaln3p5o8vlUufMsiw19kDCWsRNlIh9KassFovKmXM6nfT39xOPx8lkMnzzm99ckc+KwWAwGAwGg+HCuSzTHU+H3++nv7+fz3/+83zsYx9jdHT0lMdICt5dd93Fj370o7ZgiPNl/fr1ajDzD3/4w4ux9VWNy+XC5/Ph9Xr5tV/7NeVySZnj1NQU+XyeSqWiSvVg3kETl6ejo4NEIqGGKsvrNZtNCoWCctlEsPn9fqanp1VSopQRSq+WzGEDVBR9vV4nm80qJ0zKECuViipfLJVKas6biLRaraYEoMTYl0olIpGIEuNer1e9FtCWXikjAbxeL5lMRo0VEOdNH0cAJ2P4Yb60MpVKKedRBHChUFCDw1utlhrG3Wg08Pv9dHV1kUgkKBQKHDx4kGPHjlEsFpf1c3EpYNIdDQbDYky6o8FgWE6uuHTH01Eulzl+/DiPPPIIb3/729mzZw979+5te0yj0SCTyXDgwAHe+ta3Mjw8zNGjRy9onenpadauXcuNN954RYi0RqNBoVCgUCiwfft2VZq4adMmurq6aLVahEIhisUiU1NTqtSvVqspESz9a5lMBpfLRSKRaBvMrJcQSt+Vw+EgHA4DKDcpn88rZ0rvJZPSxXA4TLFYbOtbE/fK4/GoUkRx4/S9yrByn8+n3EG9ZwxocwBlj3owiDhr4gyK+wanhpTI17J2sVikXC7TaDRU4AnMC0Kfz0elUiEUChGJROjo6CCdTnPkyBFOnDhhBJrBYDAYDAbDJcYVI9Jg3pl45JFH+MM//EM2bdrE3NwcExMTbY9pNpvs37+f66+/Xl0gT01Nnfca5XKZVqtFJBJRcfOrwa1cDg4ePAjMzzYTcRUIBNQ8MSnvk4TGQqFAqVRSt2w2q3rZarUaHo+n7dyJcBMxJcmOLpcLv9+v5puJWyWJjiLGvF6vmtsmQk13sUKhEM1mE6/Xq1ITRajpsfgSSiJCC06Gnugli4K+ljxOT5cUYSY/k/tqtZoSZ4VCQQkzmScnM9DkeIPBIB6Ph3w+z7FjxxgaGrpkB6sbDAaDwWAwXMlcMeWOi/mVX/kVbrjhBj7/+c+f8TH33nsvGzZs4Ktf/Wqbw3EutmzZws0338zw8DC7d+++oi+Ub7vtNrZs2UJfX58KEZmdnWVubo6ZmRk8Hg+ACtkQpyoSiRCLxUgkErhcLtasWaNEW71eV6WDbrcb27ZVTL/X68Xn86kyQElQrFQqyqnTEyGlXFLvG/N6vSpApFQqEQ6HVXCIjBAQQSYlm06nk0wmowJBpDxTSh4lzESGT3u9XpxOpyqRFFEGqPCUQqGgyjrL5TIOh0PNOGs2m0qQytgHy7IYHh7m+eefX7H3+1LClDsaDIbFmHJHg8GwnJyt3PGcIs2yrEHg60APYAN/a9v2/2NZ1v8AfheQrPlP27b9g4Xn/AnwENAE/si27R+fY41lF2nSu+RwOFT/0ukes3btWh566CE++9nPnrfYsiyLzs5O/vzP/5yPf/zjpFKpi7n1SwqJhA8Gg7z97W/n+uuvV7PECoUCqVSKSqVCLpcjm83icDhIpVLKGQqHw7hcLjZu3EggEMDj8RAMBtXri5ulx86LOyaCCOYd0lQqpWLx3W434XCYarWqovrdbjfVapWenh4l0CQl0e12q+e53W5KpRJOpxOPx6MGSpfLZVW+KJ8rKZt0uVyqVNHtdqvjyGazqp+sWCySz+dJp9NqJlw+nycej5PL5VTqZD6fJxwOK3G2du1a9u7dy759+5ibm7ugPyhcyRiRZjBcWizH9YgRaQaDYTl5vT1pDeATtm3/wrKsMLDLsqyfLPzsL2zb/oL+YMuyrgUeAK4D+oCnLMt6g23bqyryUJwQubA/02MmJyf5xje+we/+7u/y+OOPnzZwZDFy0X0llTqeCREMhUKBF154gVdeeYW+vj76+voYGBggFApRq9VUJL6IIgkJkcAMmA9/ERcsFAqpYdqhUEgJMykzlL4vEYkAgUCgrUzR7XarkkhxsgKBgHKsnE4ngUCAVCqlXtvlcqnnyvGVSiUlFPVB2DJ0W54DJ3vsarWaeq6UNBaLRVKplDp++exkMhkVQOJ0OlmzZg2JRAKPx0OlUuF73/seuVyOfD5vBJrBYLicuSyvRwwGg+F0nFOk2bY9CUwufJ23LOsA0H+Wp7wP+LZt21XgmGVZQ8CtwPaLsN+LzrlEVLVa5ciRI2zdupXrrrsOv9/PoUOHzvm61WqVF154gS1btjA8PHzZDbe+UFqtFnNzc8zNzVEsFsnlcuRyOYLBIN3d3So9sVKpUK/X1cwxh8NBJpNRPVlSohiLxdR8NHHb9KRJ6QETsQW0zVZzOp0qvl9eUxIn5bnS7yZJkLqgl/sANc5B70mDk58tfTi2uLYiCsVFlBlt2WxWHafE8jebTYLBoCq19Hg8VKtVstksqVSK0dHRK/6PAQaD4fLncr8eMRgMBp0LCg6xLGs9cBPwEnA78AeWZX0A2Mn8X7fSzP8H80XtaeOc5j+ilmV9CPjQa9v28mLbNo8//jj/+T//Z3p6epicnCSXy531OeVymccee4yPfvSjtFqtK16k6SSTSZLJJPv27aOjo4Pbb7+dnp4egsEgkUgEQImWWq1GLpdTYR6SwAkoQSSBGX6/X81bczqduFwu5bi5XC6azWbbIOparYbL5VIulfScSaS/rAcnQ0skKVLKI+V1APWviEFdyEnwiKRLiiBrNptks1lqtRper1cFgoiQFLEYjUYJh8N4PB6KxSKHDx9mamrqnJ9Dg8FguBxZqusRH4Gl3bjBYDCcJ+ct0izLCgGPAR+zbTtnWdZfAZ9lvi78s8D/An77fF/Ptu2/Bf524bVXvQ2QTqf5zne+w6233sqnP/1pPvWpT5318bZtUy6XGRwcVMLDcCrpdJonnngCmC9HvO2227jqqqvo6+tTwioej5PP55XzpIunVqtFOp2mVqspcSaCq9FoEIvF8Pv9Sqi5XC7lSNm2TTgcplwuY1mWSneUSH/LsshmsypJUtaVvrBIJKLKGCWtUgZbw/xoAZnlVi6XlYsm4rNer6uEy0ajofYnQ7vD4TDRaBSfz4ff7yefz7N3795TRkcYDAbDlcRSXo9ErPiqvx4xGAxXBucl0izLcjP/H8Rv2bb9rwC2bU9rP/8q8MTCtxPAoPb0gYX7LnmSySQvvfQSc3NzfPSjH+VrX/vaOZ2MRx99lEgkwh133MFzzz23TDu9NCmXyzz33HO8+OKLOBwOent7efOb30xfX5/q5apUKnR0dKgZZNKzJuKpXq9TrVapVCqqdFIEV7VaVYmQIqTi8bgST4FAgGAwSK1WU6WOMP++S4KjRPSnUimVRlkqlUin02qAt5Q3yvBuEWQi4GR/Pp9PuYTi+EWjUbUPGVfw3HPPKZEqpZUGg8FwJWKuRwwGw5XCOUWaNV+v9ffAAdu2v6jdv2ahPhzg/cC+ha+3Af9sWdYXmW/U3QzsuKi7XiFarRaZTIZDhw5Rq9V4y1vewuHDhxkbGzvjc/bv388999xDIpEwIu0c2LatIukBNadMgkG6u7vp7u4mHA63DbKWYA0pH5TQFgkakV4x3bESV0ycLsuy2uL7JYjE4XCQTCYpFou4XC7Vm5ZOp8nlcvh8PjWYW6L8pTSyUCioGWmyrgR7SOCJz+cjGAwqtw+gWCyqktparUYymVRBIwaDwXClYq5HDAbDlcT5OGm3A78FvGJZ1u6F+z4N/IZlWW9ivrxgBPgwgG3b+y3L+g7wKvNJTP/tckpSqtVqKgDjPe95D319fVSrVWZmZk77+GQySb1eJ5FI4PV6zxj3bziVYrHI0NCQ+n5gYID169cTDodpNptK1ESjUdWDJi6WxPlLQIjE8ctAaUlYLJVK6j2pVCqUy+W2PViWpRIX5fky761UKuH1egGYnZ0FULPLZOC0CEIZuA2ogJNwONw2yLpcLlMqlUilUpw4ccL0MRoMBkM75nrEYDBcMVyxw6wvFnfddRe9vb08/PDDZ3zMLbfcQldXF6+++irHjx9fxt1d3sTjcQYHB7n22mvp7OxUaY2WZVEoFFQ6ogyHFmcKUOEdlmWpNEkZTi0/F0dtdnYWv9+P0+lUokpcM5fLRSAQYGZmRs12CwaDOBwOSqUSHo8Hn8+Hx+NRZZp6GMj09DQjIyOMj48zPj6+YufySsTMSTMYDIsxc9IMBsNy8rqGWS8HlmUlgSIwu4Lb6DTrm/XN+lfM+uts2+5axvUMBsMlgLkeMeub9c36y7z+Ga9HVoVIA7Asa6dt27eY9c36Zn2zvsFgMKwUK/3fI7O+Wd+sf+Wur+NY6Q0YDAaDwWAwGAwGg+EkRqQZDAaDwWAwGAwGwypiNYm0vzXrm/XN+mZ9g8FgWGFW+r9HZn2zvln/yl1fsWp60gwGg8FgMBgMBoPBsLqcNIPBYDAYDAaDwWC44llxkWZZ1i9blnXIsqwhy7I+tQzrDVqW9bRlWa9alrXfsqyPLtz/PyzLmrAsa/fC7VeXcA8jlmW9srDOzoX74pZl/cSyrCML/3Ys0dpbtGPcbVlWzrKsjy3l8VuW9Q+WZc1YlrVPu++0x2vN8+WFz8Ney7JuXqL1/9yyrIMLa3zXsqzYwv3rLcsqa+fhr5do/TOeb8uy/mTh+A9ZlvWuJVr/EW3tERkMu0THf6bfuWX7DBgMBsO5MNcj5nrEXI+Y65FVdT1i2/aK3QAnMAxsBDzAHuDaJV5zDXDzwtdh4DBwLfA/gE8u03GPAJ2L7vs88KmFrz8F/M9lOv9TwLqlPH7gbcDNwL5zHS/wq8APAQvYCry0ROvfC7gWvv6f2vrr9cct4fGf9nwvfBb3AF5gw8Lvh/Nir7/o5/8L+D+X8PjP9Du3bJ8BczM3czO3s93M9UjbfeZ6xDbXI+Z6ZOWvR1baSbsVGLJt+6ht2zXg28D7lnJB27Ynbdv+xcLXeeAA0L+Ua54n7wP+aeHrfwLuW4Y17wGGbds+vpSL2Lb9cyC16O4zHe/7gK/b87wIxCzLWnOx17dt+0nbthsL374IDLyeNS50/bPwPuDbtm1Xbds+Bgwx/3uyJOtblmUBvw48/HrWOMf6Z/qdW7bPgMFgMJwDcz1yEnM9cvJ+cz1irkdW7HpkpUVaPzCmfT/OMv4HyrKs9cBNwEsLd/3Bgp35D0tl7y9gA09alrXLsqwPLdzXY9v25MLXU0DPEq4vPED7L8NyHT+c+XhX4jPx28z/pUTYYFnWy5Zl/btlWXcu4bqnO9/Lffx3AtO2bR/R7luy41/0O7eaPgMGg+HKxlyPmOsRWPn/F5nrEXM9olhpkbZiWJYVAh4DPmbbdg74K2AT8CZgknnLdam4w7btm4FfAf6bZVlv039oz3usSxq7aVmWB/g14NGFu5bz+NtYjuM9E5ZlfQZoAN9auGsSWGvb9k3Ax4F/tiwrsgRLr9j5XsRv0P4/xiU7/tP8zilW8jNgMBgMK4m5HjHXI2CuRzDXI6ew0iJtAhjUvh9YuG9JsSzLzfyb8y3btv8VwLbtadu2m7Ztt4Cv8jot3bNh2/bEwr8zwHcX1poWC3Xh35mlWn+BXwF+Ydv29MJelu34FzjT8S7bZ8KyrP8CvAd4cOGXkgVbf27h613M12C/4WKvfZbzvZzH7wL+d+ARbV9Lcvyn+51jFXwGDAaDYQFzPWKuR8z1iLkeWVXXIyst0v4D2GxZ1oaFv6Q8AGxbygUXal7/Hjhg2/YXtfv1GtP3A/sWP/cirR+0LCssXzPfMLqP+eP+4MLDPgj821Ksr9H2F4vlOn6NMx3vNuADC4k6W4GsZkFfNCzL+mXgj4Ffs227pN3fZVmWc+HrjcBm4OgSrH+m870NeMCyLK9lWRsW1t9xsddf4B3AQdu2x7V9XfTjP9PvHCv8GTAYDAYNcz1irkfAXI+Y65HVdD1iL2NKyeluzCenHGZeIX9mGda7g3kbcy+we+H2q8A3gFcW7t8GrFmi9Tcyn5azB9gvxwwkgJ8CR4CngPgSnoMgMAdEtfuW7PiZ/4/vJFBnvp73oTMdL/MJOn+58Hl4BbhlidYfYr7OWD4Df73w2PsX3pfdwC+A9y7R+mc838BnFo7/EPArS7H+wv3/CHxk0WOX4vjP9Du3bJ8BczM3czO3c90w1yPmesRcj5jrkVV0PWItbMJgMBgMBoPBYDAYDKuAlS53NBgMBoPBYDAYDAaDhhFpBoPBYDAYDAaDwbCKMCLNYDAYDAaDwWAwGFYRRqQZDAaDwWAwGAwGwyrCiDSDwWAwGAwGg8FgWEUYkWYwGAwGg8FgMBgMqwgj0gwGg8FgMBgMBoNhFWFEmsFgMBgMBoPBYDCsIv5/QdbivOekeqgAAAAASUVORK5CYII=\n",
+      "image/png": "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\n",
       "text/plain": [
        "<Figure size 1296x432 with 2 Axes>"
       ]
@@ -544,7 +490,7 @@
    "source": [
     "### Create Swin UNETR model\n",
     "\n",
-    "In this scetion, we create Swin UNETR model for the 14-class multi-organ segmentation. We use a feature size of 48 which is compatible with self-supervised pre-trained weights. We also use gradient checkpointing (use_checkpoint) for more memory-efficient training. "
+    "In this section, we create Swin UNETR model for the 14-class multi-organ segmentation. We use a feature size of 48 which is compatible with self-supervised pre-trained weights. We also use gradient checkpointing (use_checkpoint) for more memory-efficient training. "
    ]
   },
   {
@@ -554,7 +500,7 @@
    "outputs": [],
    "source": [
     "os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\"\n",
-    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
+    "device = torch.device(\"cuda:1\" if torch.cuda.is_available() else \"cpu\")\n",
     "\n",
     "model = SwinUNETR(\n",
     "    img_size=(96, 96, 96),\n",
@@ -571,7 +517,19 @@
    "source": [
     "### Initialize Swin UNETR encoder from self-supervised pre-trained weights\n",
     "\n",
-    "In this section, we intialize the Swin UNETR encoder from weights downloaded from this [link](https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/model_swinvit.pt). If training from scratch is desired, please skip this section."
+    "In this section, we intialize the Swin UNETR encoder from pre-trained weights. The weights can be downloaded using the `wget` command below, or by clicking on [this link](https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/model_swinvit.pt). If training from scratch is desired, please skip this section."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "scrolled": false
+   },
+   "outputs": [],
+   "source": [
+    "# uncomment to download the pre-trained weights\n",
+    "# !wget https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/model_swinvit.pt"
    ]
   },
   {
@@ -620,17 +578,60 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 15,
    "metadata": {
     "scrolled": true
    },
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (X / X Steps) (loss=X.X):   0%|                                                          | 0/24 [00:00<?, ?it/s]Traceback (most recent call last):\n",
+      "  File \"/opt/conda/lib/python3.8/multiprocessing/queues.py\", line 245, in _feed\n",
+      "    send_bytes(obj)\n",
+      "  File \"/opt/conda/lib/python3.8/multiprocessing/connection.py\", line 200, in send_bytes\n",
+      "    self._send_bytes(m[offset:offset + size])\n",
+      "  File \"/opt/conda/lib/python3.8/multiprocessing/connection.py\", line 411, in _send_bytes\n",
+      "    self._send(header + buf)\n",
+      "  File \"/opt/conda/lib/python3.8/multiprocessing/connection.py\", line 368, in _send\n",
+      "    n = write(self._handle, buf)\n",
+      "BrokenPipeError: [Errno 32] Broken pipe\n",
+      "Training (X / X Steps) (loss=X.X):   0%|                                                          | 0/24 [00:01<?, ?it/s]\n"
+     ]
+    },
+    {
+     "ename": "RuntimeError",
+     "evalue": "CUDA out of memory. Tried to allocate 1.80 GiB (GPU 1; 15.75 GiB total capacity; 12.70 GiB already allocated; 1.45 GiB free; 13.13 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
+      "Input \u001b[0;32mIn [15]\u001b[0m, in \u001b[0;36m<cell line: 88>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     87\u001b[0m metric_values \u001b[38;5;241m=\u001b[39m []\n\u001b[1;32m     88\u001b[0m \u001b[38;5;28;01mwhile\u001b[39;00m global_step \u001b[38;5;241m<\u001b[39m max_iterations:\n\u001b[0;32m---> 89\u001b[0m     global_step, dice_val_best, global_step_best \u001b[38;5;241m=\u001b[39m \u001b[43mtrain\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m     90\u001b[0m \u001b[43m        \u001b[49m\u001b[43mglobal_step\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtrain_loader\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdice_val_best\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mglobal_step_best\u001b[49m\n\u001b[1;32m     91\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     92\u001b[0m model\u001b[38;5;241m.\u001b[39mload_state_dict(torch\u001b[38;5;241m.\u001b[39mload(os\u001b[38;5;241m.\u001b[39mpath\u001b[38;5;241m.\u001b[39mjoin(root_dir, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mbest_metric_model.pth\u001b[39m\u001b[38;5;124m\"\u001b[39m)))\n",
+      "Input \u001b[0;32mIn [15]\u001b[0m, in \u001b[0;36mtrain\u001b[0;34m(global_step, train_loader, dice_val_best, global_step_best)\u001b[0m\n\u001b[1;32m     34\u001b[0m step \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m \u001b[38;5;241m1\u001b[39m\n\u001b[1;32m     35\u001b[0m x, y \u001b[38;5;241m=\u001b[39m (batch[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mimage\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mcuda(device\u001b[38;5;241m=\u001b[39mdevice), batch[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mlabel\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mcuda(device\u001b[38;5;241m=\u001b[39mdevice))\n\u001b[0;32m---> 36\u001b[0m logit_map \u001b[38;5;241m=\u001b[39m \u001b[43mmodel\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     37\u001b[0m loss \u001b[38;5;241m=\u001b[39m loss_function(logit_map, y)\n\u001b[1;32m     38\u001b[0m loss\u001b[38;5;241m.\u001b[39mbackward()\n",
+      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py:1111\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1107\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1108\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1110\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1111\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1112\u001b[0m \u001b[38;5;66;03m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1113\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[38;5;241m=\u001b[39m [], []\n",
+      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:280\u001b[0m, in \u001b[0;36mSwinUNETR.forward\u001b[0;34m(self, x_in)\u001b[0m\n\u001b[1;32m    279\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mforward\u001b[39m(\u001b[38;5;28mself\u001b[39m, x_in):\n\u001b[0;32m--> 280\u001b[0m     hidden_states_out \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mswinViT\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx_in\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mnormalize\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    281\u001b[0m     enc0 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mencoder1(x_in)\n\u001b[1;32m    282\u001b[0m     enc1 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mencoder2(hidden_states_out[\u001b[38;5;241m0\u001b[39m])\n",
+      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py:1111\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1107\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1108\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1110\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1111\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1112\u001b[0m \u001b[38;5;66;03m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1113\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[38;5;241m=\u001b[39m [], []\n",
+      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:971\u001b[0m, in \u001b[0;36mSwinTransformer.forward\u001b[0;34m(self, x, normalize)\u001b[0m\n\u001b[1;32m    969\u001b[0m x0 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mpos_drop(x0)\n\u001b[1;32m    970\u001b[0m x0_out \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mproj_out(x0, normalize)\n\u001b[0;32m--> 971\u001b[0m x1 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlayers1\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx0\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcontiguous\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    972\u001b[0m x1_out \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mproj_out(x1, normalize)\n\u001b[1;32m    973\u001b[0m x2 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mlayers2[\u001b[38;5;241m0\u001b[39m](x1\u001b[38;5;241m.\u001b[39mcontiguous())\n",
+      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py:1111\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1107\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1108\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1110\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1111\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1112\u001b[0m \u001b[38;5;66;03m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1113\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[38;5;241m=\u001b[39m [], []\n",
+      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:841\u001b[0m, in \u001b[0;36mBasicLayer.forward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m    839\u001b[0m attn_mask \u001b[38;5;241m=\u001b[39m compute_mask([dp, hp, wp], window_size, shift_size, x\u001b[38;5;241m.\u001b[39mdevice)\n\u001b[1;32m    840\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m blk \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mblocks:\n\u001b[0;32m--> 841\u001b[0m     x \u001b[38;5;241m=\u001b[39m \u001b[43mblk\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mattn_mask\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    842\u001b[0m x \u001b[38;5;241m=\u001b[39m x\u001b[38;5;241m.\u001b[39mview(b, d, h, w, \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m)\n\u001b[1;32m    843\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdownsample \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
+      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py:1111\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1107\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1108\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1110\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1111\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1112\u001b[0m \u001b[38;5;66;03m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1113\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[38;5;241m=\u001b[39m [], []\n",
+      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:651\u001b[0m, in \u001b[0;36mSwinTransformerBlock.forward\u001b[0;34m(self, x, mask_matrix)\u001b[0m\n\u001b[1;32m    649\u001b[0m shortcut \u001b[38;5;241m=\u001b[39m x\n\u001b[1;32m    650\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39muse_checkpoint:\n\u001b[0;32m--> 651\u001b[0m     x \u001b[38;5;241m=\u001b[39m \u001b[43mcheckpoint\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcheckpoint\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mforward_part1\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mx\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mmask_matrix\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    652\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    653\u001b[0m     x \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mforward_part1(x, mask_matrix)\n",
+      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/utils/checkpoint.py:235\u001b[0m, in \u001b[0;36mcheckpoint\u001b[0;34m(function, use_reentrant, *args, **kwargs)\u001b[0m\n\u001b[1;32m    232\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mUnexpected keyword arguments: \u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;241m+\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m,\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;241m.\u001b[39mjoin(arg \u001b[38;5;28;01mfor\u001b[39;00m arg \u001b[38;5;129;01min\u001b[39;00m kwargs))\n\u001b[1;32m    234\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m use_reentrant:\n\u001b[0;32m--> 235\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mCheckpointFunction\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mapply\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfunction\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mpreserve\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    236\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    237\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m _checkpoint_without_reentrant(\n\u001b[1;32m    238\u001b[0m         function,\n\u001b[1;32m    239\u001b[0m         preserve,\n\u001b[1;32m    240\u001b[0m         \u001b[38;5;241m*\u001b[39margs\n\u001b[1;32m    241\u001b[0m     )\n",
+      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/utils/checkpoint.py:96\u001b[0m, in \u001b[0;36mCheckpointFunction.forward\u001b[0;34m(ctx, run_function, preserve_rng_state, *args)\u001b[0m\n\u001b[1;32m     93\u001b[0m ctx\u001b[38;5;241m.\u001b[39msave_for_backward(\u001b[38;5;241m*\u001b[39mtensor_inputs)\n\u001b[1;32m     95\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m torch\u001b[38;5;241m.\u001b[39mno_grad():\n\u001b[0;32m---> 96\u001b[0m     outputs \u001b[38;5;241m=\u001b[39m \u001b[43mrun_function\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     97\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m outputs\n",
+      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:591\u001b[0m, in \u001b[0;36mSwinTransformerBlock.forward_part1\u001b[0;34m(self, x, mask_matrix)\u001b[0m\n\u001b[1;32m    589\u001b[0m     attn_mask \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[1;32m    590\u001b[0m x_windows \u001b[38;5;241m=\u001b[39m window_partition(shifted_x, window_size)\n\u001b[0;32m--> 591\u001b[0m attn_windows \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mattn\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx_windows\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mmask\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mattn_mask\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    592\u001b[0m attn_windows \u001b[38;5;241m=\u001b[39m attn_windows\u001b[38;5;241m.\u001b[39mview(\u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m, \u001b[38;5;241m*\u001b[39m(window_size \u001b[38;5;241m+\u001b[39m (c,)))\n\u001b[1;32m    593\u001b[0m shifted_x \u001b[38;5;241m=\u001b[39m window_reverse(attn_windows, window_size, dims)\n",
+      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py:1111\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1107\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1108\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1110\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1111\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1112\u001b[0m \u001b[38;5;66;03m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1113\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[38;5;241m=\u001b[39m [], []\n",
+      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:475\u001b[0m, in \u001b[0;36mWindowAttention.forward\u001b[0;34m(self, x, mask)\u001b[0m\n\u001b[1;32m    473\u001b[0m q, k, v \u001b[38;5;241m=\u001b[39m qkv[\u001b[38;5;241m0\u001b[39m], qkv[\u001b[38;5;241m1\u001b[39m], qkv[\u001b[38;5;241m2\u001b[39m]\n\u001b[1;32m    474\u001b[0m q \u001b[38;5;241m=\u001b[39m q \u001b[38;5;241m*\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscale\n\u001b[0;32m--> 475\u001b[0m attn \u001b[38;5;241m=\u001b[39m \u001b[43mq\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m@\u001b[39;49m\u001b[43m \u001b[49m\u001b[43mk\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtranspose\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m-\u001b[39;49m\u001b[38;5;241;43m2\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m-\u001b[39;49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[1;32m    476\u001b[0m relative_position_bias \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mrelative_position_bias_table[\n\u001b[1;32m    477\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mrelative_position_index\u001b[38;5;241m.\u001b[39mclone()[:n, :n]\u001b[38;5;241m.\u001b[39mreshape(\u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m)\n\u001b[1;32m    478\u001b[0m ]\u001b[38;5;241m.\u001b[39mreshape(n, n, \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m)\n\u001b[1;32m    479\u001b[0m relative_position_bias \u001b[38;5;241m=\u001b[39m relative_position_bias\u001b[38;5;241m.\u001b[39mpermute(\u001b[38;5;241m2\u001b[39m, \u001b[38;5;241m0\u001b[39m, \u001b[38;5;241m1\u001b[39m)\u001b[38;5;241m.\u001b[39mcontiguous()\n",
+      "\u001b[0;31mRuntimeError\u001b[0m: CUDA out of memory. Tried to allocate 1.80 GiB (GPU 1; 15.75 GiB total capacity; 12.70 GiB already allocated; 1.45 GiB free; 13.13 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF"
+     ]
+    }
+   ],
    "source": [
     "def validation(epoch_iterator_val):\n",
     "    model.eval()\n",
     "    with torch.no_grad():\n",
     "        for step, batch in enumerate(epoch_iterator_val):\n",
-    "            val_inputs, val_labels = (batch[\"image\"].cuda(), batch[\"label\"].cuda())\n",
+    "            val_inputs, val_labels = (batch[\"image\"].cuda(device=device), batch[\"label\"].cuda(device=device))\n",
     "            val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)\n",
     "            val_labels_list = decollate_batch(val_labels)\n",
     "            val_labels_convert = [\n",
@@ -656,9 +657,11 @@
     "    epoch_iterator = tqdm(\n",
     "        train_loader, desc=\"Training (X / X Steps) (loss=X.X)\", dynamic_ncols=True\n",
     "    )\n",
+    "\n",
+    "    # TODO: use AMP\n",
     "    for step, batch in enumerate(epoch_iterator):\n",
     "        step += 1\n",
-    "        x, y = (batch[\"image\"].cuda(), batch[\"label\"].cuda())\n",
+    "        x, y = (batch[\"image\"].cuda(device=device), batch[\"label\"].cuda(device=device))\n",
     "        logit_map = model(x)\n",
     "        loss = loss_function(logit_map, y)\n",
     "        loss.backward()\n",
@@ -676,6 +679,7 @@
     "                val_loader, desc=\"Validate (X / X Steps) (dice=X.X)\", dynamic_ncols=True\n",
     "            )\n",
     "            dice_val = validation(epoch_iterator_val)\n",
+    "            # FIXME: epoch_loss is a running average at time of validation??\n",
     "            epoch_loss /= step\n",
     "            epoch_loss_values.append(epoch_loss)\n",
     "            metric_values.append(dice_val)\n",
@@ -699,7 +703,7 @@
     "        global_step += 1\n",
     "    return global_step, dice_val_best, global_step_best\n",
     "\n",
-    "\n",
+    "torch.cuda.empty_cache()\n",
     "max_iterations = 30000\n",
     "eval_num = 500\n",
     "post_label = AsDiscrete(to_onehot=14)\n",
@@ -751,7 +755,7 @@
    "outputs": [
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAskAAAGDCAYAAADZHo16AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAABsKUlEQVR4nO3dd3xc5ZX/8c/RqHeruEnuvWIbYzA9tNCCSYcUQhqbQjbJJrsh2fRssimb/NJIIY0UCAGSECAkkBA6NrgA7sa23JuaJY3aSKN5fn/MHXksq1plir7v10svZu69M/fMyFydOXOe5zHnHCIiIiIickJKrAMQEREREYk3SpJFRERERLpQkiwiIiIi0oWSZBERERGRLpQki4iIiIh0oSRZRERERKQLJcky5Mys0cymxzoOEREZHDNzZjYz1nEMJTPbYmYXxzoOiX9KkpOMme01s8u82zeb2bPDfL4nzex90ducc7nOuYphPudxM8sYrnOMFDP7opn9LtZxiEhyMrO/m9mXu9m+ysyOmlnqIJ77SS+JPqPL9j972y8+3ec+zXimeudt9H6OmdnDZnZ59HHOuQXOuSdHMjZJTEqSpUeDuXgOFzObClwAOOC6YXj+uHvNIiKD8GvgHWZmXba/E7jLORcc5PO/CtwUuWNmxcBKoGqQzzsYhc65XOAM4B/An83s5hjGIwlKSXKSMrN5wE+Ald4n6jpve4aZ/Z+Z7fc+Zf/EzLK8fReb2UEz+5SZHQV+ZWZjvE/iVV719mEzK/eO/yrhhPWH3jl+6G3v/HrOzArM7Dfe4/eZ2WfNLMXbd7OZPevFc9zM9pjZVX28tJuANcCdwLuiXlOdmS2Mev2lZtZiZmO9+9ea2cvecc+b2eKoY/d6r3kj0GRmqWZ2m5ntNjO/mW01s9dHHe8zs2+bWbUX863ea06Nes2/MLMjZnbIzP7HzHyn8Tu8zvtasM6r2MyL2vcp77n9ZrbDzC71tq8ws3Vm1uD9fr8z0POKSFJ5ACgmfK0GwMzGANcCv/GuGau968wRM/uhmaUP4PnvAt4adY27Efgz0BZ1vpSoa2qNmd1rZkVR++/zqtr1Zva0mS2I2nenmd1uZn/1rncvmNmM/gTmnDvqnPse8EXgG1F/e6K/cfWZ2WeirvfrzWySt2+umf3DzGq96+xbBvC+SBJQkpyknHPbgA8Aq732h0Jv19eB2cASYCZQBnw+6qHjgSJgCnAL4X8jv/LuTwZagB965/hv4BngVu8ct3YTyg+AAmA6cBHhJPfdUfvPBnYAJcA3gV+YnVLxiHYT4YvyXcBrzWyccy4A/InwxTniLcBTzrlKM1sK/BL4N8J/LH4KPGgnt2vcCFxDuAIRBHYT/qNSAHwJ+J2ZTfCOfT9wFeH3cBlwfZcY7wSChN/fpcAVwPsYADObDfwe+BhQCjwCPGRm6WY2B7gVOMs5lwe8FtjrPfR7wPecc/nADODegZxXRJKLc66F8HXgpqjNbwG2O+deATqAjxO+Bq8ELgU+NIBTHAa2Er7O4Z3nN12O+Qjh6+RFwETgOHB71P6/AbOAscAGwtf3aDcQvg6PAXYBXx1AfBD++zAWmNPNvv8gfP2/GsgH3gM0m1kO4Sr03d5jbwB+ZGbzB3huSWTOOf0k0Q/hZOky7/bNwLNR+wxoAmZEbVsJ7PFuX0z4039mL8+/BDgedf9J4H1djnGEE0Sf93zzo/b9G/BkVHy7ovZle48d38O5zwfagRLv/nbg497ty4DdUcc+B9zk3f4x8JUuz7UDuCjqPXtPH+/ry8Aq7/a/gH+L2neZF3cqMA4IAFlR+28Enujheb8I/K6b7Z8D7o26nwIc8n5HM4FK77xpXR73NOE/JiWx/reoH/3oJz5+vGtnXeTa7l0fP97DsR8D/hx13wEzezj2ScIFgHcQ/lA/F3jV23cQuNi7vQ24NOpxE7xreWo3z1nonbPAu38n8POo/VcTTvC7i2dq5FrcZXumt/087/5eTvyd3BG5tnd5zFuBZ7ps+ynwhVj/PvUzcj+qJI8upYQT0fXeV2t1wN+97RFVzrnWyB0zyzazn3qtEg2Ek7DCfrYPlABpwL6obfsIV68jjkZuOOeavZu5PTzfu4DHnHPV3v27vW0ATwDZZna2hfuWlxD+yg/CVfBPRF6z97onEa5oRByIPpGZ3RTVnlEHLPReD97jDvTw2Cneaz4S9difEq5EDMREot4351zIO0+Zc24X4T9kXwQqzeweM4u8lvcS/qZgu5mtNbNrB3heEUkyzrlngWrgeq9VYQXh6ydmNtvCbXRHvWv81zhxreuvPwGXEP6G67fd7J9CuC84ck3cRriCPc5rd/i61+7QwIlvxaJjOBp1u5me/0b0JPI3p7abfZMIf3PYXcxnd/m78XbC37bKKKFBSsnNdblfTbhdYoFz7lA/H/MJwl9Rne2cO2pmS4CXCFeluzu+6/naCV9stnrbJhOuiA6Ihfum3wL4LNwvDZBBOGE/wzn3ipndS7hqewx42Dnn9447AHzVOdfbV3Sdr8PMpgA/I/y142rnXIeZvcyJ13wEKI967KSo2wcIV5JL3OAGxBwGFkXFZN55DgE45+4G7jazfMJJ+DeAdzrndgI3er13bwDuN7Ni51zTIGIRkcT3G8KtEHOAR51zx7ztPyZ8Tb/ROec3s48BbxrIEzvnms3sb8AHCbd5dXWA8Ld1z3XdYWbvBFYR/mZsL+EWt+OcuN4OhdcT/vZtRw+xzQA2d7P9Kefc5ac+REYLVZKT2zGgPDIIw6tG/gz4f3ZiQFuZmb22l+fII5xY13kDLb7QzTm6nRPZOddBuBfuq2aW5yWf/wGczpRn1xOuPMwnXCVeAswj3BMd6bW7m/BXZG/3bkf8DPiAV2U2M8sxs2vMLK+Hc+UQTpqrAMzs3YQryRH3Ah/13rtC4FNRr/kI8BjwbTPL9waszDCzi3p5bSlmlhn1k+Gd4xozu9TM0gh/WAkAz5vZHDO7xDuulfDvJ+TF+g4zK/V+13Xe84d6ObeIjA6/IZyIvp/wjBcReUAD0GhmcwknuqfjM4Rb2PZ2s+8nhP8OTIHOgdWros4fAGoIf9P5tdM8/ynMbJyZ3Ur479anvetiVz8HvmJms7y/D4stPEPHw8BsM3unmaV5P2dZ1ABqSX5KkpPbv4AtwFEzi7QofIrwwIc13ldb/6T7wQwR3wWyCFeF1xBuz4j2PeBNFp6d4vvdPP4jhPugK4BnCSevvzyN1/Iu4FfOuf0uPGL5qHPuKOFBhG83s1Tn3AveuSYSHggCgHNuHeE/DD8kXKHYRbgfulvOua3At4HVhD8ELCLcwxfxM8KJ8EbCFZhHCA/U6/D23wSkE66eHwfuJ9yD15MbCSe6kZ/dzrkdhPv8fkD4vX8d8DrnXBvhCvrXve1HCbdyfNp7riuBLWbWSPh3c4MLD9wRkVHMS16fJ1wEeDBq1yeBtwF+wte2P5zm8x/22jq68z3vnI+ZmZ/w35KzvX2/IdxadojwNXPN6Zy/izozawI2Ee5hfrNzrqe/O98hXJR4jPCHhV8QHlPiJzwY8QbC3+wdJfyNXcLPzy/9Z8719m25iPTFwtPW/cQ5NyXWsYiIiMjQUCVZZIDMLMvMrrbwfMplhL/K+3NfjxMREZHEoUqyyACZWTbwFOHpjlqAvwIfdc41xDQwERERGTJKkkVEREREulC7hYiIiIhIF0qSRURERES6iNliIiUlJW7q1KmxOr2IyKCsX7++2jlX2veRyUHXbBFJZKdzzY5Zkjx16lTWrVsXq9OLiAyKme3r+6jkoWu2iCSy07lmq91CRERERKQLJckiIiIiIl0oSRYRERER6UJJsoiIiIhIF0qSRURERES6UJIsIiIiItKFkmQRERERkS6UJIuIiIiIdKEkWURERESkCyXJIiIiIiJdKEkWEREREemizyTZzH5pZpVmtrmP484ys6CZvWnowjvZpoP1rNtbO1xPLyIiIiJxaldlIy8fqBux8/WnknwncGVvB5iZD/gG8NgQxNSjbz22g6/8ddtwnkJEREREeuGc44ntlXztkW280kPSGuwIcf/6g1z67Sf56D0vEQh2nPb56lva+crDW7nyu0/zpYe2nPbzDFRqXwc45542s6l9HPYR4I/AWUMRVE8KstLYX9M0nKcQERERSTjOOcxs2M+zbm8t3/z7Dl70vtm/4+kKlk4u5OZzp3LVwgmkGDzw8mF+8K+d7KtpZnppDn95+TBV/gB33LSc3Iw+U89OHSHHvesO8H+P7qC2uY0bzprEJ66YM1wv7RT9j7QHZlYGvB54DcOcJBdmpVHf0j6cpxAREZEYa24L8kJFLRfNLiUlZfgTv/4KdoT444aDvHbBeAqz0/v9uPX7jvPlh7awqLyAL123EN8QvKaOkGPjwTqe3VnNM7uqeXl/HWdPL+K2q+ayYGLBKce3tndw3/qDrN1TS/mYLKaV5DCtJIepJTnkZ6bR0tZBU1uQ5rYOWto6MIM0XwppPiPNl0JtUxs/+NdO/rmtktK8DP7n+oVcs2gCf3n5EL9evY+P3vMyX83bRla6j301zSyYmM/PblrOZfPG8ueXDvGf92/kxjvWcOe7z6I4N+OU+ALBDiobAhxtaOVoffjngZcPseVwA2dNHcOvX7eChWWnvq7hNOgkGfgu8CnnXKivTzBmdgtwC8DkyZMHfKICL0kOhVxc/U8jIiIiQ6OlrYObf7WWF/fU8po5pXz3rUspyE6LdVgA/H3LUT71x038ds0+7nrfORRk9R5XQ2s73/r7Dn73wj4KstJ45WA99S1BvvOWM0jznd7cCfXN7Xzj0e08/MphGlqDACyYmM8bzyznb5uPcO0PnuUNS8v55GtnM6EgC39rO3e9sJ+fP7OH6sYA4/IzeGRTG8GQG/C58zJT+c/XzuHd500lOz2cQt583jRuWjmVp3ZW8evn99IUCPLZa+Zz2byxnZXtNywrpzA7jQ/dtYE3/2Q1v3nvCsoKs9h2xM8TOyp5fNsxXj5QR9eQygqz+P6NS3nd4gkjUiXvypzr+03y2i0eds4t7GbfHiASeQnQDNzinHugt+dcvny5W7du3YCC/fkzFfzPX7ex8YtXkJ8ZH//DiMjoZGbrnXPLYx3HSDmda7Ykt2MNrYzNyxhw8vL0q1UU5aR3WxVsbe/g/b9Zx3O7qnnrWZO5f/0BJhRk8dN3nsm8CflDFfpp+8Bv1/Pc7mpa2ztYVFbAb997Njk9tA/8ffNRvvDgZqr8AW4+dxqfuGI2v12zj6//bTuXzx/HD9+2lIxU34DO//fNR/jcX7ZQ29TGqiUTuXjOWM6bUdxZma1vbudHT+7iV8/txQyuXDieJ7ZX0tAa5IJZJXz4NTM5e1oRHSHHweMt7KlpYm91E02BINnpqWSn+8hK93UmwO0dIdo7QrQFQwBcNm8cY3L6X0Hvat3eWt5z51rSU32k+Ywj9a0ALC4v4IJZJUwpymFcQSbj88M/+VmpQ5Ycn841e9CVZOfctKgA7iScTD8w2OftTr73ia2+uV1JsoiISAys31fLtx97led31/DW5ZP48vUL+p3svXKgjpt/9SIA7z5vGv9x+ezOJLMtGOLDd23gmZ3VfOtNi3nz8km8eXk5H/zdel7/o+f4xhsXs2pJ2YBi7Qg5jtS30BYMMb00d2AvtIvGQJAndlRyw1mTWDmjmA/f/RLv/fVafnXzCrLST7z+9fuO891/vsozO6uZNyGfO965nDMmFQLwgYtmkJXm4wsPbuH9v1nPT99x5kmP7UmVP8AXHtzMI5uOMn9CPr+6+axuP2QUZKfx6avn8Y5zpvB/j+3goVcOc/n8cXz4NTNZXF7YeVyqz5jqtVowci2+LJ9axH0fOJf//vMmSnIz+PhlY7l4Tilj8zNHLogB6DNJNrPfAxcDJWZ2EPgCkAbgnPvJsEbXRWEkSW5pZ9JInlhERCSB7K5q5OZfvUi1v43MtBQy03xkpKaQn5XGyunFXDJ3LGdOGUPqAL7y33Swnu/8YwdP7KiiJDedVUsm8od1B9hV1chP3nEmpXmn9plGawuG+NQfNzI2L5PXzB3LL57dw983H+V/rl/IBbNK+NgfXuLx7ZV85fqFvHl5+K/8ssljeOgj53Pr3S/x0Xte5rEtx5gzPi9caSwI/4Sc41hDgGMNrRyrb+VoQysHj7ewv7aZg8ebae8If2P++Wvn857zp/UWYq8e33aMQDDEtWdM5KypRXznLSE+9oeX+bffrednN53JxoP1fO+fO3l2VzVFOel89pp5vOvcqae0Vbzr3Klkpfn41J828q5fvchXVi1k9rjcbiumuyob+fvmI/z82T00Bzr4z9fO4ZYLp/fZqjGpKJvv3bCU7751SUzaFHozZ3we93/w3FiH0S/9md3ixv4+mXPu5kFF04eCqCRZREQkUXR4zZa9DdhyzvHMzmq2H23gcF040Ttc10IwFOJDF89k1ZKJ/Up4ahoDvPtXa2kOdPCOcybT2h6itb2D1mCIKn8rv3xuDz99uoL8zFQunjOWC2eXsri8gOklOSclzc45KqqbeH5XNf/aXskTO6ooyErjU1fO5V3nTiE7PZUr5o/nk/e9wnU/fJafvvPMk6qVXf3oyV1sP+rnF+9azqXzxvGGZWV8+k+bePeda5lekkNFdROfvWYe7zxnykmPG5uXyV3vO5tvPbqD+9cf5K+bjvT6+guz05g0Jpv5E/J57YLxTCnO5p9bj/E/f93KrHG5XDCrtNvH+VvbCTl67DN+6JUjjM/P5MzJYwBYtaSMQHuI//rjRi785hMcawhQkpvOZ66eyzvOmdLZstCdt5w1iYy0FD5x7yu89rtPU5KbwbkzijlvZjHTS3N5+tUq/rb5KLsqGwFYOb2Yr1y/gJlj83p97V3FW4KcaPrVkzwcTqe/bfvRBq787jPc/rZlXLN4wjBFJiLSN/Uky0B8+K4N7Kps5P4PriSvh3bB37+4n0//aRMAuRmplBVmUTYmi6P1rWw90sCKaUV8ZdVC5ozvOVFqbe/gxp+tYevhBu655RyWegldtMZAkGd3VvH4tkqe2FFJdWMbAJlpKcybkM/CiQU0BYI8v7uGow3hntGywizevLyc95w/7ZR2xy2H67nlN+upbgzwjTcu5vqlp7ZE7Djq59ofPMPViybwvRuWdm4PBDv4yZMV/PipXXzkkll8+DUz+3gnT8yCcMSrGqcYjM/PZFx+JqV5GWSmndq+0BQI8sYfP8/huhb+cuv5TCvJOWn/6t013Hr3Bopz03nk3y84pcJe39LOWf/zT965cgqfu3b+SfvuemEfv3hmD28/ZwpvWzG5X+0TEYfrWnh2ZzXP7a7m+d01VPkDAKQYnD2tmCsXjueKBeOYUJDV7+eU7p3ONTuhkuSj9a2c87+P87XXL+JtZw98dgwRkaGiJFn6a+vhBq7+/jMAXLN4Aj+8cekpFb4dR/1c98NnWTGtiB++bdlJ1cyOkOMPaw/wzUe3428NcvO5U/nYZbNOSbZDIceH797A37cc5cdvX8aVC/suJoVCjl1VjWw5XM/mQw1sPlTP1sMNpKWmsHJGMefNKOG8mcVMLsrutSpZ0xjgg3dt4MU9tbzujIl88XXzOweTBTtCvPHHz3PgeAv/+PiF3U7/1RFyQzItWm8O1DZz3Q+fpSgnnT9/+DzyM9NwzvHL5/bytUe2MSY7nerGAP/7hkXcuOLkHOP+9Qf55H2v8OcPndvtB4+h4JxjV2Uju6saWTGtmKJBDJCTU8Vk4N5Iilw06lraYhyJiIhI//zkqd3kpPu46dyp/PjJ3Zw9rYibVk7t3N/S1sGtd28gLzON77xlySlf9/tSjLedPZmrFo7nW4/t4JfP7eHetQe4dN5Yrl40gQtnl5KZ5uPrf9/O3zYf5bPXzOtXggyQkmLMHpfH7HF5vN4r8EaKZwP5qr44N4O73nc2P3lyN9//106e31XNl1Yt4JpFE/jlc3t45WA9379xabcJcuQ1DrdJRdn86O1n8s5fvMBHf/8SP3jbMv77z5v4y8uHuWL+OP7vLWfw7l+t5Tv/eJVVSyae1C7x8MbDlBVmscQbgDcczIxZ4/KYNW5gLRUyfBIqSc5MSyE9NUU9ySIikhD21zTz8MbDvP+C6fznFXN49aifrzy8lTPKCztnPPjyw1vYWdnIb9+7otfBb2Ny0vna6xdx41mT+e2avTy29RgPvHyYnHQfSyYX8tyuGt61cgrvHcTgNDj9PtY0XwofuXQWVywYz3/d/wq33v0Sf557iGd3VXPZvHG8Lg7aJFfOKOaL1y3gsw9s5vxv/Iv6lnY+ecVsPnTxTFJSjM9cPZc3/ng1v3hmDx+5dBYAx5vaeHZnNe89f5p6fEeZ05vJOkbMjIKsNBqUJIuISAK445ndpKak8J7zp5GSYnz7LWcwNi+TD921gbrmNh565TC/f/EAH7x4Ro8DyrpaVF7AN990Bmv/+zJ++94VXLekjO1H/Lx2wTg+/7oFMU/k5ozP448fPJfbrprLM7uqSU9N4auvXxjzuCLecc4Ubj53KgC/vPksbr1kVucCZWdOKeKK+eP46dMVVDeG+4Mf3XKUYMhx7eKJsQpZYiShKskQngaurllJsoiIxLcqf4B71x3kjWeWMc6bB7YwO53b376MN//keT74uw1sPlTPssmF/Mflswf8/Gm+FC6YVcoFs0r53zcsGurwByXVl8IHLprBNYsmEAiGOl9/vPjidQv43LXzu23z+K8r5/La7z7NDx7fyZdWLeThjUeYUpzNwrLYL2YiIyuhKslwYmlqERGR4eScI3QaS/dG/Oq5PbR3hLjlwhknbV8yqZDPXjOf1RU1mMH3blh62ksUx7tJRdnMHDu4RTyGS0990DPH5vLWsyZx1wv7Wb/vOM/vrubaGC2LLLGVcJXkgqy0zmUMRUREhsu//XY9jYEgd757BempA0tiG1rb+e3qfVy9cMIp040B3LRyCq3tHZwxqZBJRdlDFbIMkY9dNosHXjrE+369lpBDrRajVMJ9dC3IViVZRESGl7+1nX9tr+T53TV88aEtA378XWv24w8E+cBFM7rdb2b820UzOGd68WBDlWEwNi+T910wnePN7cwozWFuL3NTS/JKvCRZ7RYiIr0ysyvNbIeZ7TKz27rZP9nMnjCzl8xso5ldHYs449lzu2oIhhwXzCrh7hf287s1+/r92Nb2Dn7x7B4umFXCovKCYYxShtMtF05nSnE2N66YrFaLUSrhkuTCrHQaA0GCHaFYhyIiEnfMzAfcDlwFzAduNLP5XQ77LHCvc24pcAPwo5GNMv499WoVuRmp/Pxdy3nNnFK++OAWXqio6fNx9S3tfObPm6huDPDBHqrIkhhyM1J58pMX874Lpsc6FImRhEuSC7LCbdQNrcEYRyIiEpdWALuccxXOuTbgHmBVl2McEBmqXwAcHsH44p5zjqd2VHLezGIyUn1878alTC7O5kN3beDg8eYeH/O3TUe4/DtP8cBLh/jARTNYOUOtFIlOFeTRLeGS5MLs8DKNdc1adU9EpBtlwIGo+we9bdG+CLzDzA4CjwAfGZnQEsOuykYO17dy0eyxAORnpvGzm5bTFgxxy2/Ws7uqkSp/gNb2DpxzHKlv4ZbfrueDd22gJDeDBz58HrddNVcJlkiCS8jZLQD1JYuInL4bgTudc982s5XAb81soXPupD42M7sFuAVg8uTJMQgzNp7cUQXARXNOLO4xozSX79+4lPf8ei2Xfvupzu2p3jRiqT7j01fN5T3nT0va6dxERpuES5LzvSS5TkmyiEh3DgGTou6Xe9uivRe4EsA5t9rMMoESoDL6IOfcHcAdAMuXLz/9CYMTzFOvVjFrbC5lhVknbX/N3LE8dOv57Kz009gaxB8I0hQIEgw53r5iCpOLNZWbSDJJuCS5MDucJGtpahGRbq0FZpnZNMLJ8Q3A27ocsx+4FLjTzOYBmUDViEYZp5rbgry4p5Z3nTul2/0LywpYWKYZK0RGg4T7TijSbqGlqUVETuWcCwK3Ao8C2wjPYrHFzL5sZtd5h30CeL+ZvQL8HrjZOTdqKsW9Wb27hraOUGc/soiMXglXSVZPsohI75xzjxAekBe97fNRt7cC5410XIngqVeryErzcda0MbEORURiLOEqyWm+FHLSfUqSRURkSDnneHJHFefOCE/9JiKjW8IlyRCeBk7tFiIiMpT21jSzv7b5pFktRGT0SsgkOV9LU4uIyBB7akd4co+LZitJFpEETZILslKpb9FiIiIiMnSefLWKaSU5TCnOiXUoIhIHEjJJLsxKVyVZRESGTGt7B2sqalRFFpFOCZkkF2SlqSdZRESGzAt7amltDylJFpFOCZkkF2arJ1lERIbOH9cfJCM1hXOmF8c6FBGJEwmZJOdnpREIhmht74h1KCIikuCe313Ng68c5v0XTCcrXVO/iUhYQibJkaWpVU0WEZHBaAuG+NwDm5lUlMWtl8yMdTgiEkcSMknWqnsiIjIUfvZMBburmvjSdQvITFMVWUROSOgkWYP3RETkdB2obeYH/9rJaxeM45K542IdjojEmYRMkguz0gFVkkVE5PR96aEtpJjxhdctiHUoIhKHEjJJPlFJ1oIiIiIycP/Yeox/bqvko5fOYmJhVqzDEZE4lJhJsgbuiYjIaWpuC/LFB7cwZ1we7zl/WqzDEZE4lZBJcl5GKmbQoCRZREQG6PFtlRyqa+Fz184nzZeQfwZFZAQk5NUhJcXCq+4pSRYRkQHaVdmIGSyfOibWoYhIHEvIJBnCfclqtxARkYGqqG6irDBLU76JSK8SOknWFHAiIjJQe6obmV6aG+swRCTOJXSSrEqyiIgMhHOOPVVNTC/JiXUoIhLnlCSLiMiocawhQFNbBzNKlSSLSO8SNkkuzFaSLCIiA1NR1QigdgsR6VPCJsmRSrJzLtahiIhIgthd3QTAdFWSRaQPCZskF2al0xFyNAaCsQ5FREQSREVVI9npPsbnZ8Y6FBGJcwmbJEeWplbLhYiI9FdFVRPTSnIws1iHIiJxLmGT5HwvSdY0cCIi0l8Vmv5NRPopYZPkwuxwkqylqUVEpD9a2zs4eLxF07+JSL8kbJIcabfQ0tQiItIf+2qacU6D9kSkfxI2SY5UktWTLCIi/dE5/VuJ2i1EpG8JmyRr4J6IiAxEhTf92zRVkkWkHxI2Sc5K85HuS9HAPRER6ZfdVY2My88gNyM11qGISALoM0k2s1+aWaWZbe5h/9vNbKOZbTKz583sjKEPs9vzkq+lqUVEpJ8qqprUaiEi/dafSvKdwJW97N8DXOScWwR8BbhjCOLql/DS1G0jdToREUlQzjkqqho1aE9E+q3P75ycc0+b2dRe9j8fdXcNUD4EcfVLgSrJIiLSD7VNbTS0BjVHsoj021D3JL8X+NsQP2ePCrLS1JMsIiJ9igzaUyVZRPpryJJkM3sN4ST5U70cc4uZrTOzdVVVVYM+Z6EqySIi0g+R6d9mqCdZRPppSJJkM1sM/BxY5Zyr6ek459wdzrnlzrnlpaWlgz6vBu6JiEh/VFQ1kZ6aQtmYrFiHIiIJYtBJsplNBv4EvNM59+rgQ+q/wuw0/K1BOkJuJE8rIiIJZndVE1OLs/GlWKxDEZEE0efAPTP7PXAxUGJmB4EvAGkAzrmfAJ8HioEfmRlA0Dm3fLgCjhZZUKShpZ0xOekjcUoREUlAFdWNzB6bF+swRCSB9Gd2ixv72P8+4H1DFtEARJamrlOSLCLSycyuBL4H+ICfO+e+3mX//wNe493NBsY65wpHNMgR1N4RYn9NM1cuGB/rUEQkgST0skNamlpE5GRm5gNuBy4HDgJrzexB59zWyDHOuY9HHf8RYOmIBzqCDtQ2Eww5Tf8mIgOSsMtSw4kkua5ZC4qIiHhWALuccxXOuTbgHmBVL8ffCPx+RCKLkYoqTf8mIgOX4ElyuMVClWQRkU5lwIGo+we9bacwsynANOBfPewf0mk7Y6WiWtO/icjAJXiSfGLgnoiIDNgNwP3OuY7udg71tJ2xUlHVRHFOOgXeOBYRkf5IiiRZq+6JiHQ6BEyKul/ubevODSR5qwWEk+RpJWq1EJGBSegkOT01hZx0H7XqSRYRiVgLzDKzaWaWTjgRfrDrQWY2FxgDrB7h+EZcRXWj+pFFZMASOkkGmFiYxaHjLbEOQ0QkLjjngsCtwKPANuBe59wWM/uymV0XdegNwD3OuaRejamuuY3qxjbNbCEiA5bQU8ABTCrK5oCSZBGRTs65R4BHumz7fJf7XxzJmGJl/b7jAJxRXhjbQEQk4SR8JXnSmCwO1jaT5MUQERE5DS/uqSXdl8LSyYWxDkVEEkziJ8lF2fgDQU0DJyIip1izp5YzJhWQmeaLdSgikmASPkkuH5MNwIFatVyIiMgJTYEgmw/Vs2JaUaxDEZEElPBJ8qSiLAAOHG+OcSQiIhJP1u87TkfIcfa04liHIiIJKAmS5EglWUmyiIic8OKeWnwpxrIpY2IdiogkoIRPkvMz0yjISlMlWURETvLCnhoWlhWQm5HwEzmJSAwkfJIM4ZYL9SSLiEhEa3sHrxyo52z1I4vIaUqOJHlMtirJIiLS6aX9dbR1hJQki8hpS44kuSibg8dbCIU0V7KIiIT7kc1g+VQlySJyepIjSR6TRVswRFVjINahiIhIHHhhTw3zxudTkJUW61BEJEElRZJcrhkuRETE0xYMsWH/cc6eriqyiJy+pEiSJ0UWFFFfsojIqLfpUB2t7epHFpHBSYokuXyMt6CIZrgQERn1XthTC8BZ6kcWkUFIiiQ5M83H2LwMtVuIiAgvVNQya2wuxbkZsQ5FRBJYUiTJEJ7hQu0WIiKjW7AjxPp96kcWkcFLniR5jBYUEREZ7bYeaaAxEGTFtOJYhyIiCS55kuSibI7Ut9DeEYp1KCIiEiMvev3IGrQnIoOVPEnymGxCDo7UtcY6FBERiZE1FbVMLc5mXH5mrEMRkQSXNElyeZE3w4X6kkVERqUth+t5ckclF88ZG+tQRCQJJE2S3DlXsma4EBEZddo7Qnzyvo2MyUnnY5fNinU4IpIEUmMdwFCZUJCJL8VUSRYRGYVuf2IX24408LObllOYnR7rcEQkCSRNJTnVl8LEwkzNcCEiMspsOVzPD/+1i+uXTOTy+eNiHY6IJImkSZIh3HKhSrKIyOgRabMozE7ni9ctiHU4IpJEki9JViVZRGTUiLRZfO31C9VmISJDKml6kgEmFWVR3Rigpa2DrHRfrMMREZEhcqC2mduf2EVmmo/8rDQKstJITbHONosrFoyPdYgikmSSLEkOz3Bx8Hgzs8blxTgaEREZKr9ZvZc/rDtAbnoq/kCwc/uEgky+8Dq1WYjI0EuqJLk8Mg2ckmQRkaTy+PZKzp9Zwm/fezYdIYe/tZ2GliAleelkpyfVnzIRiRPJ1ZPsLSiyv0aD90REksWe6iYqqpq4bF545gpfilGYnc7k4mwlyCIybJIqSS7NzSAzLYUDxzV4T0QkWTy+7RgAl8zVSnoiMnKSKkk2M8rHZGvVPRGRJPL4tkrmjMvrHHciIjISkipJBpg0JkuVZBGRJFHf0s7avbVcMk9VZBEZWcmXJBdlc7C2GedcrEMREZFBevrVKoIhx2VKkkVkhCVfkjwmG38gSH1Le6xDERGRQXp82zGKctJZMmlMrEMRkVEm+ZJkb4YLrbwnIpLYgh0hnny1iovnlOJLsViHIyKjTNIlydFzJYuISOLasL+OuuZ2Lp07LtahiMgolHRJ8pTicJK8T3Mli4gktMe3HyM1xbhwdkmsQxGRUSjpkuS8zDSKctLZX9sU61BERGQQHt9WydnTi8jLTIt1KCIyCiVdkgwwuShblWQRkQS2r6aJXZWNarUQkZhJyiR5SrGSZBEZvczsSjPbYWa7zOy2Ho55i5ltNbMtZnb3SMfYl8e3VQJwqaZ+E5EYSc4kuSibI/UttAVDsQ5FRGREmZkPuB24CpgP3Ghm87scMwv4NHCec24B8LGRjrMvj28/xsyxuUwpzol1KCIySvWZJJvZL82s0sw297DfzOz7XsVio5ktG/owB2ZycQ4hBwc1w4WIjD4rgF3OuQrnXBtwD7CqyzHvB253zh0HcM5VjnCMvfK3tvNCRa2qyCISU/2pJN8JXNnL/quAWd7PLcCPBx/W4HTOcFGrJFlERp0y4EDU/YPetmizgdlm9pyZrTGzbq/xZnaLma0zs3VVVVXDFO6pdlU2Egw5zppSNGLnFBHpqs8k2Tn3NFDbyyGrgN+4sDVAoZlNGKoAT0ckSd6vvmQRke6kEi5sXAzcCPzMzAq7HuScu8M5t9w5t7y0tHTEgqv0BwAYl585YucUEelqKHqS+1O1AEauKlGam0F2uk+D90RkNDoETIq6X+5ti3YQeNA51+6c2wO8SjhpjgtVXpJcmpcR40hEZDQb0YF7I1WVMDMmF2VrrmQRGY3WArPMbJqZpQM3AA92OeYBwlVkzKyEcPtFxQjG2KtKfwAzKM5Nj3UoIjKKDUWS3J+qxYibXJTNXlWSRWSUcc4FgVuBR4FtwL3OuS1m9mUzu8477FGgxsy2Ak8A/+mcq4lNxKeq8gcoyk4nzZeUEzCJSIJIHYLneBC41czuAc4G6p1zR4bgeQdlSnE2T75aRSjkSEmxWIcjIjJinHOPAI902fb5qNsO+A/vJ+5U+VvVaiEiMddnkmxmvyf8tVyJmR0EvgCkATjnfkL4Qnw1sAtoBt49XMEOxOTiHNqCIY75W5lQkBXrcEREpJ+q/AElySISc30myc65G/vY74APD1lEQ2RKkTcNXE2zkmQRkQRS6Q8wc2xerMMQkVEuaRu+NA2ciEjiCYUc1Y0BxuarkiwisZW0SfLEwix8KcY+zXAhIpIw6lraae9wlOYqSRaR2EraJDnNl0JZYZbmShYRSSCROZJVSRaRWEvaJBnCLRf7tTS1iEjCqPS3AqiSLCIxl9RJ8uSibFWSRUQSyIlKspakFpHYSuokeWpxDvUt7dQ3t8c6FBER6YdKLUktInEiqZPkyd4MFxq8JyKSGKr8AbLTfeRmDMVaVyIipy+pk+TINHBquRARSQyVWkhEROJEUifJk70FRTR4T0QkMVT5WxmrJFlE4kBSJ8nZ6amU5mWwt1rtFiIiiUCVZBGJF0mdJEN4eep9qiSLiCSEKn+AsXma2UJEYi/pk+TJxdlamlpEJAG0tnfgbw2qkiwicSHpk+QpRTkcbWiltb0j1qGIiEgvKhs0/ZuIxI/kT5K9GS4OqOVCRCSuVTV6q+0pSRaROJD0SfJkTQMnIpIQIpVkzW4hIvEg6ZPkKUWRBUWUJIuIxLOqxkiSrIF7IhJ7SZ8kF+Wkk5uRyv4aTQMnIhLPKhsCpFj4ui0iEmtJnySbGVOKNQ2ciEi8q/IHKMnNwJdisQ5FRCT5k2QID97TNHAiIvGt0t+qQXsiEjdGRZI8uSiHA8eb6Qi5WIciIiI9qGoMaNCeiMSNUZEkTy/Nob3Dsae6MdahiIhIDyobtCS1iMSPUZEknzOtGIDVu2tiHImIiHSnI+SoaWrTzBYiEjdGRZI8qSiLssIsnleSLCISl2qb2ugIOVWSRSRujIok2cw4d0YxqytqCKkvWUQk7lT5tZCIiMSXUZEkA5w7s5i65na2HW2IdSgiItJFpV9LUotIfBk1SfLK6SVA733JT+6oZNsRJdEiIiPtRCVZPckiEh9GTZI8viCT6SU5PfYl+1vb+cDv1vPJ+17BObVkiIiMpEovSVYlWUTixahJkgFWzijmhYoa2jtCp+z768YjtLaH2HK4gVcO1scgOhGR0avKHyAvI5WsdF+sQxERAUZZknzujBKa2jrYdOjUJPi+9QeZWpxNdrqP363ZF4PoRERGryp/gNJ8VZFFJH6MqiT5nOlFwKl9yRVVjazfd5wbV0xm1ZIyHnrlMPXN7bEIUURkVKryByjNVZIsIvFjVCXJxbkZzB2fx/O7q0/afv/6g/hSjNcvLeMd50wmEAxx/4aDMYpSRGT0qfS3MjZfg/ZEJH6MqiQZwi0X6/YeJxDsAMKrPP1pwyEuml3K2PxMFkwsYOnkQu56YZ8G8ImIjJBKVZJFJM6MwiS5mEAwxEv76wB4dlc1RxtaefOZ5Z3HvP3sKVRUNbG6Qiv0iYgMt6ZAkOa2DsaqJ1lE4sioS5JXTC8ixeicCu6+dQcozE7jknljO4+5dvEECrLSuOuF/bEKU0Rk1Oic/k2VZBGJI6MuSc7PTGNReSGrd1dT39zOY1uPcf2SMjJST0w7lJnm481nlvPo5qOdq0CJiMjw6FxIRJVkEYkjoy5JBlg5vZiX9tfxh3X7aQuGeFNUq0XE286eTDDkuHftgRhEKCIyemhJahGJR6MyST53RjHBkOO7/9zJ3PF5LJiYf8ox00tzOW9mMb9/8QAdIQ3gE5HEYWZXmtkOM9tlZrd1s/9mM6sys5e9n/fFIs4ILUktIvFoVCbJy6eOIc1nNLd18KYzyzGzbo97+9lTOFTXwpM7Kkc4QhGR02NmPuB24CpgPnCjmc3v5tA/OOeWeD8/H9Egu6j0B0hNMQqz0mIZhojISUZlkpydnsrSSWNI9eZG7snl88dRkpvOvevUciEiCWMFsMs5V+GcawPuAVbFOKZeVfkDlOZlkJLSfcFCRCQWRmWSDPAfV8zma69fRHEvo6nTfCm8fmkZj2+rpKYxMILRiYictjIg+pP9QW9bV280s41mdr+ZTeruiczsFjNbZ2brqqqqhiNWwJsjWf3IIhJnRm2SfM70Yt5yVrd/F07y5uWTCIYcD7x8eASiEhEZEQ8BU51zi4F/AL/u7iDn3B3OueXOueWlpaXDFkyVP8BYJckiEmdGbZLcX7PH5XFGeQH3rTugFfhEJBEcAqIrAOXetk7OuRrnXOTrsZ8DZ45QbN2q8reqkiwicUdJcj+8afkkth/1s/lQQ6xDERHpy1pglplNM7N04AbgwegDzGxC1N3rgG0jGN9JnHPUNbdTmJ0eqxBERLqlJLkfrls8kfTUFO5brwF8IhLfnHNB4FbgUcLJ773OuS1m9mUzu8477N/NbIuZvQL8O3BzbKKFQDBEMOTIy0yNVQgiIt3SVakfCrLTeO2C8fzl5cN85up5ZKb5+n6QiEiMOOceAR7psu3zUbc/DXx6pOPqTkNrOwB5mZr+TUTiiyrJ/fTmM8upb2nnn9uOxToUEZGk4W8NApCXoZqNiMQXJcn9dN7MEiYWZHLfuoOxDkVEJGk0RpJktVuISJxRktxPvhTjjWeW8/TOKo7Ut3Ru31PdxC2/Wcdbf7pas1+IiAxQZyVZ7RYiEmf6lSSb2ZVmtsPMdpnZbd3sn2xmT5jZS97k9FcPfaix96Yzy3EO/rThEPXN7Xzl4a1c8f+e4h/bjvHCnlpePdYY6xBFRBJKYyDck5yrdgsRiTN9Jslm5gNuB64C5gM3mtn8Lod9lvAI6qWEpxv60VAHGg+mFOewYloRv3puLxf/3xP88rk9vOnMcv7y4fMAeGbn8K1IJSKSjBrUbiEicao/leQVwC7nXIVzrg24B1jV5RgH5Hu3C4CkXZ7ubSsmU90YYO74fB7+yPn87xsWs7i8kBmlOTyzszrW4YmIJBS/kmQRiVP9uSqVAdETBB8Ezu5yzBeBx8zsI0AOcFl3T2RmtwC3AEyePHmgscaFVUsmsmRSIVOKszGzzu0XzCrlnrX7aW3v0BRxIiL9FBm4p3YLEYk3QzVw70bgTudcOXA18FszO+W5nXN3OOeWO+eWl5aWDtGpR5aZMbUk56QEGeCCWSW0tofYsO94jCITEUk8/tZ2stN9pPo0jlxE4kt/rkqHgElR98u9bdHeC9wL4JxbDWQCJUMRYKI4Z3oxaT7jabVciIj0m781qCqyiMSl/iTJa4FZZjbNzNIJD8x7sMsx+4FLAcxsHuEkeVSNYsvJSGXZ5DEavCciMgCNgaD6kUUkLvWZJDvngsCtwKPANsKzWGwxsy+b2XXeYZ8A3m9mrwC/B252o3DS4AtmlbDlcAM1jYFYhyIikhAaWtvJ1RzJIhKH+vXx3Tn3CPBIl22fj7q9FThvaENLPBfMKuX/HnuVZ3dVs2pJWazDERGJe42BIPmqJItIHNJIiSG0sKyAwuw0TQUnItJP/la1W4hIfFKSPIR8KcZ5M0p4dme1lqgWEekHf2u7Bu6JSFxSkjzELphVwtGGVnZVaolqEZG+NLYGyVNPsojEISXJQ+z8WeGZ7zQVnIhI7zpCjqa2DrVbiEhcUpI8xMrHZDO9JIdnNRWciEivGgNabU9E4peS5GFwwawS1lTUEgh2xDoUEZG45W9tByBf7RYiEoeUJA+D82eV0tLewXotUS0i0iN/q1dJVruFiMQhJcnD4JzpRaSmGM+qL1lEpEeRdgv1JItIPFKSPAzyMtNYPnUMf99yVFPBiYj0INJuodktRCQeKUkeJm9YVk5FVZNaLkREetDZbqGBeyISh5QkD5NrFk0gO93HvesOxDoUEZG4FEmStSy1iMQjJcnDJCcjlWsXT+DhjUc6++5EROSESJKsdgsRiUdKkofRW8+aRHNbB49sPBLrUERE4k5joB1fipGZpj9FIhJ/dGUaRssmj2F6aQ5/GGTLxcMbD/PIJiXaIpJc/K1B8jJTMbNYhyIicgolycPIzHjr8kms33ecXZWNA368c47vP76TW+9+iS89tGUYIhQRiR1/a1CD9kQkbilJHmavX1aGL8W4b/3AqsmhkOPLD2/lO/94lYkFmRxrCFDpbx2mKEVERl64kqx+ZBGJT0qSh9nYvEwumTuWP64/RHtHqF+Pae8I8Yn7XuFXz+3lPedN49tvWQLAlsMNwxjp0NhyuJ7tR+M/ThGJPX9ruxYSEZG4pSR5BLxl+SSqGwM8uaOqz2Nb2zv4wG/X8+eXDvHJK2bzuWvnsbAsH4DNB+uHO9RB++R9G/nyQ1tjHYaIJAB/a5A8tVuISJxSkjwCLp5TSkluBn9Y23fLxZ3P7+Xx7ZX8z/ULufWSWZgZeZlpTCvJYfPh+E6SW9s7ePWYn7rm9liHIiIJoDEQVCVZROKWkuQRkOZL4Y1nlvHEjso++4rX7zvO9NIc3nHOlJO2L5iYz+ZD8d3GsPVIAx0hhz+gJFlE+uZvbSdXSbKIxCklySPkzWdOoiPkePDlw70et/lQPYvLCk7ZvqisgEN1LRxvahuuEAdtk9cOElkgQESkJ845r5KsgXsiEp+UJI+QmWNzmVKczYt7ans8ptLfypH6VhZ2kyRHtsVzy8WmQ+HYGluDOOdiHI2IxLNAMER7h1O7hYjELSXJI2jZ5DFs2F/XYwK52UsyF5cXnrJvwURv8F4ct1xE4g+GHK3t/ZvJQ0RGp4bWcFuWBu6JSLxSkjyClk0ZQ3VjgIPHW7rdv/FgPWYnEuJohdnpTCrK6rGS3BQI8ol7X+FAbfOQxtxfre0d7KxspCgnHQj3GoqI9KTRa8tSu4WIxCslySNo2eRCADbsP97t/k0H65lZmktOD5WVhRMLOqu1XT265Sh/3HCQf247NiSxDlRk0N7K6cUA+APqSxaRnvk7k2RVkkUkPilJHkFzxuWRne5j/b4ekuRD9SwqP7UfOWJhWQH7apqpbzm1SvvIpiMAVFQ1DU2wAxRJ3lfO8JJkDd4TiRkzu9LMdpjZLjO7rZfj3mhmzsyWj2R8EJ7+DdCy1CISt5Qkj6BUXwpnlBd2W0k+1tBKpT/Q7cwWEZHBe1u7rLzX0NrO069WA7C7qnEII+6/jQfrKc5JZ/a4PEDtFkPtLT9Zzfcf3xnrMCQBmJkPuB24CpgP3Ghm87s5Lg/4KPDCyEYYFrlGqN1CROKVkuQRtmxKIduO+GluO7nSutGbPq23SvKJwXsnt1w8vu0YbR0hppfmxLSSvLCsoPOr00ZVkodMa3sHa/fV8szOvlds7Oq+dQf44O/Wa7aR0WUFsMs5V+GcawPuAVZ1c9xXgG8AvU/ePkwa1G4hInFOSfIIWzZ5DB0h15kUR2w6WEeKwfwJPSfJJbkZTCjIPGXw3l83HmFCQSavX1LG0YbWzq8xR0pk0N7i8oLOr07VbjF09lQ34RxsP+ofcLL7j63H+Nvmo+yO0YcniYkyIHp5z4Petk5mtgyY5Jz760gGFq1RSbKIxDklySNs6eQxwKmD9zYeqmf2uDyy0n29Pn5Bl8F7kVaLqxZOYObYXAD2jHBCFBm0t7CsgHzvq9OGBGi3+P2L+3l0y9FYh9GnyLcD/tYgR+oHVvTbVxOe7eQfW2MzoFPij5mlAN8BPtGPY28xs3Vmtq6qauDfZPQm8kFaPckiEq+UJI+wopx0ppfksCFq8J5zjs2H6lnUSz9yxKKyAiqqmzqrxZFWi2sWj2eGlySPdF9yJGlfVFbQucTsSFezB6q9I8T/PLyV/7p/Y9wn9BVRv8/tR/s/T3Yo5NhXG06wYzXricTEIWBS1P1yb1tEHrAQeNLM9gLnAA92N3jPOXeHc265c255aWnpkAbpb20nK81Hqk9/hkQkPunqFANLuywqcqS+lerGNhb30o8csbAsH+dg25FwsvTXjUcZn5/J0kljmFKcTYqdnFSNhE3eoL0JBZn4UozsdF/ct1tsOlRPU1sH9S3t/PyZPbEOp1e7qxopzA5X6Lcf9ff7ccf8rbS2h5hQkMmG/cep8geGK0SJL2uBWWY2zczSgRuAByM7nXP1zrkS59xU59xUYA1wnXNu3UgGGV6SWlVkEYlfSpJjYNmUQmqb2jq/Co/0J3e3HHVXnctTH6rH39rO0zuruGrReFJSjIxUH5OKstldPbLtFpu8QXtmBoR7DON9dovVu2sAWDm9mF8+u4faprYYR9SziuomFpUVMLEgkx0DSJL3Vof/fb37vKk4B//armryaOCcCwK3Ao8C24B7nXNbzOzLZnZdbKM7wd+qJFlE4puS5BhY1qUvedOhOlJTjHkTTl1pr6tx+ZmU5mWw+VADj2+rpC0Y4ppFEzr3Ty/JYXflyFWSI4P2oltF8jLT4r7dYk1FDXPG5fHlVQtoagvy06d2xzqkbjnnqKhqYnpJDnPG57H9SP+T5H014Q9LVy2cQFlhFv/YWjlcYUqccc494pyb7Zyb4Zz7qrft8865B7s59uKRriJDeMGhXE3/JiJxTElyDMwel0duRmpnkrzxYHjQXmZa74P2IhZOzGfzoXr+uukI4/MzO5NugBmlueypbiIUOv0pv3ZV+jn/G//i5l+9yN83H6G9I9TjsdGD9iJyM1Ljut2iLRhi7d5aVs4oZta4PF6/pIxfr95LZUNMZsLqVZU/QGMgyPTSXOZOyGd3VSNtwZ5/H9H21jST5jMmFmZx+fxxPLuripa2jmGOWKR//K3t5KuSLCJxTElyDPhSjCWTClm/r65z0F5/+pEjFpYVsLPSz1OvVnHlwnCrRcT00lwCwRCH6lpOK7ZKfyvv+uVamts62HakgQ/8bgMr//dx/vdv29jTTRtH56C98uhKcmrnHKgjwTnHkfoWnt1ZzZ3P7eFzD2zmP+97hdb27hPClw/U0doe4hxvCe2PXjaLYIfj9id2DWucbcEQP3py14BaUSJTt80ozWXu+DyCIUdFdf++KdhX08Skomx8KcZl88bR2h46rbmWRYaDvzWomS1EJK7pChUjyyYX8sMndrHjmJ/jze29LiLS1cKyAkIunHRds3jCSfuml+YA4T7WSUXZA4qpKRDkPXeupbapjT/82znMn5DPU69Wcc/aA/z8mT387OkKvvb6RdywYnLnYzYdrKcoJ52JBZmd2/Iz0047SY/20CuHeWJHJcunFLFiWhEzSnM6+56r/AGe2FHJkzsqeWZn9UmV6+x0H81tHZw/q4RVS8pOed7Vu2swg3OmFwEwpTiHNy+fxN0v7uf9F06nfMzA3rf++tf2Y3zz7ztIS0nh/RdO79djIjOVTC/NoSQvHYAdR/3MHd93a87emmamFof/PZw9vYi8zFT+ue0YVywYf5qvQGToNKonWUTinK5QMbJ0yhhCDn63Zh9Av6Z/i4i0NozLz+DMqFYLCFccAXZXNnLR7P5P2RTsCHHr3RvYeriBn79rOYvLCwG4dN44Lp03jmMNrfzX/Ru57U+baGrr4L3nTwNOHbQH4XaLoVhx73bvQ8SfNoRnryrJTefMKWM4Ut/aOdhxXH4G1yyawMKyAmaU5jJjbA4lORlc+K0nuH/9we6T5Ipq5k/IpzA7vXPbRy6ZyR/XH+QHj+/iG29aPOjYu/PcrvBgwYc2Hu53klxR1URWmo/x+ZmU5GaQ5jO2HfGzaknvj3POsa+mqfODQJovhdfMGcvj2yrpCDl8Ud8+iMSCv7VdS1KLSFxTkhwjyyaFk9s/bThEms+YMz6v34+dWJDJlOJsrlk04aRWCwgnknmZqf3+Sh7CCdXn/rKFJ3ZU8dXXL+SSueNOOWZcfiZ33HQmH/39y3zl4a00B4K8/8Lp7Kxs5LJ5Jx8fnt1icElyfXM7O475+fhls3ndGRN5cU8NL+ypZf2+4xTnpPPJK2bzmrljmT8h/6QEPeKNy8r5/r92criuhYmFWZ3bW9s72LC/jpvOmXLS8RMLs3j7OZP5zep9fODiGUwryRlU/N15bnc1vhRj48F69lY3MbUf56iobmRaSQ4pKUZ6ijGjNJcd/ZgruaoxQHNbR2clGeCy+eN48JXDvHzgOGdOKRrUaxEZjI6Qo6mtQ+0WIhLX1JMcIwXZacwozaG5rYO54/PJSO3foD0AM+Oxj1/IJ66Y0+2+GaW5nau09cdPn67g9y/u50MXz+DtZ0/p8biMVB8/fNtS3rC0jG//41VuvXvDKYP2IDy7RUt7R68D/vqybl8tzsFZU4uYVpLDW8+azHfesoSn/vM1/OlD53HrJbNYMLGg2wQZwkmyc/Dnlw6dtH3DvuO0BUOsnFF8ymM+dPFMnHM80OUxQ+FofSsVVU28a+VUAB7eeLhfj6uoaupsoQGYMz6vX9PARaYXnFJ8onXk4jmlpPmMx7T6nsRYZPYbtVuISDxTkhxDkVkpBtKPHJGR6uvxK/PppTn9XnUv2BHi9id2ccncsXyym6S7q1RfCv/35jN4xzmT+ee28JRiXeOP/OFrGsQ0cC/uqSXNZyydXHhaj59cnM2KaUX8cf3BzkVbAFZX1OBLMVZMO7WSWpqXwcyxuWw8WHeaUffsuV3VALzxzDJWTC3ioVeO9PmY1vYODhxv7myhAZg7Pp/D9a3UN/c++C8yyDK6Ip6fmcY504uHdInqUMgRHMSHIRmdIoNX89VuISJxTElyDJ05JZwkLx5AP3J/zCjN5VhDoF9zFW853IC/Ncj1S8tOad3oSUqK8ZVVC/n3S2Zy6dyxJw3aAzqXph5My8WLe2tZXF7Y72nxuvOmM8upqG5iw/66zm2rd9ewsKygx17IxeWFbDpUf1JiPRSe211NUU4688bn87ozJrDjmL/PivC+mmac46RK8lyvLWfHsb4e20RqilEW1WoCcNm8cVRUNQ3Z0uWfvP8V3vzT1QmRKH/k9y/xxQe3xDoM4UQlOVeVZBGJY0qSY+iSeWO5cHYpr5k7dkifd0Zkhot+JELPR608NxBmxn9cMYdf3HzWKS0PkblPG05z1b2Wtg42Hazvtto7EFcvmkBWmo/71x8EoLktyMsH6np9rYvLC6hubONI/dDNmeyc4/ldNaycXkxKinHVogmkWHj2jt5Efn/RleRI73pffcl7a5opH5NFqu/k/8Uvmx/uH//nEFWTd1c28tL+On7+bHwv7e1vbedvm47wl5cPDWoOcRkakQ/QarcQkXimJDmGxuZl8pv3rGBcfmbfBw9AJKnqT1/y87urmTMuj9K8jCE7f6RKe7ozXLy0/zjBkGPF1MElybkZqVy1cDwPbzxMa3sHa/eGn7e7fuSIyCwjkdkzhkJFdRNHG1o5d2b4vCW5GZw3s4SHNh7utWJd0U3LxISCTPIzU9nWZxW6iSnFpw4MLCvMYp43td9QqG8JfxD6f/94tdt5tOPFmopagiHHcW9AqMRWpN1CA/dEJJ4pSU5Ck4uzSTH6/Eo9EOzoXHluKEX+8J1uu8WLe2sxg2VTxvR9cB/edGY5/tYgj209xurdNaSmGGdN7fl5503IJzXF2HSobtDnjoj0I583o6Rz2+sWT2RfTTObDvWcjO+ubGRCQSY5UYmEmTF3fH6vrRrOOfZVNzO1uPv5nldMHcMrB+qGpEWirqWdy+ePIz01hU//aeOQt6kMlWd2VpHmC3/jsdr79kRi50QlWT3JIhK/lCQnoYxUH5OKsvusJL+0P7zy3HkzS3o9bqAiX6H6A6fXbvHinlrmjc+nIGvwf0DPmV5MWWEW968/yOqKGpZMKiQ7vefqVWaaj9nj8oa0kvzcrmrKCrNOmmnitQvGk+azXlsudlefPLNFxJzxebx61N9jQlrb1IY/EOy2kgzhDx9NbR2DrqiGQo6GlnZmj8vlM1fPY01FLfesPTCo5xwuz+ys5vyZJUwpzmZ1hZLkWIskyVqWWkTimZLkJDWjNLfPSvLzu2tIMQbd+9vVYNot2jtCvLS/bshiSkkx3rCsjGd3VrHpYF2/quZnTCoYssF7HSHH6t01nDez+KTe7YLsNC6aXcrDG4902yPrnKOiqpHpJbmn7Js7IQ9/INjjqoZ7venfppZ0X0mODBjdsO/4gF9PtMa2ICEHBVlp3HDWJM6ZXsTXHtnGsYah6+ceCgdqm9lT3cQFs0pZOb2YFypq6FBfckxFkmQN3BOReKYkOUlNL8lhT3VTr4OUVu+uZlFZwZBUbKPldQ7c6zlJPtxDgrf5UD0t7R1Dmri/cVk5IQchR7+S5EVlhdQ1t3OgdvBLa285XE9Da7Dbav3rzpjIkfpW1u8/NVmtbmzD3xrstpIcmeFi+5HuK8H7asLfIPRUSS4rzGJcfgbrB5kkR6ahK8xKx8z4+hsW0xYM8bkHNsdV28UzO8PtLhfOLmHljGIaWoNsPdz3giwyfBoD7fhSjKxBzF4jIjLc+pUkm9mVZrbDzHaZ2W09HPMWM9tqZlvM7O6hDVMGasbYXALBUI/VxqZAkJf213HuELdaAGSkppDmsx57krcfbeDcr/+LP204eMq+F/fUAuFFRIbK1JIczpo6hvTUlM65qXuz2Jv3eeMQ9CVHlqLuLjm/bN44MtNSum25iHwLML301Ery7HG9TwO3t6aZFINJY7qvJJsZZ04Z021yPhCRQXv53oesqSU5fPzy2Ty29RiPbomfBUue2VnFhIJMZpTmds5ssrqiOsZRjW7+1iC5Gak9LgYkIhIP+kySzcwH3A5cBcwHbjSz+V2OmQV8GjjPObcA+NjQhyoDMd2bEaGnlou1e8Oj/c8d4kF7EE7C8jLTaOyhJ3lvdbgd4Ot/237KgiNr99YyrSRnSGfbAPjyqoV8/4Yl/Zp3efa4PNJ9KWwagr7k53dXM3tcLmPzTp3BJCcjlUvnjeOvG48QCHactC/STz6jm0pyXmYa5WOy2N7D4L291U2UjckiPbXn/72XTR7DgdoWKgfRGtHgJcnR30S87/xpTCnO5ndr9p328w6lYEeI53ZVc8GsEsyMsfmZTC/N0eC9GGtsDWr6NxGJe/2pJK8AdjnnKpxzbcA9wKoux7wfuN05dxzAOVc5tGHKQM0Y2/s0cM/vriHdl8LyKUPbjxyRm5HaYyW5qjEAQKU/wE+f2t25PRRyrN17fNBTv3Vn3oR8rlw4oV/HpqemMG/C4AfvRWYPOXdGz9X6t62YTE1TGz97uuKk7RVVjWSmpTCxIKvbx80dn8f2I923DOyraWJqD60WEZGZQzYMoppc302SnOpL4drFE1hdUUNtU9uAnu+eF/fz6T9tpNI/dD3NGw+F210umFXaue3cGcW8uKd2UMumy+A0tAY1s4WIxL3+JMllQPSQ9YPetmizgdlm9pyZrTGzK7t7IjO7xczWmdm6qqqhmadVuleck05+ZmqPleTnd1ezdHIhWenD0xOYl9lzklztD2AG1yyawE+fruhsCXm10k99SztnDfFAwtOxuLyQzYfqB7XwxIZ9fc8ect7MEq5aOJ4f/GsXB2qbO7dXVIcT3Z5WQZwzPo+K6qZTKtAQbreY0sP0bxELJuaTnpoyqL7kziQ5++Rk5+pFE+gIOR7dcrTfz/X87mo+8+dN/P7FA1z27af4w9r9Q9LX/Myr1Zhx0u9g5fQSmto6ep1+T4aXv7WdPM2RLCJxbqgG7qUCs4CLgRuBn5lZYdeDnHN3OOeWO+eWl5aWdt0tQ8jMmDE2t9tKcl1zG1sON/Ra4RysvMzUHme3qGoMMCY7nc9cMw+Ab/59OwBrvX7ks+MgSV5UXoA/EGRPzekvkPH87mp8KcbZ03t/PZ9/3XxSU4wvPLilMzHcXdV40kp7Xc0dn09HyLG78uT46prbqG9p77OSnJHqY3FZwdAkyV0Gfs6fkM/U4mwe2XSkX89zrKGVf//9S0wvzeWhW89n7vh8PvXHTdxwx5p+rRrZm2d2VrGorICinPTObed4vw+1XMROY0DtFiIS//qTJB8CJkXdL/e2RTsIPOica3fO7QFeJZw0SwxNL8mlovrUJGNNRQ3OwXkzh74fOSI3I63HZamr/QFKczMoK8zilgun85eXD7Nh/3Fe2FPL+PxMysd032IwkiKD9wbTl/zsrmoWlxeQ38fXyhMKsvj45bP51/ZKHtt6jECwgwO1zd32I0fMmxAevNd1zt/I9G89zWwR7cwpY9h8qKHbanR/1LWEZyjI6fJthJlx9aIJPL+775aL9o4QH7n7JZoCHfz47ctYVF7APbecw9ffsIhtRxq48nvP8LkHNrPmNKZta2ht56UDdVww6+QPg8W5GcwZl8cazZccM371JItIAuhPkrwWmGVm08wsHbgBeLDLMQ8QriJjZiWE2y8qkJiaXprDsYbAKauzPb+7hux0H4vLC4ft3Pm9tFtUNQYoyQtX9j5w0QzG5mXw5Ye2snZvLWdNK4qLEe8zS3PJTEs57b7kUMix6WB9v2fpeNe5U5k7Po8vPbiF7Uf8hFz3M1tEzCjN5ZzpRfzgXzs5HpWIRqZ/62m1vWjLpoyhrSPE5kOnNx1afUs7hVlp3f6++tty8X+P7uDFvbV8/Y2LmOXN2pGSYtywYjL//MRFXLtoAveuO8ANd6zh7K89zmcf2MTzu6r71Yqxenc4sY7uR45YOaOYtXtraQuqLzkWGgNBzZEsInGvzyTZORcEbgUeBbYB9zrntpjZl83sOu+wR4EaM9sKPAH8p3NOZZoYu2rheEpy03n9j57jLy+fKP4/t6uaFdOKep39YLDyMlNpDPTQk9wYriRDeIaH/3ztHF4+UMexhsCQL2xyulJ9KSyYWHDay1P7W4MEQ46x/ZylI82Xwv9cv5DD9a186o8bAbqdIznCzPjSdQvxtwb55qM7OrfvrW7GDCYV9SNJnjy4RUXqW9p7nGN7wcR8pvTRcvHolqP89OkK3nHOZFYt6TrMAcbmZfKdty5hw+cu54dvW8rZ04r44/pDvO3nL/DhuzfQ2t57BfyZnVVkp/u6nfZv5YxiWttDvHygrvcXKUPOORfuSdbAPRGJc/3KkpxzjzjnZjvnZjjnvupt+7xz7kHvtnPO/Ydzbr5zbpFz7p7hDFr6Z3ppLn/99wtYMDGfj97zMp99YBMHapvZXdU0LFO/RQtPARc8peLnnKPKH6Ak90Ty+MZl5SwsywcYlpktTteisgI2H2o4rdXZjjeHq7tjstP7OPKE5VOLeOvySZ1Tu00r6b1lYs74PG4+dyr3rN3PxoN1QLiSPLEgq19T3ZXmZTClOLvbvuQdR/285Sere51poqGlvXOO5K7MjGt6abnYV9PEJ+99hcXlBXzu2vndPMMJORmpXLt4Ire/fRkbPnc5n7pyLn/bfJQbf7aGam+mlO48s7OaldOLu/0weM60YszUlxwLgWCI9g5HrgbuiUic04p7SW5cfiZ3v/8c/u3C6fxuzX5W3f4cwLAO2oPwcrMdIUdz28nVvsZAkNb20EnzIKekGN960xl84KIZzBrbc4vBSDtjUgEt7R3sqhz44LHOJDlnYNWyT101l8LsNMbmZfSr0vaxy2ZRkpvB5/6yhVDIsbemqc+ZLaItmxxeVCT6w0xHyPFff9zIi3tr2dLLynS9VZLhRMvFY11aLlrbO/jA7zaQkmLc/rZlZKT2f4aVrHQfH7x4Bj9++zK2HWng9T96rtvfz76aJvbVNJ/SjxxRkJ3G/An5WlQkBiJtWPlqtxCROKckeRRI86Xw6avnccc7z6S9I0RxTjrzJ+QP6zkjg3K69iVXN4aTx66LhcybkM9tV83tccqzWFhUVgjQWaUdiLrIks0DqCQDFOWk8+O3n8mXrlvQr+PzMtP4zNVzeeVAHfeuO+BN/9b3oL2IZVPGUOUPcPD4iZUZf7dmH694bQg1jT0PvOsrSY60XPy1S8vF5/+ymW1HGvh/bz2jX20h3bly4QTuuWUlLW0dvPHHz/PUq1VsP9rAkzsquefF/fzfY68CcMHsnmfRWTm9mA376/ps25Ch5fcG9KrdQkTinT7KjyJXLBjPPycV4m8NDnsyGvkDGF5178Rqc1X+8Nfj0e0W8Wp6SQ456T42Harnzcsn9f2AKJEWg4G0W0R0t4R1b65fUsbdL+znf/+23Zv+rf+J55lev+76fceZVJTN0fpWvvXoDs6aOoa1e49T00s7Q19JcmSWizuerqC2qY2inHTuXXuAe9cd5NbXzOSSueP6/yK7sWRSIX/+0Hm8+861vOuXL56yf8W0os6VJ7tz7sxifv7sHjbsOz4sy7NL9yIfnNVuISLxTlepUWZcfibjhreIDNC5UEDDKZXkcNI11MtOD4eUFGNhWcFpzXARabcoOo0keaAig/iu/cEzQP+mf4uYMz6PnHQf6/cd5/qlZXzhwc0EQyG+/eYlXP7/nqKmhyncQiFHQx9JMoQXjPnxk7t5bMtRFpUX8Lm/bOa8mcV8/PLZ/X+BvZhUlM0fP3guj2w6Ql5mKuPzMxlfkMnYvMw+B6aeNbUIX4qxuqJGSfIIigzo1RRwIhLvdJWSYdFTu0UiVZIhPF/yr1fvo70jRJqv/91Jdc3tpNjIJQLzJ+Zz08qp3Pn83l7nV+7Kl2IsnTyG9fuO89iWozy65RifunIuk4uzKcnN6LHdorEtSMidupBIVwsm5jO5KJv71x/kx0/tZkx2Ot+7YSm+IfwmoyArjRtXTB7w4/Iy07j7fWczb+IIfGqUTmq3EJFEoZ5kGRad7RbdVJJTjJNWQItnZ0wqpC0Y4q41+wb0uOPNbRRmp49oj/VtV83lV+8+q3O+4f5aNmUM24828Lm/bGbu+Dzed8E0AIpz06lp6r7dor65+yWpu4q0XKzbd5xDx1u4/e1L4+oD0tnTi/tc7EWGVuSDsyrJIhLvlCTLsMjtrCSfvOpelT9AUU7GkFYSh9Pl88fxmjmlfPGhrXzz79sJ9XM6uLrmdgr7SCCHWmaaj9fMGTvgx505ZQwhB5X+AP/7hkWdFfOinPQeK8k9LUndnVVLJpJi8Jmr53HmlPiZ4k9iQ0myiCQKXaVkWPQ8u0UgIfqRIzJSffzspuV8/sEt/OjJ3Rw83sK33ry4z2nLjje3ndagvVhYMqmQzLQU3rp8EkujFt4ozsng1S6rNUYMJEmeNyGfDZ+7fMAzfUhyilwTcjRwT0TinK5SMixy070kOXBqT3JJbmIlS6m+FL56/ULKx2Txzb/v4FhDK3e8c3mvrQa1TW2Ujzm96c1GWkFWGk988mLG5mWetL0kN53qpjacc6csPT2QJBkGPhWeJK/GQDtZab4B9fiLiMSCrlIyLFJSjNyM1G7bLRKpkhxhZnzo4pl874YlvLS/jnf84oVej69rbmfMCLdbDMaEgqxTWmCKc9NpC4Zoajt1HuGBJskiEf7WoFotRCQhKEmWYZOXmXpSu4VzjurGtoRMkiNWLSnj1ktmsulQPU1dquTRjje3MSZBBif2pCgn/Hvqbq5kJclyuvytwc4xCyIi8UxJsgybvMzUk2a3aGgJ0tYRojSOZjc4HWWFWcCJOZ+7amnrIBAMjfjAvaFW7LXFVHczeK++pZ3UFCM7vf9LSosANLS2a0YREUkISpJl2ORlpuEPnGi3qEqghUR6E4k/MudzV5GFRBJl4F5PSvqoJBdkpZ3SqyzSl0RtuRKR0UdJsgybcE/yiUpyoi0k0pO+kuTBLEkdTyKV5NpuVt2rb2nvc45kke5U+gOMVZIsIglASbIMm67tFom0JHVvIkl+VQ/tFnXeQhuJNHCvO5EFX7pbmrq+ue8lqUW6aguGqG1K7HEJIjJ6KEmWYZOXmUZDElaSi3LSSbF+tFsk+MC9zDQfuRmp3fZeR9otJD6Z2ZVmtsPMdpnZbd3s/4CZbTKzl83sWTObPxJxRVZw7DrdoIhIPFKSLMMmPLvFiZ7k6sYAqSlGYYInV74Uozg3o8ckuc5LkhN94B54S1P3MHBPSXJ8MjMfcDtwFTAfuLGbJPhu59wi59wS4JvAd0YitsqGSJKc2B+URWR0UJIswyYvI5VAMERbMASEK6/FuemkJMiS1L0p7SVJPu61WxRmJXYlGaA4J73nnmQlyfFqBbDLOVfhnGsD7gFWRR/gnGuIupsD9G+99UGq9CdHy5WIjA6arFKGTWTBgMZAkKLUdKoSbEnq3pTmZfTYk1zb1EZeRirpqYn/GbQ4N4MDtc0nbQuFHA2tSpLjWBlwIOr+QeDsrgeZ2YeB/wDSgUtGIrDIB8ux+clxHRCR5Jb4f8UlbuV6c6FGWi6qGwMJP0dyRGle7+0WhTnJkUAW56SfMnDPHwjinBYSSXTOududczOATwGf7e4YM7vFzNaZ2bqqqqpBn7PS3wpAcU5yXAdEJLkpSZZhE6kkR6aBq/IHEn7QXkRpXgbVjQGcO/Vb6uPN7Qk//VtEcW643SIUOvE6G7TaXrw7BEyKul/ubevJPcD13e1wzt3hnFvunFteWlo66MCq/AGKctKT4lsWEUl+ulLJsIlOkkMhR02CL0kdrTQ3g/YO17k8c7S65jYKkyVJzsmgI3Ty69SS1HFvLTDLzKaZWTpwA/Bg9AFmNivq7jXAzpEITHMki0giUZIswyYv40S7RV1LO8GQS6pKMnQ/DVy4kpwcCWRkQZHolovIPNBKkuOTcy4I3Ao8CmwD7nXObTGzL5vZdd5ht5rZFjN7mXBf8rtGIrZKrbYnIglEA/dk2ERXkpNlIZGIzgVF/AFmjcs7ad/x5rbkabeIWpp65thcIKqSnCQfBJKRc+4R4JEu2z4fdfujIx4UUO0PMKM0JxanFhEZMFWSZdhEz26RLAuJRHRWkrvMcNHeEcLfGkyeJLmbSrLaLeR0OOeo8ge0kIiIJAwlyTJscjsrye1JV0nuqd2ic0nqZJndIpIkR30YUJIsp6OuuZ22jlDSXANEJPkpSZZhk5HqIz01BX/riUpysvyBzM8Mz4N8apIcWW0vOSrJRdndV5LTfEZWmi9WYUkCinzrooF7IpIo1JMswyo/MxV/IAgG6b4U8jOT45+cmXW76l5ktb1kGbiX6kuhMDvtpKWpI6vtmSX+yokyciJLUifLB2URSX7JkbFI3MrLTMPfGqS1vYPSvIykSqy6W3XvuFdJTpaeZIgsKHLidTZoSWo5DVWN4YVEVEkWkUShJFmGVW5GKv7WdjpCjpLc5EkcIZwkd12y+bjXljAmJ3lea3FuBtXdVJJFBiJSSR6br4F7IpIY1JMswyovM5XG1iDVSbSQSER3S1MnW7sFQElu+ikD95Qky0BV+gNkpfnISVcvu4gkBiXJMqzyMlM7B+4ly/RvEaW5GdQ2t9HeEercVtfcRnpqSlINaivKCS9NHVHX0qYkWQasyh9gbH5ytVyJSHJTkizDKjcjjbqWNmqbkm+lrdK8DJzjpAQyvJBIcg1qK87J4HhzO0Hvw0B9syrJMnCV/lb1I4tIQlGSLMMqLzOVYw0BQi55FhKJiF51LyK8JHXy9CMDnb3ktc1thEIOfyCoJFkGTEtSi0iiUZIswyp6yrdk+wPZ3ap7x5uSZ0nqiOLcyNLUbfhbgzgH+UqSZYC02p6IJBolyTKscpM4SR7bzap7x5vbkma1vYgib6aO2qY2rbYnp6W1vQN/azDprgEiktw0BZwMq7zME8nUaGi3qGtuT5rV9iIi7RbVjQHyvd9nsr1GGV7JtuKmiIwOqiTLsMpL4kpyVrqPvIzUzgTAOUddS3tSTf8G4YF7EG63UCVZTkelXwuJiEjiUZIswypSSc5MS0nK+VGjV91raA3SEXJJ15NckJWGL8WoaQooSZbToiWpRSQRKUmWYZWbEa4kJ9uS1BElUQuK1CXhktQAKSnGmOx09STLaYt8kNTAPRFJJEqSZVhFZrdItn7kiNK8DKq9JLm2c0nq5EsgS3LTqW5so64l/BqVJMtAVDYE8KVY5yBQEZFEoCRZhlWk3aI0WZPk3OhKcrjKmoyD2oq9panrW9pJ96WQmaZLh/Rfpb+V4px0fCnJ922SiCQv/aWTYRWZAq4kSXsRS/My8AeCtLR1cDxJ2y0gPHivpqmNhpZ28rOSa0VBGX6RJalFRBKJkmQZVjnpPkrzMpgzLi/WoQyLSIW8ujHAca+SXJSMSXJueufsFgVZmjlSBqZSC4mISALSXzsZVmbGM//1GtJ9yfl5LHrVvbrmNlLs5GnvkkVxTjqNgSBV/kBStpPI8KryB1g4sSDWYYiIDEhyZi4SVzLTfKQkaS9iadSqe7VNbRRmpyfla40sTV1R1aRBezIgHSFHdaPaLUQk8ShJFhmE6CQ5vNpeciaQxd6sBDVNbUqSZUBqmgKEnOZIFpHE068k2cyuNLMdZrbLzG7r5bg3mpkzs+VDF6JI/CrKSccsnCQfb25LykF7cKKSDJr+TQYmMvuLVtsTkUTTZ5JsZj7gduAqYD5wo5nN7+a4POCjwAtDHaRIvErzpVCUnU6VN3AvaZPkqPlt85UkywBU+iOr7Wngnogklv5UklcAu5xzFc65NuAeYFU3x30F+AbQOoTxicS9Um/VvbrmNsYka7tF7okkWZVkGYiqBlWSRSQx9SdJLgMORN0/6G3rZGbLgEnOub8OYWwiCSGSJNc2tTEmSVcUy81IJT01fLlQkiwDEVmSWj3JIpJoBj1wz8xSgO8An+jHsbeY2TozW1dVVTXYU4vEhZLcDA7UNhMIhpJ24J6ZUeJ9AChUkiwDUNnQSn5mKplpvliHIiIyIP1Jkg8Bk6Lul3vbIvKAhcCTZrYXOAd4sLvBe865O5xzy51zy0tLS08/apE4UpoXXo0OknO1vYgir+WiIEk/CMjwqPQHVEUWkYTUnyR5LTDLzKaZWTpwA/BgZKdzrt45V+Kcm+qcmwqsAa5zzq0blohF4kxp1MwPyZwkF+eEX6faLWQgqrTanogkqD6TZOdcELgVeBTYBtzrnNtiZl82s+uGO0CReBddJUvWgXtwYvCekmQZiEq/FhIRkcTUr/VznXOPAI902fb5Ho69ePBhiSSOk5LkJB24B+Hea1CSLP3nnKPKHzjp2xYRkUTRryRZRHoWnSQn68A9gNctnkiazzQAS/qtMRCkpb1DlWQRSUhKkkUGKbpKVpiVvJXkReUFLCoviHUYkkBOLCSiJFlEEo+SZJFBKshKC1dYU32dcwmLSPSS1Bq4JyKJR0myyCClpBgluRmk+izWoYjElUq/VtsTkcSlJFlkCJTkZmDKkUVOUtnQCqjdQkQSk5JkkSHw3vOnEXIu1mGIxJXSvAzOm1msGVFEJCEpSRYZAtcvLYt1CCKdzOxK4HuAD/i5c+7rXfb/B/A+IAhUAe9xzu0b6jhWLSlj1RL9vyEiiUmjjEREkoiZ+YDbgauA+cCNZja/y2EvAcudc4uB+4FvjmyUIiLxT0myiEhyWQHscs5VOOfagHuAVdEHOOeecM41e3fXAOUjHKOISNxTkiwiklzKgANR9w9623ryXuBvwxqRiEgCUk+yiMgoZWbvAJYDF/Ww/xbgFoDJkyePYGQiIrGnSrKISHI5BEyKul/ubTuJmV0G/DdwnXMu0N0TOefucM4td84tLy0tHZZgRUTilZJkEZHkshaYZWbTzCwduAF4MPoAM1sK/JRwglwZgxhFROKekmQRkSTinAsCtwKPAtuAe51zW8zsy2Z2nXfYt4Bc4D4ze9nMHuzh6URERi31JIuIJBnn3CPAI122fT7q9mUjHpSISIJRJVlEREREpAslySIiIiIiXShJFhERERHpQkmyiIiIiEgXSpJFRERERLow51xsTmxWBezr47ASoHoEwhkuij92Ejl2UPyx1N/YpzjnRs0KG/28ZsPo+N3HK8UfO4kcO4yO+Ad8zY5ZktwfZrbOObc81nGcLsUfO4kcOyj+WErk2ONBIr9/iRw7KP5YSuTYQfH3RO0WIiIiIiJdKEkWEREREeki3pPkO2IdwCAp/thJ5NhB8cdSIsceDxL5/Uvk2EHxx1Iixw6Kv1tx3ZMsIiIiIhIL8V5JFhEREREZcXGbJJvZlWa2w8x2mdltsY4nwsz2mtkmM3vZzNZ524rM7B9mttP77xhvu5nZ973XsNHMlkU9z7u843ea2buGMd5fmlmlmW2O2jZk8ZrZmd77sct7rI1A/F80s0Pe7+BlM7s6at+nvVh2mNlro7Z3++/JzKaZ2Qve9j+YWfoQxj7JzJ4ws61mtsXMPuptT4j3v5f4E+X9zzSzF83sFS/+L/V2TjPL8O7v8vZPPd3XNRrF63thumbrmt3/2HXNPrFd12wA51zc/QA+YDcwHUgHXgHmxzouL7a9QEmXbd8EbvNu3wZ8w7t9NfA3wIBzgBe87UVAhfffMd7tMcMU74XAMmDzcMQLvOgda95jrxqB+L8IfLKbY+d7/1YygGnevyFfb/+egHuBG7zbPwE+OISxTwCWebfzgFe9GBPi/e8l/kR5/w3I9W6nAS9471W35wQ+BPzEu30D8IfTfV2j7See3wt0zdY1u/+x65qta/ZJP/FaSV4B7HLOVTjn2oB7gFUxjqk3q4Bfe7d/DVwftf03LmwNUGhmE4DXAv9wztU6544D/wCuHI7AnHNPA7XDEa+3L985t8aF/2X+Juq5hjP+nqwC7nHOBZxze4BdhP8tdfvvyfsEfwlwv/f46PdiKGI/4pzb4N32A9uAMhLk/e8l/p7E2/vvnHON3t0078f1cs7o38v9wKVejAN6XUMVf4JJtPdC12xds7uLXddsXbNPEq9JchlwIOr+QXr/RY8kBzxmZuvN7BZv2zjn3BHv9lFgnHe7p9cR69c3VPGWebe7bh8Jt3pfb/0y8tUXA4+/GKhzzgW7bB9y3tdASwl/Mk64979L/JAg77+Z+czsZaCS8B+q3b2cszNOb3+9F2O8/n8cT+L5vdA1W9fsAdM1W9dsiN8kOZ6d75xbBlwFfNjMLoze6X06TJgpQxItXs+PgRnAEuAI8O2YRtMHM8sF/gh8zDnXEL0vEd7/buJPmPffOdfhnFsClBOuIsyNbUQSA7pmx17CXDNA1+xYirdrdrwmyYeASVH3y71tMeecO+T9txL4M+Ff4jHvaxS8/1Z6h/f0OmL9+oYq3kPe7a7bh5Vz7pj3P1II+Bnh3wF9xNnd9hrCX4+ldtk+ZMwsjfDF6i7n3J+8zQnz/ncXfyK9/xHOuTrgCWBlL+fsjNPbX+DFGK//H8eTuH0vdM0+Zbuu2b3QNVvX7K6BxN0PkEq40X0aJ5qrF8RBXDlAXtTt5wn3pX2Lk5v6v+ndvoaTm/pf9LYXAXsIN/SP8W4XDWPcUzl5EMWQxcupgxCuHoH4J0Td/jjh3iOABZzcrF9BuFG/x39PwH2cPCDgQ0MYtxHuOftul+0J8f73En+ivP+lQKF3Owt4Bri2p3MCH+bkQSD3nu7rGm0/8fpeoGu2rtkDi1vXbF2zT45pqP/nGMI362rCIzN3A/8d63i8mKZ7b+orwJZIXIR7YB4HdgL/jPqfwYDbvdewCVge9VzvIdxMvgt49zDG/HvCX6+0E+6/ee9QxgssBzZ7j/kh3gI1wxz/b734NgIPdrkA/LcXyw6iRg339O/J+52+6L2u+4CMIYz9fMJfy20EXvZ+rk6U97+X+BPl/V8MvOTFuRn4fG/nBDK9+7u8/dNP93WNxp94fC/QNVvX7IHFrmt2bN//uLtma8U9EREREZEu4rUnWUREREQkZpQki4iIiIh0oSRZRERERKQLJckiIiIiIl0oSRYRERER6UJJssQFM2v0/jvVzN42xM/9mS73nx/K5xcRGW10zZbRQEmyxJupwIAuuFEr8fTkpAuuc+7cAcYkIiLdm4qu2ZKklCRLvPk6cIGZvWxmHzczn5l9y8zWmtlGM/s3ADO72MyeMbMHga3etgfMbL2ZbTGzW7xtXweyvOe7y9sWqYCY99ybzWyTmb016rmfNLP7zWy7md1lZhaD90JEJN7pmi1Jq69PcyIj7Tbgk865awG8C2e9c+4sM8sAnjOzx7xjlwELnXN7vPvvcc7VmlkWsNbM/uicu83MbnXOLenmXG8AlgBnACXeY5729i0lvLTlYeA54Dzg2aF+sSIiCU7XbElaqiRLvLsCuMnMXgZeILw86Cxv34tRF1uAfzezV4A1wKSo43pyPvB751yHc+4Y8BRwVtRzH3TOhQgv7Tl1CF6LiEiy0zVbkoYqyRLvDPiIc+7RkzaaXQw0dbl/GbDSOddsZk8SXtf9dAWibneg/1dERPpD12xJGqokS7zxA3lR9x8FPmhmaQBmNtvMcrp5XAFw3LvYzgXOidrXHnl8F88Ab/V66EqBC4EXh+RViIiMDrpmS9LSJy2JNxuBDu8ruDuB7xH+2myDNxCjCri+m8f9HfiAmW0DdhD++i7iDmCjmW1wzr09avufgZXAK4AD/ss5d9S7YIuISN90zZakZc65WMcgIiIiIhJX1G4hIiIiItKFkmQRERERkS6UJIuIiIiIdKEkWURERESkCyXJIiIiIiJdKEkWEREREelCSbKIiIiISBdKkkVEREREuvj/chiVBsTZn+8AAAAASUVORK5CYII=",
       "text/plain": [
        "<Figure size 864x432 with 2 Axes>"
       ]
@@ -793,7 +797,7 @@
    "outputs": [
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA8wAAAF1CAYAAAA9cmP6AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOz9ebhk51neC9+rqtZU87TnntXdUkuyBrcs2xLGdmyMzPiFBIKBixyGQEJICOfkOwfIAElIICeEOMQEMFdyDCbGB/hMTMBgwEZIYFuDbVmSNXRLrR73WFW75rVWTev7Y/f97reqd89D1d79/K6rL+1du2rVu1ZtL+/7vZ/nfowwDCEIgiAIgiAIgiAIwjCRcS9AEARBEARBEARBECYREcyCIAiCIAiCIAiCsAUimAVBEARBEARBEARhC0QwC4IgCIIgCIIgCMIWiGAWBEEQBEEQBEEQhC0QwSwIgiAIgiAIgiAIWyCC+TbGMIyvGobxrnGvQxAEQdjEMIyThmG89wqeFxqGcfAa3+OaXysIgiAItxMimG9jwjC8JwzDx8e9DkEQBEEQBEHYbtzozUfZzJxMRDALgiAIgiAIgiAIwhaIYL6NYdmfYRg/YxjG7xqG8VuGYTQMw3jBMIzDhmH8pGEYq4ZhnDEM433a677PMIyXzz/3hGEYPzxy3P/TMIwlwzAWDcP4QX23zDAM2zCMXzAM47RhGCuGYfyqYRjurT53QRCESccwjIcNw/i8YRjV8/fUDxmGYY087RvO34dLhmH8B8MwItrrv//8vXrdMIxPG4ax9xafgiAIwrbAMIwjhmE8fv5++1XDML7l/OOPG4bxg9rz/jfDMP7q/NdPnH/4K4ZhNA3D+DuGYbzLMIyzhmH81Pn78knDML5be/1VHe9mn7dwZYhgFsg3A/gogByALwP4NDZ+PxYA/GsAv6Y9dxXANwFIA/g+AP/JMIw3A4BhGI8B+N8BvBfAQQDvGnmfnwdwGMAD53++AOBf3oTzEQRB2O70Afw4gCKAtwN4D4AfGXnO3wTwEIA3A/hWAN8PAIZhfCuAnwLwbQCmADwJ4LdvyaoFQRC2EYZhmAD+F4A/BTAN4B8B+B+GYdx5qdeFYfi157+8PwzDZBiG/+/572excd9eAPB3AXz4cse6zPGEMSOCWSBPhmH46TAMewB+Fxt/YP18GIZdAB8HsM8wjCwAhGH4R2EYvh5u8JfYuMG84/xxvgPA/xOG4VfDMGwD+Bm+gWEYBoAfAvDjYRhWwjBsAPh3AL7z1pyiIAjC9iEMwy+GYfiFMAx7YRiexMbG5TtHnvbvz99PTwP4IIAPnH/87wP4uTAMXz5/X/93AB4Ql1kQBOEC3gYgiY2/ezthGH4WwB9i8356LfyLMAyD838n/xE2/j4WtikimAWyon3tASiFYdjXvgc2biYwDOP9hmF8wTCMimEYVQDfgI2dNACYB3BGO5b+9RSAOIAvni95qQL4k/OPC4IgCBrnW2P+0DCMZcMw6tgQvcWRp+n32FPYuAcDwF4A/1m711YAGNhwPARBEIRN5gGcCcNwoD12Ctd+v1wPw7A1cqz5iz1ZmHxEMAtXhWEYNoD/H4BfADAThmEWwKew8YcYACwB2KW9ZLf2dQkb4vueMAyz5/9lwjBM3vyVC4IgbDt+BcArAA6FYZjGRom1MfIc/R67B8Di+a/PAPhh7V6bDcPQDcPwczd91YIgCNuLRQC79QwIbNxPzwFoYcPsIbNXcLycYRiJkWPx3nwtxxPGjAhm4WqxANgA1gD0DMN4P4D3aT//HQDfdz48IQ7gX/AH53fufh0bPc/TAGAYxoJhGF9/y1YvCIKwfUgBqANoGoZxF4B/sMVz/r+GYeQMw9gN4McAsOftVwH8pGEY9wCAYRgZwzC+/VYsWhAEYZvxFIA2gP/TMAzTMIx3YSPb5+MAngPwbYZhxM8H2P7AyGtXABzY4pj/yjAMyzCMd2Aj9+d3zz9+rccTxogIZuGqON93/I+xIYzXAXwXgD/Qfv7HAH4JwF8AeA3AF87/KDj/3/+Lj58vMfxzAJcNQhAEQbgN+afYuMc2sLHZuFUAzCcBfBEbf4T9EYD/BgBhGP4+gH8P4OPn77UvAnj/zV+yIAjC9iIMww42BPL7sVEN+V8BfG8Yhq8A+E8AOtgQsr8B4H+MvPxnAPzG+fYX9ikvY+Nv5MXzz//754+FazyeMGaMMAzHvQZhB2MYxhFs/KFmnw+eEQRBEARBEIQdx3l3+rfCMNx1macK2whxmIUbjmEYf/P8vOUcNhyO/yViWRAEQRAEQRCE7YYIZuFm8MPYmNX8OjbmiG7VdycIgiAIgiAIgjDR3DTBbBjGY4ZhvGoYxmuGYfzEzXofYfIIw/Cx8+nX+TAM/2YYhkvjXpMg3I7IfVgQBGH8yL349iEMw8elHHvncVN6mA3DiAI4BuDrAJwF8AyAD4Rh+NINfzNBEAThAuQ+LAiCMH7kXiwI25+b5TA/DOC1MAxPnE+e+ziAb71J7yUIgiBciNyHBUEQxo/ciwVhmxO7ScddAHBG+/4sgLfqTzAM44cA/ND5b4/epHUIgnB5SmEYTo17EcIN57L3YWD4XhxF9Ggc6VuzOkEQFD5a6ISBMe51CDeFq/qbWO7DgjA+Gljf8m/imyWYL0sYhh8G8GEAMAxDZlsJwvg4Ne4FCONDvxenjXz4VuM9Y16RINx+PBV+ZtxLEMaI3IcFYTL48/D3tvyb+GaVZJ8DsFv7ftf5xwRBEIRbg9yHBUEQxo/ciwVhm3OzBPMzAA4ZhrHfMAwLwHcC+IOb9F6CIAjChch9WBAEYfzIvVgQtjk3pSQ7DMOeYRg/CuDTAKIA/nsYhl+9Ge8lCIIgXIjchwVBEMaP3IsFYftz03qYwzD8FIBP3azjC4IgCJdG7sOCIAjjR+7FgrC9uVkl2YIgCIIgCIIgCIKwrRHBLAiCIAiCIAiCIAhbIIJZEARBEARBEARBELZABLMgCIIgCIIgCIIgbIEIZkEQBEEQBEEQBEHYAhHMgiAIgiAIgiAIgrAFIpgFQRAEQRAEQRAEYQtEMAuCIAiCIAiCIAjCFohgFgRBEARBEARBEIQtEMEsCIIgCIIgCIIgCFsgglkQBEEQBEEQBEEQtkAEsyAIgiAIgiAIgiBsQWzcCxCEW4lhGDAMAwAQhiHCMBzzigRBEARBEARBmFREMAs7GgrkaDSKSCQC0zSHft7tdmEYBvr9vvrvYDAAABHTgiAIgiAIgnCbI4JZ2JFEIhHEYjFEo1HEYjHYto1YLIZIJKKEcCQSQb/fh2ma8DwPsVgMnU4H3W4Xg8EArVZrzGchCIIgCIIgCMI4EcEs7Aii0SgMw0AkElFOsuu6ME1TuceRSEQ5zmEYKkEdiURg2zZs24ZlWYhEIuh2u+j1eur4/X5/6HtBEARBEARBEHY+IpiFbY1hGLAsC47jIBKJDAngaDSKaDQKz/MQhiEGgwEMw0AsFkO/3wewUXbNsmzTNFU5diwWQyKRUI93Oh20Wi2EYYhOpzPOUxYEQRAEQRAE4RYhglnYtkQiEViWhWQyCdd1MRgMVMm1ZVkq1Et3mGOxGMIwRDQaBbApmB3HUSKaTnIstvE/j0QioZxnAKhWq+I2C4IgCIIgCMJtgIyVErYlkUgE8XgcyWQSjuMoAUxhTEFLx5hYlqVezx5mlnMPBgNVqt3v95XA7vV6CMMQpmkikUjAdV3E43H1noIgCIIgCIIg7EzEYRa2HQzxSiQSiMViKt263+8jDENVds10bJZpAxgSx+xlHhXMvV5PiWlgI0lbd6cdx4Fpmmg2m2g2mzKeShAEQRAEQRB2KOIwX4Rv+qZvwjvf+U7E4/FxL0XQiEajSCaTSCaTQ0KXLjD7i1lOrc9dpoDm151OB7ZtK7eZgpnuMo8ZhqEKDKMjzVLwVCqFVCql3kMQBEEQBEEQhJ2DCOYtKBaL+MZv/EZks1lxDicIwzDgOI5yeOkG8zOiaKWbTMcZgBK+fL4+k3kwGKh/fD1fx77oSCSCTqejnOxoNArHcZDL5ZDNZlWptyAIgiAIgiAIOwcpyd6Cb/7mb8axY8fw1FNPwfO8cS9nbLAMmTOMKTpZ9kxHt9vtqvLlmwnXQkEMYMgB1vuRKWxHn6f3NluWpRKy+Vqed6fTUa6zLqx7vZ76XnesWR4us5sFQRAEQRAEYecggnmE3bt348iRI/jEJz6BUqk07uXcUuje0lHlv2g0qnqCe70e+v0+bNtWCdOjLq/neaq390YyuhY6wgz6GgwGql+Z5dU60Wj0gsf5Wm4C0FHu9/sXHC8ajaLf7w+FgvF5lmXBNE30ej10Oh2pTBAEQRAEQRCEHYAI5hHe9a534XOf+xxefvnlHT86yHEcpFKpoR5ejmNi7y/d116vB8dxhkKx6L5Go1HYto1OpwPTNGGaphr5ZBgGfN9HvV6/rrXqvcimaSqhCmwGebHPmOXVemk1he9oMBgd5jAMEQQBbNseEtV66nYsFkOn01Hv2+/3EQSB6mnu9/tIpVJot9tot9vXdb6CIAiCIAiCIIwfEcwa+/btQyKRwOc+9znUarVxL+emYNs2ksnkkKgFoMqJR3t+KVK73a4Sw61WC4PBAP1+H51OR41e0ucd27YNy7JUGbVpmojFYqjX69dc5q6XQHONXB/Lryl+dceYzjCAofRrOuSmaarjUESz7JvXp9frIRqNqg2DVCoF3/eHSsJ7vZ4qX6fTLAiCIAiCIAjC9kUEs8bXfd3X4fXXX8fi4uK4l3LDsCwL6XQayWTygp+FYaiEL0Uke3kpDnXxGQQBLMsaEp0sUeZ4JrqvFI10hhOJhAraotAsl8tXdS66YNbFMh1lvVSbPxsVzvpxgiAYKuvWxW4sFlMbCvpro9GoEtjcIAA2NhQAKMfddV30er0L5kALgiAIgiAIgrB9EMF8nvvvvx/FYhGf//znt3XQl23bSKVSyhml6GMZMwBVVkxRrAs/fj3as6yXOjMsKxaLKZFMF7fX66lS7CAIAEC9liXPiURCiUrP81CpVK6o51cvEx/tVaYjPPo9z0EPKuO5ep6nxk+FYQjbttU1YOl2LBZDu91Wgnn03FgGzvfk9bAsC/F4fMi1FwRBEARBEARheyGC+TwPP/wwnn/+eZw4cWLcS7lq9NnErusOOcDdbhfdblc5nRSUdJdHS7D5GgpmvXyZZcfRaFSJUH7NY48eS++D5rFs20av10MikYBlWbAsS/U5XyptmyKYfcejAlg/R10w63OUCfuwTdNU7jTd9iAI0O/3h1ziwWAAx3FUyBmFOEUyNwO4wUBh3e/3t/UGjCAIgiAIgiDczohgBvDe974XlmXh5Zdf3lZhTa7rolAoqLJhAEP9s1vNE9ZLqFl+TdFIAcyUZ5Y283WjZcgUoQzg4jgmuqx0gClGKWL5uiAIEIlE4LouUqkUCoUC1tbWtuwfp0Cm+NfLvwnXy/MAMBT6pa+JYpkbALxW+pzmSCQCz/PUMSmG+fPBYADbttX5MyGc70eXmb3egiAIgiAIgiBsL0QwA3jooYfwyiuv4OzZs+NeyhXhOA6KxSJisZjqteU8ZApm3e3VGR2dxPFIANR8YYpNOql0kUdTqXUBrQtJvg8FKd+DCdw87qgT7DgOZmdnkcvl1HM6nQ7q9boS3nqfsn6OPA7Fsp6E3ev1VKk430sfRcXX6a657mDzeAAQBAESiYQ6tu66DwYDdLtdxONx9XrTNJFIJC6YCy0IgiAIgiAIwuRz2wvmxx57DM1mEy+88MLEu4AUlHQ/e72eEslk1HXVS6MBDPX98p9lWeh0Okog8njsS6b4pPDVk6H5OiZLsxRZn83MJOpR8Urhy7nGvu/DcRwkk0k4jqPc23q9jna7PRTmpbu6o+XZdIbj8bhyt/UQMDrcfF/9eGEYwjRNdLtddDodVWbN4zAQTT8m338wGCCdTqvz58/j8Ti63S6azeZN+b0QBEEQBEEQBOHmcFsLZtd1sX//fhw/fhynTp0a93IuSjQaRT6fRzqdVmKOpcm6yKXrqvcoA5t9vPrYJ/Yd60KYj/Mxy7KUsxyJRFQSNPt2+Xr9eHqpNJ1XYDNNG9gQ0I7jDCVSd7td1T9MZ9lxHGSzWbiuiyAI0G634fu+WhPPC9jcKKAYNk0TYRgOBZzpCeD6JgPPX3fWGQKml3T7vq9ceH3TgNeWs6j5mO5Uu66r1i4IgiAIgiAIwvbgthbM3/md34l+v48vfelLEytkLMvCnj17EIvF4Pu+En0Uqrpg1PuO9Z5mPRF7MBioecp6mbDuOHe7XTViiaXbdG/pmlJcUpjqAp1ikWOoKCi5dpYuUzBTLAMYepzznOnatttt5Tbz5zym7mhzDdxA0BO0eU30sVl8P9d11flHo1F0u101S9owjCH3nGOogA3BzevQ7Xbhui4Gg4ESz/wMBEEQBEEQBEHYXkTGvYBxMTc3h0KhgFOnTqFUKo17OVtCsawLM6Y082tgOJ2awpSimI4r/6sLN7rKFJa6I6q7zRSXfD4AtcHAPmW+l14mrSdT62LesqwhEc7zYAAZj99sNrG+vq7GP+VyOeRyOaTTaWSzWcTjcXVt9J5hwvLqMAzVGCg6yxToFLS8dpFIBI7jwDRNOI6jvo/FYmqTQT8HXiNuXoyWaXPzIBaLIR6P39DfD0EQBEEQBEEQbi63rcP8tV/7tXj99dfx+OOPT+Sc3Hg8jl27dl0wR5hsFdrFsmPdCR49N5Yp66XTegI1RSZFN0ux6eTqjrY+xonl2yyx1pOnKWLp2AJQz+PjrutesM5ut4t6va6E+tTUFPL5vBLJ7J/W3WbP85SrzTLr0VAynid/po/Forh2HGcoJVw/V/Zd6yKbx9UFuuM4QwFktm2r2dOCIAiCIAiCIEw+t6Vg3rdvH9LpNF544QXlPE4SjuNgfn4eg8FA9QFTgOoil2XDehAXxZk+JgoYnlGs/5zHYlkynWCOl6KY5vvrDqoelEVh3Ov1VFgXhW0QBENl2xTUnU5nKPVaF91ca7fbRaPRUGOgkskkEomEOl6j0VAJ4QCGNhb0knUAQ33L3DTQA870nwNQc5h5/p7nDY20Yn+0bdtot9twXReJRAKNRkNtALAsndeDI6YkMVsQBEEQBEEQJp/bUjA//PDDKJVKePbZZ8e9lCEMw0ChUEA+n1ciU5+BzNFHepk1RTTdUgpGus0UjRR2wPC4KJYV27YN0zTheZ4qRdZdYApjfQ6zvgbf91XJMV/H/mTC57LMmQFfdGUpwHXoGus91AxAo/gOggC+76PRaCghys0E0zQvEPzcGGAqut5jzDXwuLZto9VqqevA9fL5XLtlWYhEIirNG8DQOfHz4fNEMAuCIAiCIAjC5HPbCeY777wTruvi2LFjExf0FY1GMT8/D9/3AUAlRVMo68JOF8SDwWAoAZvijOXDesk1U7YJhS2wOXKp0+moWcL685hezfArvQQ5Ho+r43iep9xZvTRbd5kpSinER0Wk7l6zZ5ti1TAMxONxFAoFuK6rgsC63S6CIFBBXyxV5/XSx0iZpqkEsx4Mps9WZr84PwuWpTuOg2azqcS3aZrq+M1mU11jCmS9lJ090/pnIAiCIAiCIAjCZHLbCebDhw+jWq3i+eefH/dShohGozh48CB6vR48z1NCUx+FpDu7dInpnuqiTA/qorvJ57OHmKXcAGDb9lBAlR6IxbUxOVtPutbnKfM5uhPLnmXHcdQmAMUvy7L199HRQ7Momj3PQ6lUQrfbRT6fx9TUFNLpNCzLUqXZwGZqNfuWgc1ybM6K1oUuU7F931el5rxedPTpSMfjcXUuLEOncx0EwdD58zlEF9xSli0IgiAIgiAIk891CWbDME4CaADoA+iFYfiQYRh5AP8vgH0ATgL4jjAM169vmTeGt771rZiZmcETTzwxUWN+IpEIDh06NFTCTNd1dPwThR4AJYrpxtJVpdNMR1UPt+L8ZopQPSCMaxl9DNic5cwxSxTLFJNBEChnlevjOgAMlZVzZBVFs15mrot8fq2XlXueh16vhyAIlCOfSqVUTzSwsRnAEVwUzbwGpmmi1Wopcc+xUrrI1gPReM4cEaUHjOnjqejk83Piueip4txYiMfjaLfbIpiFG8Z2uxcLgiDsNOQ+LAg7lxsxVurdYRg+EIbhQ+e//wkAnwnD8BCAz5z/fiKYmppCrVbD8ePHx70UBfuWLcsamt1Lh5luM8uudaeZZb8UZzweABWepZcp9/t9BEGgenj15+ghYL1eTwVa6bOLe72eOg77ivVQK30WNHuOKUp1QcsEbL6vviEwKvxHHwMA3/dRLpdx5swZLC0twfd9uK6L6elpFItF5PN5pFKpoZFTdLPprNMZ1oPU+F765oBpmgiCAJ7nqefqYpz/bNtW56L3SjPkazStnE6/INxAts29WBAEYYci92FB2IHcjDnM3wrgN85//RsA/j834T2umocffhiZTAZ//dd/PVFjpAzDwJ49exAEgRJyFKpMjh4MBkrgjY6XAqBKi1laDUC5sJ1ORwlVfc6y7tzqs50ponWhx5Jlilr2M48mco/OI6YTqwttriMIApVQzpnRuoilG8webrq1FPrdbhfr6+tKNHe7XcTjceRyOWQyGaTTaaRSKViWpcrG6cazt9pxHABQY6Yodvl8/TroKdrs29bTy0dHfNFx5/xm9obz+huGccEoLUG4wUzkvVgQBOE2Qu7DgrADuN4e5hDAnxqGEQL4tTAMPwxgJgzDpfM/XwYws9ULDcP4IQA/dJ3vf0WYpol8Po92u43FxcVb8ZZXhGEYmJubU26rLjb10mxgs6RX78XVYWmzXuar9wDTne73+0PONEOuWAZNMUuRThGo90DrSdn6qCiKWr032vd9JfpZssxyajrqXEen01Elz4Tr0h+jQO10OiiVSkoAF4tFpNNpJJNJ2LatxD5d4l6vB9/31fnyevE5dKF5bMdxhhxkvg9Du+jUs8Sc1yUajQ5tAPAcbdtW721ZlgR/CTeSG3IvdhC/FWsVBEHYich9WBB2KNcrmL8mDMNzhmFMA/gzwzBe0X8YhmF4/sZxAedvJB8GgIs950bxyCOPYHZ2Fr/zO79zM9/mmpiamkKz2VRBVPqoJoo827aVuGKyNJ+njywCMCRA9VJuvoaiWe8vplDUBTb/ARsl0BSBdGjpAlMss2Rbd5kpSl3XVWKc6wM2NwH05OpIJKLOlevnOkdLpyl4G40Gzpw5A9/3cccddyCdTqvjDAYD1Go1mKaJarU6JN7Zf0zBq193lljrY7rotNPhjsViiMfjiMVialY04aYHz5sVAxylxeAvz/Nu2u+WcFtxQ+7FaSM/OeU3giAI2wu5DwvCDuW6BHMYhufO/3fVMIzfB/AwgBXDMObCMFwyDGMOwOoNWOc1k06nkcvlsL6+jna7Pc6lXECxWFROZafTGSpJZqgWRR/FaqfTUWJYD/fi+KNR4c3H9bRtln3rZcYUwZzXHIvF4DgOBoOBclopjPv9PnzfV2XJnN9Mkaofl44010B3mGFbdLrp1lLMcg4yS8RjsZgSsRT9fO9er4dSqQTf95FIJGBZFlzXxczMjLpOnucp0dpsNhGLxdBut9WGxFbBaUy+dhxn6Dr2+301q9q2beWYdzodde08zxvq39ZHfFFYb9VDLQjXwna4FwuCIOxk5D4sCDuXaxbMhmEkAETCMGyc//p9AP41gD8A8HcB/Pz5/37yRiz0Wjl69Cimpqbw67/+6+NcxgVEIhHcfffdKJfLSjwCUCKL7i0A5cLq7jDFZhAEQ8nMFNBEd21159a2bViWhXQ6rZxUik8Ku1QqpRKnWWKs9/sOBgOUSqUhQcnRSpzXrJeQs09bTwDXQ84omjnTWS/vHnW2ec7saQaAdruNkydPIhqNYnZ2FrZtI5vNAgAqlQrCMESr1UIymYTneej3+6rHmZsMfK94PI56vT7Ui82NBv1662Fitm0r4cyNA2500InXe9Rd10W73VY95oJwLWyXe7EgCMJORe7DgrCzuR6HeQbA758XGTEAHwvD8E8Mw3gGwO8YhvEDAE4B+I7rX+a1MTs7i5mZGZw6dWpcS7gkHFdEcaw7ybpA1p1ZCmC6uCzz1VOlgeHyZYroSCSCZDKJfD6PXC6HVCqlnqO713SZ6SR3u100m031/uwpdhwHU1NTaLVaQ2XJnGesj2zqdDpwXXfIyQUwFGKmB3tRiHJteom4XlKul6b3+33U63WcOXMG/X4f09PTiMfjmJ6eRjabxdmzZ1UZN/un9T5qYhgG2u32kIOtfz6u66rXUBwzCEy/3vrGhd7DzQRuPiYI18nE34sFQRB2OHIfFoQdzDUL5jAMTwC4f4vHywDecz2LulHccccdSKVS+PjHPz7upVxAJBLB+vq6CqHSg6QorCjuKLzo7AJQQpkiloFU+ngo3bmORqNIJpPYu3cvFhYWkEwmEY1GUavV4HkePM9DKpVCJpNR5dUsia5UKlhZWVGilaI6m82qNdNZ1v8lEgm0Wi31/hTaukjm43rJuN6XrZeQ8/zpsOuuM89zMBigXq/D8zwEQYBdu3YhkUioczcMA41GAwDQarVUmJfuorOvmBsTnuepALPR8m1+BhS/3ARhBQAdeoaQ8fPkZ+S6rkoLF4RrYTvciwVBEHYych8WhJ3N9YZ+TSy7d+/Gnj178MILL4x7KRdgGAbuu+8+VY6sh2/RLdaDuIie2gxAiUaKWJYm62OY2G+bzWaxsLCA+fl5JBIJABt9y41GA6VSacgBBoBGo4HFxUWsr6+j0+nA9301ZoriUR9HRSE52u+r9yBTEOuiG9hMvaZwZvkyoevOcnC6uTx3XiM6zpybzMqC+fl5WJaFMAwxNzeHTCYz5Dazt53vT/c4l8spF5rl23TQKegplJl8zVJ0rosjqvgZszda3/gQBEEQBEEQBGEy2bGCmcLoC1/4wriXsiWO42BxcXFojJQurOhCshSYPbJ0XfkYnU6KR6KPocpkMio9mmFavu9jfX0di4uLWF1dRTKZRBAEWFpawrlz5+D7PhqNhnJeAai18Pi+76v3owDUNwEoRDlLmeXP7F9m3zKwmdRNB1kvgeb3uujU/8uSbL2/ORqNwvd9nD17VpWP04nO5/Oq7LvZbKpjsf+YvdUAVHm5PtKL58Sea4ri0RAz/pzl6RTZfCwWiykRLsJZEARBEARBECaPHSmY9+zZg0OHDuGv/uqvxr2ULXnwwQfR6XSUONTnEFNY6j28TJqmM0snmo/Rkd5qvrJpmpiamkKhUFDCvN/v4/Tp0zh37hyazSYGgwF838fS0hIajQbW19cBbIp23TWluKPLy8f1/l2mc28lBJlYzeAxpl+HYajK0/ne+lxpzqkencfM6zXaw83XtFottQGwsLCAbDaLSCSCXC6nyqN5TIr3Xq+nRDPPD4Aqz+a15nXQ+9C5ecFz5QYFNwgoxOPxuOp/FrEsCIIgCIIgCJPJjhPMhmEoJ/XFF18c93IuwDAMFItFnD59GsBm2bIu/ChAOa+Xj/d6PeVsBkGgyoFZpjw6tzgajSqxTJHXarWwurqKU6dOKXcV2BCXa2tr8DxPlVBTCNLZBjZTt+Px+JDgZBK3XqZMAUrxyzFMeiAWS6v13mCWbnNjgCJ/NOSLQpPClkI1Eomg1WqpnulKpYJ2u41ut4vdu3ejWCxiMBggHo8PnRNDy+hQsxSbPdTcpNBL4fU5zXS1ee25Nm5o6KOlWFovYlkQBEEQBEEQJpcdJ5jvuOMOPPLII/jDP/zDcS9lS+bm5hAEAdrt9lACM0WVPpuXopOClYKLAV8saaarTNHN18TjcczOziKVSqFer+Ps2bNYX19Hs9lEu92G7/uqnLjdbisRSGdXL4d2HAe5XA6JRAKpVAqu66pe3jAMEQQBms3mkEtLkas735ylDGyWNfO/tm0r4a9vBHBeMt1uBoDpvcAUvdxM0Gcd0xk+e/ascngdx1FBaPpIK9/3EYvFUK/XhzYv+Bmx7Fo/tv4cbg7ofdtcJz8/x3HUddXHdQmCIAiCIAiCMFnsKMFM57Pb7WJxcXHcy7kAwzBw6NAhNYaJglIfp6SPTKLI0mckM2hLL4GmWAOg3NBYLIZsNotisQjTNLG6uopyuYx6vY5Op6OEdrfbVQFcfE+6xolEAo7jwDRNzM7OIpPJIJ1OIx6Pw7Is+L4Pz/NUv3M2m1Wl5p7noV6vK/HIecfpdBqWZanjUlSyxJvnRuFLV7zdbiuXmO+hXwNgwyVnybQeKsbrFoYhVldXMRgMMDU1henpaTiOA8dxkMlkVHBXu92GYRhotVrq+uul1XwvfoYM9+L78fPUNzIYGsZzpuPOTRJBEARBEARBECaPHSWYd+/ejXvuuQef+MQnxr2ULZmfn1eBW3RGKZoJhSvdTDqi/BmFH78HoAKl9BAt0zSRy+WGAqwAqK/1EU4UtJy/HIvFUCgUMDc3h1wuB9u2USgUVLk0RSPXmU6nVXI2z4/903R7k8kk4vG4ei1DuOLxOCKRCFzXVWXZvC7AhjANggCNRgO1Wg3NZhPr6+tKqHNzIRKJqMAuOvQU3hSv/LpSqajPYG5uTon5XC6HdrsN27axuroK27bVNdPTsFlazhAz13XVteGagM0gMb7v6KYAhb+UZQuCIAiCIAjCZLJjBHM8Hsfc3ByWlpZQq9XGvZwLiEQiuOuuu3Du3DkAUD3F7HWlIKMjyZJjPaFaF1d6qrRewsze4OnpaRSLRViWpUrAfd8fmtNMsU6X2nVd5HI55PN5zM7OolAoDIljfc4z10ZXFgBSqRSazSaCIEAikVBiPB6PI5lMIhaLqRFReu8ve5r1sVMU7hwnlclkMDU1hV6vh3a7jUajgbW1Nayvr6PdbqPT6SjBrc+r5nnqo7ZYPs7vU6mUum62bSMajSoXnT3mlmUp4avPXaYAprjmefG68rm8VqMuNM9fSrIFQRAEQRAEYfLYMYJ5amoKhw8fxm/+5m+OeylbMj8/j0ajocScLuw4/zcSiahZvvqsZWDTTdZTmNlPSyFGEZ1IJFS/MUuLK5UKPM9TYlkX56ZpIp1OY35+HgsLC0in0wA2y7vp2Nq2PZRcHYvFkE6nEQQB4vE4PM9DsVhEo9GA53lKcHJ8Et+Xx6L7zesRj8eVsOW56OtlX3Y2m0W328XMzIxynldXV+H7PlqtlkqepphlnzaTtnnt2dedyWRUeTYFbCqVUj3NjuOocm9eL/Zl69dJD/TSA8p4jrZtDwlpbh4IgiAIgiAIgjCZ7AjBbNs2MpkMTpw4Me6lbInrujhw4ABWV1cRBIESTnRPKbYoniigdfeZ5b90kwkFnj6fmcFcnU4HjUYDy8vLqNfryu0cHQ0Vj8exsLCAAwcOIJFIqPAsAMrl5TooPPWQrV6vh2azqdbcbDbVWvkcCl4ASjxT1PJ4FON0mvXz1IU6AJUgblkWUqkUisUims0mKpUKarUaqtXq0MYDg9L0zYper4dqtapSuOfm5hCPxzEYDJBKpdBoNJDJZNDv91GtVocCy9g3ro+d0sW+/tlRLPO89bWMjsISBEEQBEEQBGFy2BGCOZPJ4PDhw/i93/u9cS9lS971rndhZWUFnucBAIIgUK6nnsDc6/VUaTHHNunlxHSb9VLgaDSqjsU+ZAZzhWGIZrMJz/OGSqkBqD5c13UxNTWFhYUFVTbN9yAsl9bLqLmmSqWCVquFIAjgeZ5yabvdrkqc1ucWsweaQpIBWBT7PH4QBGqOMddiWZYS4aZpwvM8db3oqmcyGeU4l8tltNvtoXM3DANBEADYdO2r1aoaB0XRTJeYJenJZFIlX/u+rzYbWFbN49GFBzBUNbDVDGkGuEkPsyAIgiAIgiBMJtteMDMwii7fpOG6rhKPdFopsIDN8C19FjP7WXXxSjdSF36xWEyVdbPsOZPJoFgswjAM+L4/JMooUg3DQDKZRLFYRDabxdTUlBKgqVRqKDSLIlCfqxwEAer1OgzDQLlcVmulgOT7MG2aQl4vAdf7lunKctwSXeZYLKbKymOxmHLmKXrZu8yNhX6/j0wmo66FZVmo1+tqvjLTsnldKVrp9FcqFQAbvdgMKKOzbds2ksnkBefItXY6HSX+WYoNQIW06Y45n8f3lf5lQRAEQRAEQZhMtr1gzmQyeMtb3oJjx46NeykXYJomjh49ivX19aExQsBwEBW/pwOru8kU03QhOfKJ4lV3mhOJhBojxRJsz/NQrVbRbrfVcU3TRCqVQjKZRDabVW4p+4v1kmImRA8GAyX62bOczWYxPz+vxHCz2VT/WKpNV1bvxWYydiQSgeM4Qy674zgqlbtQKKBYLGJqagrpdBqJREKJ2EqlgnK5jEajgXq9jlKppPqXbdtWx2k0GqhUKmr+tO7mclNAL5FniNjMzIwqo2ZwWr/fVz3NDAOje6/3a/Oc9dFXwGZlgP699DALgiAIgiAIwuSy7QVzMpnEoUOHJrIc+8CBAyo4CthMtqaoAjb7gjnXmG6qHu6lC1ndISV0OkfFdKfTQa1WU+nYLMGOx+PI5XKqX3cwGMB1XaTT6aGSbQrEbrcLx3EwOzuLYrGIXbt2YWZmBoPBAKVSCbVaDSsrKxgMBgiCQLnDdFktyxqai6yLZLrlLF1OpVLIZrOYnp7G/Pw8pqamUCwWkUwmh865UCjg0KFDCMNQvf/KygqWlpawtraGer0O0zSH+qH5/hwpRTdY33To9/uqHJxOs+u6cBwHANSoLFYO8Nh6mbculnkNGP5F+J4sVRcEQRAEQRAEYfLY1oKZfbgUiJNEMpnE/Pw8VlZW0Ol0hoKh6Gyyx1V3jynygiBQritFJbApnvWAKcuyEI/HVQgWRyA1m03lNNOBTqVScF0XwGaQFn9OJ5tOLIX13NwcHn74Ydxzzz0quArYEJ7pdBrtdhumaaqwr3q9roRqIpEYWjfdV753p9MZmktcq9XQaDTU+5umCdd14bruUCk7MQwD2WwWmUwG8/PzmJmZwdLSEs6cOYNyuQzTNJUjTFELbGwmUPDqmxB02iuVCoIgQDqdVs43Z0UDUAnXdN31z5VimEnY+rr5OerjvQRBEARBEARBmEy2tWCOx+M4cuQIvvSlL417KUNEo1Hcd999KpSKIoz9u7rAAqCEMZOn9V5mOph6//JogJdpmsjn88hms8qJrVarqNfrqofWdV2kUinVT22aJhKJhBK2i4uLQ+L50KFDePjhh/HOd74TuVxuy/N0HAeZTAbNZhMnT57Es88+i/X1ddRqNQRBAMdxLkgDdxwHhUIBg8EA6+vrKnCL52NZFnK5HKanpzE7O4tdu3bhrrvuwsGDB1EsFpFIJJBMJlXQGaEjPD8/D2Czt911XSwvL2N9fR2O4yihzPPkmoIgUOXc3IChuz4YDJDNZlU/NWc183NgqX0QBGpDgO45e7v13mxgs7dZEARBEARBEITJZVsL5mQyiXvuuQe/9Eu/NO6lDDE7O4tIJIJ6vT40a5mCbHTGMKF7SeFFcawHhY0mPkciEaTTaSwsLCCfzyMWi6mU6Gq1CmCzJLrVaqmkbtM0UalUlDPd6XRgWRb27t2LRx99FN/+7d+OmZmZLcce9ft9nD59GktLS3jiiSfwxBNP4IUXXkCz2US73R6aq5xMJmHbthL/LGVmqTpnG7N0ud/vo1aroVwu46tf/Sosy0I2m0WhUMCePXtw8OBBHD58GPPz80gmk6rE3HVdFSZGoa6nkDOdOx6Po9vtqqRyvmev11OJ2ACU89/r9bC+vq42MOLxOGzbHgoqY49zLBZDEATq82HfOkdxAcN9zBL4JQiCIAiCIAiTzbYVzJFIBIVCAYlEAu12e9zLGeLw4cNYW1tDt9sd6p3Vxykx3AvYOBe9JFhHf73eD8ugKbqqFMue52FtbQ2lUgme5ylR1+v1VGo1Xe5Wq4VOp4NEIoHp6Wm86U1vwnvf+158zdd8jRJ/7N31fR+lUglnzpzB4uIiPvnJT+Izn/kMFhcXLzh/CtNCoTA0qornyZ5f3VWnS0tXli48ADQaDTSbTZw5cwZPPfWUOv7MzAzm5uYwPz+Pffv2oVAoqPdqNptqHjPnPGcyGbVhQRHLEvR4PK7SvEeTrj3Pg2EYaLfbmJ2dVZ+f4zhIp9PwfV8lZfMzZEk9z1cvv2cYmh76JgiCIAiCIAjC5LFtBXMymcTb3vY2vPTSS+NeyhD5fB6DwUA5p3SEWZqrw5nCdEWZlK0nZ+vzj+k007mMxWIoFouYm5uDZVloNpsol8tYXV1Vqdj66CRgc/wT1xSPx7Fv3z48+OCDePTRR5HP53H8+HFEIhEVCraysoI33ngDX/rSl/DZz34Wr7766gXnwZAv0zSRy+VQKBRgmqZyWnu9HnzfV6KRwVoMHWMZMzcHGMDF86ewpvAvl8s4c+aMuhbpdBqFQgG5XE4FmtHN5bVKJpNwHEclaHuepwLZgA2Hn6XZ+rgufp7RaBSlUgmO4yCXyykxzPXrVQAcM0UBr58fHW0Ry4IgCIIgCIIw2WxrwfzWt74VP/3TPz3upQzx5je/GdVqVfXB0tHl10xj1h/Xy5EBqOfoKdIUuRRcdJdnZmaQSCQQBAGazSYqlQpqtZqat6wnMLPkOwxDOI6DVCqFubk53HvvvSgWi3juuedQr9dV+Fi73cbLL7+MZ599FouLixcIPNM0Yds2UqkUbNtWc5MZjhUEgQrF4n8pIPUSaLqv+jmOjs3qdDpDI7m4ycBjV6tVNJtNnDp1Sgl39m0vLCyokVQsT7dtW/V4MziNnxE3GOgM60FeTB0HNkrdWf5tWZb6HFmezXRyboLwPFiKPYlhdYIgCIIgCIIgbLItBXMkEkE2m4XrulhaWhr3chTRaBSpVArLy8tKXFJ0sYSaTiPDtiik9ARpPUiKPcQUbfxnWRZmZ2eRyWTUvOVWq4V6vT404ojrYv8wxXgul1P9v8ePH8ezzz6LUqmEdruNwWCAM2fODKVKE5ZM02VNp9NIJpNDAWV6/zDPVz8vHocl67qLDmyOyaKIpmjldWT6NGGKNdPBWW5eKpUQhiGWlpaQzWaxsLCAubk5Je7DMFQJ2ZlMBt1uF9VqFZFIBO12W4WR6fOUea3pEBeLRcTjcQBQQWQsvWfJN7/2fR+dTgdBEKDdbg+NmRIEQRAEQRAEYfLYloLZcRw8+OCDeP7558e9lCHm5ubQaDRUzyyFFkUmhbKekE0ByTJgPkcfJTU6r9h1XRQKBRSLRUSjUfi+j3a7jUajgXa7PZQCDWwI5mQyiXg8PhREtba2hmPHjmFpaUmVkF8K13VVSnU2m1Wjlpjybdu2+prlyBTGFKAUwZZlqfekuKbDTpE5umHATQX92ull3RSx8Xgc2WwWnU4H7XYb6+vreOONN3DixAns379fjcpiajhLyoEN0cv3ZS8yKwMcx0E0GkW73Uav10O5XFal6Bw/ZVkWut2uem+O2uIIrVarNVT2LQiCIAiCIAjC5LItBXMymcSjjz6KD37wg+NeyhCHDh3C2tqaKsnVxTKwOT6KCcvsL9bFMkUURSADsQAoNzaTyWB2dhau66o5xo1GA+VyWYVpUXBaloVkMol8Pg/HcVAqlbC6uqoc2MulNLOXOZlMIp1OI5fLIZlMwjRNBEGAbDarXF32DcdiMVU2bZomPM9TIpH9vd1uF47jqMRqHoNimw67Lor5PQWznkDO43KMVSwWg+u6mJqaQj6fx+rqKsrlMp5++mm4rot9+/ahWCziwIEDKpis3+9jfn5elVfr5fB6sBdDwYIgQK1WU89Jp9OqRJtp3NwsoFCmgy8IgiAIgiAIwuSzLQVzPB5HLpfDsWPHxr0UBftjOXuZUCDrTijFGMWd7kTrvbos59bLnF3XxfT0NLLZLIANZ7per6NSqcD3fVUSbBiG6lPO5XIwTRPLy8s4e/YsSqXSZftnI5GIem0+n8f09LQ6z2QyqUQfS8KZtj0YDOC6Lvr9vkrYZjkyS6C73a4acdVut5Vjq4+kouuru7D6ddgqMAzYLItmGXS320WtVgOwsdHS7/cRBIEKi1tdXcXevXuxf/9+NZeaYWirq6tYW1tDpVIZCunieUQiEfi+j0qlotxyutCu66pNkFarhfX1dXieJ2JZEIRbRjSXG34gHKBfrY1nMYIgCDucaC6H/vr6BY9dFLknbxu2nWC2bRtHjx7FuXPnxr2UIQ4dOqTKb9mXzJ7lXq831JPrui48z1MlyLowZvm1HoJFMW3btppJzPLnIAiwtraGcrmsHFaGWmWzWeRyOQwGA5w6dQonT568ohFc6XQaU1NTmJ2dxdTUFFzXVS5sGIZIJBLKxU2lUvB9H61WC4lEQpVE8zzpCLdaLXVOrVZLlZG3Wi1Uq1X4vg/bttFoNFQZtO6UA5vl6xSddORZjs3n8Zr7vo9yuaw2CCzLUkFlnEv9yiuv4PTp0zh37hz27t2Lw4cPo1gsYmFhAVNTUzh37hzeeOMNJY65ycG+cDrj+nzrXC6nNgsMw1Dn2Gq1buwvnSAItz3RQv7iPyyMCuZQ3ScV/b78wSYIgnCFRHM5IGJs/cNCDtHRn43eh3W0v3HlXjzZbDvBnEgk8P73vx+/9Vu/Ne6lDHHw4EGcPn0aYRgqMTuaCh2NRlVYFUU1BTFLkAFcMF6KfbzxeBzT09NIpVKIRCLwPE+5oHSXOT4pmUwilUqh0WjgzJkzWF5e3jLES8dxHMzMzGDfvn3Ys2cPLMtCPB5X4WCWZanz43uxNLtarSKRSKhe5n6/j0QioUR2KpVSo6J0R7lWq8E0TTSbTdi2PdTfrbvLvDZ6kJgulimmeT07nQ7K5TJWVlaUm+55HjzPU9UAuVxOudwvvvgizpw5g5WVFXzLt3wLpqen1Yxnlmivra2p9yf6Gsrlsvp88/m8St7O5/Mq8Kter1/rr5ggCLcx0UIeMLb4Iy2fvfKDGMaFf7wNBojGYkCvJ3+sCYIgjBDN5YBoZPOBXGbrezG5lEAeJRLZfP5ggCgg9+EJZdsJZtM0MTc3h89//vPjXsoQLNMdDAZDM4P1Emz2IHc6HVVOTOGnB1zxeBRjTKUuFovKXfZ9H2tra1hZWVHOp2VZyh1mSfGrr76KSqVy2VLgZDKJPXv24J577sGePXvgOA5M00Q0GkU8Hkc6nR5aK3urWYZMBxzYEN68DuxH5iYCACWo4/G4CgZzHAe1Wk2FZbGsnanYurvM68j16D9n8nYYhqjX68rF12GptuM4mJqaQqVSQavVQq1Ww7PPPgsA+MAHPoB7770XhUIBqVQKpmni+eefV5sgej8136PT6ajy7X6/r4LAEokEZmZm1CipK3H5BUEQSLSQ3xDGl/oj7VqJRIB8Fkavj6gRuaCcUBAE4XYlWshvCORI5PJPvl4iERj5HKKmBXQ7IpwnjG0lmKPRKPbv349SqXRZt/RWUigU0Gw2hwK7AChBpYdY0f2kkNaDwHRHWZ+9bBgGstks5ubm1OzhcrmM5eVlNTfZtm24rotisYhcLodSqYTjx4+jXC5fNo05lUrh0KFDOHz4MPbs2YNMJgPTNJFIJGAYhpphzHAyfWwUzzWZTKrH9LnG+nvToaZbzOfo84tZxq4LZv4jvI6jY6p4bblZwfL4i1GtVhGLxZDP5xGLxVSvM0Xzj/3Yj2FhYQEPPvigconZj83PiAIdwJB7zh7qTCajStfn5+cRBAHOnTt32bA1QRBub6KFPBA5X6qXTd0csawRxqIwcpkNt1n+WBME4Tbnlorl84SxKJBLw/ACQO7BE8W2EsyO4+Dd7343Hn/88XEvZYi77roLjUZDBXbpLizLdRnGxZFPFJ2E/cvApshk+W8qlcKuXbuQzWYRiURQrVaxurqKRqOhZlKzb9a2bZw9exbHjh27IrGczWZx11134dChQ5ibm0M2m4VlWchkMnAcB2EYqhCxyMhNg4KV5eYUrRSTPE+KQ/b/0oFnPzQA+L6PZrOpNhZGZxTrGwr6Y3q/t97vfDmxDGwI3HK5jD179mBmZgaRSATr592VZ599Fh/72MfwIz/yI0gkEnjggQfU63q9HqrVqvp89JFYDC0rl8tqMySTycC2baTTaUxPT8PzPKysrFxybYIg3L5EiwUgm76lf6gB8seaIAgCMB6xLEw22+o3wbZtvPnNb8af/MmfjHspikgkohxmusl6DzJLiDnfl4nSejI2v6YwpRCNRCJwHAdzc3OYnZ2FaZrwfR8rKyuo1WpwHAfz8/O44447cPDgQRQKBSwtLeH555/H6urqluXIOul0GgcPHlRl2HSnOWNZT5wGoPpzgeEebH30kz43mYKVmwf6bGhgM/XacRyVLk1nVhe7oy6yLsp5fGA4BOxK6fV6WF1dRSwWw8zMjCorB4C/+Iu/wJNPPol2u42lpSV8zdd8Db79278dd955J1KplNoEYQAY1whsCOdGo4FSqYRmszn0u8KRYIIgCDrRYgGx2ZmxiGVBEATh/IbluMWyGbt0oKNwy9lWDnM8Hkc8HsfS0tK4l6I4dOgQPM9TacksM6Y45tcAVG+rnn496oxScLFPN5lMDs1c5pijWCyGhYUFzM/Po1AoAABefPFFvPbaa1cULJXL5VQZ9q5du1AoFJDJZJQ4jUQiQ2XRdFL1UmiKfV2wMhBLT7imkB4tWQeGe7WDIEAQBKrkWUcXyLqz3O/3Ve+4LsZH3fBL0Ww24fs+0uk0isWiSmD3fR+//du/jXe84x3o9/tYXV3F137t16LZbKJareLUqVOqfzwWi6Hb7Sqn3TAM5URzLYlEAqlUCjMzM6hWq1haWrrASRcE4fYjWizAME2ECRehCGVBEISxMK7qngsIQ4SeP941CENsm/9ntiwLjzzyCE6cODHupQxx6NAhNBqNC3psR11HXSzTqaWLPFpWrIdpTU1NIZfLqVLsxcVFRCIR3HXXXTh69CjuvPNO5HI5VKtVvPbaa6qk+FIUCgXcf//9uO+++7CwsKBcZY6BAjAkQLlunovuivd6PTWTmCKZj/O1FIUs3+bzgA3x63keWq0WGo0G2u22OgZLnVn6zF5w9jSPimrTNGGaplr76M8vRavVQhiGSCaTQ2L7zJkz+MVf/EXccccdaDQaWF1dxfve9z488sgjyGaz6rNj6Jou2vWxX6urq+h0OjBNE6lUCrOzs0in01e8PkEQdib8Ay1MJcb/RxoAo9vDoFwZ9zIEQRBuKRMjlgGg18dAAmInign4rbgyLMvCO9/5zonrX2aZNIUu+5WJnpZNIaj3L+vlzACU6GLQ18LCggr6qlQqiEajOHLkCB566CEcOHAAuVwO9XodX/7yl3Hq1KnLliMXCgXcd999uP/++1UZdjKZVCnWdEv5NYW/LpqBYeFMwai76oT9y3rJtu5U9/t9NJtNlEolrK+vqwRpnsdocjgAJZz1r/UeajrQF8wbvQSNRkO9bmpqauh9n3zySbzxxhu48847cfLkSQRBgO/6ru/C3XffjXg8rs5bX4tOu91GtVpFpVJBu91WQWPFYhHxePyK1ygIws5iov5AI/KHmiAItxkTeS8WJopt85sRjUaxa9cufPaznx33UhTJZBLtdluV5XLkkGVZQ+OQKBYpJvUZy8Bm+TUFn2maKvE6lUqh3++r0UeHDx/GI488gj179iCZTKLf7+OFF17Aiy++CM/zLrnemZkZvPnNb8Z9992H3bt3o1gsIpvNqr5dvTxaL4HmGvUeZt3x1R3j0XFZHDfF8+N14H9rtRrOnj2Ls2fPolqtqs0FvbSZ6dcU7XpPM/u+WZqtJ3nn8/mhNV8K/fPJ5XLYs2eP6uXu9/v46Ec/imw2i4MHD+IrX/kKdu3ahb/9t/829u/fr9LEudnA9eifK0dONRoNGIaBeDyOfD6PZDJ5ResTBGHnEC3kEdu1MFF/oBndHsJzyxislca9FEEQhFtCtFiYyHux3Icnj23Rw2wYBjKZDNrt9hWVHN8qFhYWEATBULgVgCH3mCKaIs40TfUaupEUdRSZpmmq5GvbtpU7eeDAATz22GMoFAoqSfqJJ57AE088gVarddm1Pvjgg7j77rsxMzOjjq2LU5aCs4RaT5xmvzDdbjrp0WhUhX9RLPr+Rt+FbdtDvczsZ2ap8vr6Ol577TWcOHFClSzzuvE6bjV+6WJut17KHYlEkEwm1XzlKwkCYz+2bduqd7zb7aJSqeCrX/0qPvaxj+E7vuM70Gw28dprr+G+++7De9/7Xvzpn/4p2u32BUFl+voHgwF838f6+jocx4HruqqnuVarTdSYNEEQbi6G4yCMO+NexjBhiMFl/n9EEARhp6DaYSZEKAMbYrm/uIxQ/iacOLaFYI7FYnjHO96BV199ddxLGWLXrl2o1Wqqj1fvn2UAWL/fh+/7yrm1bRu2bQ+VblOo0ZGOx+PI5XJIJpPodrsol8tIJBL41m/9Vtxzzz2oVCqwbRsvvvgiHn/8cVSr1Yuu0TAM7NmzBw8//DDuvfde5PN5xONxlfJMkTpaxqyXROv/uAnAjQC953kwGMCyLLRaLeWw2ratQrw4VisIAiwtLeH48eM4efIk1tbWhsZW6b3LFPF0f0dD0vg+TCfnc3u9HhzHwfT0NFqt1mXDtdhPzPN0HAfxeFyNwfJ9Hx/96EcxOzuLRx99FM888wzCMMTb3vY2LC4uYn19HdVqVZXnm6ap3HKeQ7fbRa1Wg2EYmJmZQTwex9TUFOr1OtbW1q7lV1AQhG1EtJCHEY8jtK+s8uVmY/T6GCwuAwAGlxnDJwiCsFOYtBJs3osHYShieULZFoLZNE28853vxG/+5m+OeylDWJY1VCpMIReLxZRjqM8m7na7sG17KOUZ2JxPzPFT09PTyOVyiEajaLVaiEQieOc734l3vetdaDQaSKVSWFtbw6c+9Sm88sorl1zjnj178Pa3vx333HOPGk2lz1XWy6lH2WpslN7TrL+eCdEM/KpWq6jX63BdF67rIggC9Ho9lEolLC0t4fTp01haWkK9XlcCkwnb3GCgI83RUxSgwOaoKgBqLXoYGOc/x2IxOI5zwagqHYp5PeWbnwvL64MgwBtvvIFf/dVfxfT0NO6991688sorCMMQe/bswd69e9Hr9dTa+bnrvc3AhlvfaDRUSXYqlUI8Hh9KERcEYWdimCZC1778E28UYYjwzCKMPQsX/fnAlyRWQRBuH6KF/MSIZRHK24dtIZhjsRgOHz6MZ599dtxLUezbtw9hGKr+Zc4s7vf7yjGmg0wRpjuzfJ4urOLxOKanp1EsFlEsFhEEAarVKu69915893d/N0zTRLvdRjabxZNPPolPfepTl3ROc7kc7r33XrzpTW/C/Py8En+xWEylVY+Ku9FQLq4/FovB8zwlbtmHTEdWL9/udDpYWVmB53lK7IZhiHK5jJWVFVSrVTW3utPpwLKsIVdZv2a8lgwiAzbTxflzPeiLUOT2+30Ui0V4nrdleXckEkGxWESv10On00E8HlcbBdxYoNvseR5efvll/NIv/RL++T//55iensbx48eRz+cxPz+vZi63tcAc/m7w/CKRCIIgUJsJtm0jk8nAdd2h1wmCsLOIFvIIk7c+5G/g+zBOnb3ID2WTThCE24doIX/rZyyHIQanz239s0GIsCvjRbcD20Iw05X1J2gnfP/+/QiCYCjsir2pLEHW5zDT4dTLiwEMCU7LspBKpZDJZBCLxVAulzE1NYXv//7vRzKZxNraGlKpFEqlEn7/938fjUbjouvLZDK45557cN9992FmZkb1F+vOLJ1PusMUc/yaDm0sFlO91nqJtH6swWCAXq+HIAhQLpextLSEZrOpZiRT/PPa6NeB4p3H1gO/9McZlqW/JwX2qBve6XSGZkLPzs6iVCoNBaOl02kkk0mYpqk2HlhKz55s0zThOA5831fPe/LJJ/GhD30If+tv/S0VFpbNZpHNZlEqldTvKTcKAAw53GEYot1uo1QqoVgsKqdZBLMg7EzG8keahjgXgiAIAGKxsdyH5R68/Zl4wRyLxfDud78bx48fH/dShigUClheXlZlwnp6sz6PuN/vw7IsVaqru6jsj6WQjcViSKfTcBwHtVoNjuPgG77hG7Bv3z74vg/P85DL5fDEE0/gySefvOjaXNdVc5r379+PZDI51JOsCzd+z7nJDCjr9XqqxJrClsKVM6XpKpumCc/zYBgG6vU6yuUyarUa2u02Go0Gms0mOp3O0Cxn27bVebO/lwLccZwLysS5Hgak0a0FoMq2Wf5Mwcx+a2Bj02V+fl69lqFpFNjc4KBrThednw/xfR/dbhd/9md/hnw+j3vvvRftdhv9fh+pVAqpVAq+7yt3m8e3bVsdH9gszbZtW/Uyt1otNJvNq/9lFARhshnTH2mDkxdxlgVBEG4zooU8kErc8veV+/DO4LL/D24Yxn83DGPVMIwXtcfyhmH8mWEYx8//N3f+ccMwjF8yDOM1wzCeNwzjzde7wGg0ikcffRSf+cxnrvdQN5ROp4Nms6lEG7Dpsupzgemwcr6yaZpKGNKxpYObTqeRzWbV8Q4dOoTHHnsM/X4f1WoVU1NTOHv2LH73d3/3om57NBrF3r178dBDD+HgwYNIpVLq/Vhyrff6sicZ2HSc2U9M15vnpidhU1gzVTsMQ6yvr6NUKmF5eVnNHG6326p/WS+b5nWybVsdi8Lcsiw1G5ql2Hq6No/Fc+DPRt1orldfP9+/2+2i3+8rYcuSej2ITf+sWZbNz8zzPDz99NN4/fXXhxxl9ogDm5sTkUgEnU4HvV5Pff4sR6/Vauh0Oshms8hkMtf3SynsaMZ9LxaujXH9kQZASv0E4QYj9+HtyTirfOQ+vDO4kt+cjwB4bOSxnwDwmTAMDwH4zPnvAeD9AA6d//dDAH7luhcYiWD37t345Cc/eb2HumHk83nVE8ugrlHhzHJtpiUDGHJ26YYCUInOnLtcq9UAAN/zPd+j5gAHQYBIJIJnnnkGTz311EXXNjU1hQceeAB33HEHstmsEsMUvxS7dFj1oDK9h5nCTx8hxTCzMAxhWRba7TZqtZoSyufOncPJkyextLSEdrutnGVeE45schxHCWJdJNNp5mt5DenKUiSz/5vX1XEcpFIpJBKJoc0IfjaO46jNAIptHo/no/88CAI101m/frFYTIWYdbtdnDp1Ck899RRWVlbUeabT6aHxYlwzv+emSr/fR6fTge/7aLfbsG1bjfoShIvwEYzxXixcPeP6I21w4hQGJ07d0vcUhNuEj0Duw9sKuQ8LN4LLlmSHYfiEYRj7Rh7+VgDvOv/1bwB4HMD/df7x3ww3FMkXDMPIGoYxF4bh0rUukOWxk9S/vHv3bniep8QW1whsljRTjBIKZD1QCticcey6LgqFAoAN9/POO+/Erl270O120Wg0sGfPHrz44ov40Ic+tGV4FbBxrYrFIvbv349isajcYbquFKR8XE/w1sOz+HxCMdlutzEYDFCr1RCJRLC+vq5Ki1mGvbS0hEqlolxV3fVlPzQAJRr5M12c684w18PNCYpRros9xkyo5mfiuq4qH4/FYrAsS5V162JYXw/FKtfWbrcRi8XUZ8nAtHg8jmaziUqlgqWlJViWpcQ/19NqtdQx9cRv/ZrT9fc8D5ZlIZFIIJPJYHV19Tp/Q4WdyLjvxcLVEc3lgHwW0Fo6bhb9104OPzDob/k8QRCuD7kPb0MM45aJ5aF7sdyHdxTX2sM8o/0PfhnAzPmvFwCc0Z539vxjF9wcDMP4IWzsuF2USCSC++67D4uLi9e4zJvD9PQ0Go2GEniO46jZyhwzRXHGmcwUTPqcYD0ROpvNIp1Ow/d9JJNJ/PAP/7ByJB3HQRAE+PKXv4wXXnjhouuamZnBm970JszOzg4JztHeaQpSPSSL50KnVw8to/AcDAZot9tYXl5WSdfNZhPVahXr6+toNBpKzPPY0WhUbXbQ/WU5Mh1thmNR0LMknO+vj+DideRmhN5zPTpySu9JBjB0bgDgeZ7auOBnw80EiuZ6va7cfV1wMyBtdXUViUQCyWRSvSeddD2ojq6+Lv75Pq1WC4ZhIJFIIJvNol6vT9QGkTDR3NB7sYNbn+K8Y4kYt04syx9mgjBO5D48oURzuQ13+SajhLLci3cs1x36FYZhaBjGVc+mCMPwwwA+DAAXe300GsVjjz2GJ5544jpXeWOxbVu5qBRpFH+maaJer6vvgU3xCED1M9N9DcMQjuMgk8kgEomg2Wzi/vvvx/T0NIIgwOLiIvbv34+XXnoJv/ZrvzbUWzu6pj179mDfvn1IpVJKMI/28nINFKZ64BgFPwOvKHh5fhwN9dxzz6ne23a7jVarhU6no0Q2hanjOGpjgS4vBXKhUFBiURe/sVhM9S7r19d1XSVieV15TbvdLnq9nhLr7DHm1+xvpoOsp2frvdMsl+b7N5vNIafdcZwLRocNBgNUq1UkEgklclkSzu/1oLWtxnHRjbdtG/l8HvV6HcvLy1f3Sync9tyIe3HayMucoRuAcpdvBfIHmiBMDHIfnjBuwcalbFreHlyrYF5hWYlhGHMAWEN6DsBu7Xm7zj92TUQiERw9ehQf+chHrvUQN4V+vz/kElIMUXDpc5Yp2HRHmf28FIMUkJzp/B3f8R2qrNs0TQRBgBdeeAFPP/30RddULBZx5MgRzM/Pw3Vd5W7rbjZFmh58RTeX4564Pv6czihLmKvVKs6cOQPP85SA1cul9VRqHoslyTzXubk55cguLy+jXC6rHmCGlIVhiEQigX6/rxLGeS0pMnVnnO6wLmT5te4+U5wCmyXdego2A8z0vnPXdZXQ5+dLR5xl2/Pz80gmk2g0GkNhbtxI0OdX673NLFlnn3osFkOxWESr1brk2DBBOM8tuRcLV8ktcpcFQZgI5D48gdyyjUsRy7cF1yqY/wDA3wXw8+f/+0nt8R81DOPjAN4KoHa9/cvT09M4dWpymuanpqYwGAyUwKRrytJjy7KUW6uXZgObpdD6aKF4PI5sNgvDMFCtVrFv3z7s3r0bvV4P1WoVu3btwiuvvIKf/dmfvWDUEjFNE3Nzc9i7d+9QKjaAIeFIKOa5doZrUThSbHuep/prKTwrlYpyom3bhmmaQzOG9VJjYGMedDweRzqdRrFYxNTUFObm5lTy92AwQCKRQLPZVKXJurNM4d3pdIbGXDGFnGnbwHBwGIWtPlqKM6Z1V5muNEuz+XyKWtd1lVCORCJDojwSiaDVasE0TdRqNSWs2S/NNXFTQU9Q58/1lG/P89RYrXg8LoJZuBJuyb1YuHKi2cwtc5f7x09c1fOjuRzyf3Rx8+oLb+zHHd/95etdliDcbsh9eBK5Fe7yVd6DgYvfh794bjf2fPvF2y6F8XJZwWwYxm9jI8ygaBjGWQA/jY2bwu8YhvEDAE4B+I7zT/8UgG8A8BqANoDvu94F6r23k4DjOPB9X4lOHdu24fu+ciAp6gjFsj4z2DRNFItF5Xp+27d9GwCoXubBYIBz587hq1/96kXXlMlksG/fPhQKBTXfWJ9TrLuYenl4GIbwPA/9fh+2bavk7CAIYFkWXNdFEASqhxeACrPSA7Msy1Ljs1jOHI/HMTMzg0KhgNnZWRSLRWSzWbiui0QiAc/z0Ol0cPfdd6sRXaurq6hWq6rUm8Kc78XRTKNp45ZlqY0KBnRRAPO609llSTTdd14TXgM+h+fBzQKOumJgmGVZCIJAbTiwlzmVSqHT6aiyewpiYEOc6/OnASgBz/JwVhkUCgV4nod6vX6Nv6nCTmPc92LhCjEit85dvor/b4ym05j74y5S5sXzER47/BJaX0gBAB5/7ggO//2LVzUJwu2I3Ie3B7ds4/Iq7sGReByzn40igq3vw1+3/xU0zt9/df7yi3fj0D+8+HQc4dZwJSnZH7jIj96zxXNDAP/wehcFbDhvb37zm/H666/fiMPdMCzLgmmaQwnMFF96CrTeowpsCGCGQVFcRiIRZDIZuK6LVquFmZkZHD16FP1+H5VKBXv37sXi4iI+8pGPXHLToFAoYO/evUMjiei+jo5GYikxy5T1kmM6q3SeeRw6pM1mU20I0G2lSGaps+M42L9/Pw4cOIA77rgDuVwOmUwGCwsLym3u9/sol8sIggCu6yKdTqNWq6FcLqvwsJWVFZRKJbTbbdTrddTrdVXuzBnKLFtnojVTrIHNcnOWWNM5ppBmqTvHZbEk2nEctXHA0upoNIpMJjMkaPk4RbXv+yiXyyqd3HEctSmgb1Dozjg/CwpqvaQ/Ho8jlUqh1WpdtG9duL0Y171YuHKi2Qwwlb8l73UlzoZhWtj9V+erZ4wu3OjWExZILNJHJrKR//CNR7+C4Knh+dFPfvo+7P2Xn7/GFQvC9kfuw9uEW7lxeSkiUez5vLPxpTGAG/Uu/lQjRMa88Off8PBzQ/diuQ+Ph+sO/bpZRCIRvOc978GXvvSlcS9liEOHDsHzvKHAJvbM0nnm9yy9Zb8tw7ZYdpxIJJDL5QBsJDYfPXpUCe14fCMlsVKp4I/+6I8uup5kMoldu3Yhn8+rUUp0QPWkbDrOXAfFve/7SoBS+FMQ6vOX6fb6vo9er6dCrvSe3kKhgMOHD+PBBx/E3Xffjbm5OczMzCCbzSIejw/Nqj5w4MDQefR6vaHEbY5sWllZGRLSHB3VarXQbrfVGKkgCNS1Znk5E8Ypmin0OQKKTjSFKtfH82k2m2r2smEYqqe73W6rjQPXddX1qtVqaDabyGQyyjVvtVpD11XfWGG5PDcu+N681rlcDq1WC9Vq9Yb9/gqCsEO4nLMRieLA5yKwIsE1HT4W6SMWGd6se983PIs/6z+CPf/qc9d0TEEQhJvNrdy4vCSGgTueMq/5HkxG78Xv+4Zn0Xu/g7/8xJux6+fkXnyrmGjBfODAAfybf/Nvxr2UIdLpNNbW1hAEgerh1cWwPh6JZb10edkvS/HEXlUKone84x2qTLpYLKJSqeCjH/3oJQVTIpHA9PQ0ksmkKqsGNhOgAQy5mHr6NJ1Lin6WGfMcWG5smqYS//V6fei1jrOxc+a6Lh588EE8/PDDePDBB/HAAw8MlUVfjlgshnQ6rcQkr1GhUEC5XMba2hpqtRp830e73Ua1WkW73VbilaXSTMem6GT5uF5Gzs+LpdwUzgz2YsiY67pIpVJDmwd8rZ7UTbEejUaxuLiIu+++G+l0GuVyGYlEYsjlp7vPJG+uSRfNLCu3LAuZTAZBEKjzEgRhMomm08B0cdzLAAAceub8RIBI74YeN2KE+LpvfgaDb9o4/pP/4yhmPyh/sAmCIIxy6Gnrht+DgY37sGX08J6/9QwG37ZxL/7rjxzF9C/LvfhmMrGCGdhIfj5x4uob6m8Ws7OzWF9fVy4xy4JZwkznmCJZn/NLF5KYpolkMol4PI5yuYzZ2VkcOHAA/X4fnuchkUigVCrh8ccfv+SaHMdR7rI+8ojp1/r8YpZk83u91JehVACGzo3zhIMgUEKShGEI0zSRyWRw8OBBvOUtb8EjjzyCu+++e+hcrxRuIliWhXg8ruZbz8/Po1aroVQqoV6vK+fW933lSPd6Pfi+r2ZBe56nRl5x3BfR+8gZuMXScr2nm+50KpWCbdtKGDMErNFoqJJrXutyuYxGo4FUKoVMJoPBYIByuTzUU85rTPHO665XBHBNqVRKOekXC30TBEEgdz5rImLcvDaOiBGq47/ze54Bvmfj3vr5X34I+f9HygQFQRBu9n0YGL4Xv+P7ngG+b+Ne/PQHjyLzW1+4qe99OzLRgnnSXLU9e/ag1+uptGV9tjEdSDqFFMicR8xQLzqM6XQa2WxWCdEjR44gDEOsr68rR/HJJ5/Ec889d9H1mKaJbDaLdDoN13VVzy4dYLrEeqiUHvyli0U95ZrnNRqy5XkeqtUqDMOA7/vIZDJIJBI4ePCgcpb37dt3TWKZxGIxTE1NoVgsDh0nDEPUajUsLi6iXq/D8zw0m02srKxgbW0Nvu+jVquhXq8r4cySaqaVB8FGWQyFJ11fPYyLLjM/U8dx1DioWCw2VOINbCZz8/UAsLKyggMHDiCVSqFWqyGRSKhrz98RPYSMGxwUyiz157XnyDFJzRYE4VJs/JF260Iy9fd69EefweAfbvzB9uV/92bEPyEhNYIg3H7c6vswMHwvfviffBH4J5sm0Ys/dT/MP332lq5nJzKRgtkwDLz1rW+duN5NltcyMGswGKgkaTqI7EHVS3B1sQxAhULl83msra3BMAwcOnRIJVAXi0W022185StfUa+92Hrm5+eRyWQAYEjc0u3WBbLuLgNQZch8LoChfl4GVLHsuNlsKvGYTqfViKiDBw9i//792Lt3LxKJxBYrvTq4piAI0O12kUwmYRiG2hwIggD1eh3Ly8sqdIvCNJ1OK4e31WohkUhgfX0drVZryLllCTp7t1mKPtrjTSee/eG8RhS8/Lx1YV6pVJBOpxGPx5XIjkQiqpxdv/a8zvz9GJ2Fzd+pXC6nEswFQRC24lb/kXax9z/6z74I/LONPy9e/Kf3IfoXk5VFIgjCzmQSWmMm5T5M7vu554Cf27gfv/wPjgBPy+iqa+HarcCbiGEYePTRR/HCC5P1oXI2MUt59ZJm9sLq/aos/QU25xLzdalUSpUJFwoFzM/Pq15cy7KwtraG11577YrWpPcux2IxlZat99zy/U3ThGmacF1XhVsxwIzOph5OppcNt1otdLtdpNNpFAoFZDIZ7N69G/v378f+/ftVUNmNotPp4OTJkzh+/LgSityAiEQiSrBzHrLjOJiamsKuXbuwd+9ezM7OYmpqCplMBqlUSglYVgfomwiRSOSCsm32RbMqYPTzBqCCz3gNe70ePM/D2toaotEoZmZm1PP168Pj872BzfnYuvPPx1gWLgiCsBV3PnvlmRG3knt/4XkMvuaByUisFQRBuEH0j104xWdS78PkyK+8jMgDd497GduSiXWYH330Ufzsz/7suJdyAXQk9SRpljTTjdSdQgrZXq+nRFYsFkMqlUK328X6+jp27dqFhYUFdDodxONxdDodvPjii2i1Wpddjy7m6AjTudTDu1iarZeO82sA6hzoUuuJ0RTbrVYLrusq4ZlKpVAoFFAsFpFKpVCv1xGJRG6YcE6lUrjjjjtUX3C/34fv+0q4c11TU1NK/DNtG4BK8qYDr58nMNy3zXJ2povrDi+vHbDZCz4YDJT7H41G0W63h2ZzVyoV5HI5TE1Nod1uo1KpqLnarVZrKIRN/1z09+Pnyc8wn8+rpHJBEG5ftvpDbZK55z+/CCCKV3/gMAbPvTTu5QiCINy2hFHZvLwWJtZh3rNnD55++ulxL+UCmHYdBIEqlwWgHEaOLBoVZXyeno7N0CjLspTbaNs2SqUSvvjFL6JcLl9yLSxH5lgoPViMTrEuvHRXvNfrDZWOU2Drop//Zeo3BX0ymUQqlUIymUQ2m0WhUFCp4VcqlnnMy+G6LhYWFpBKpS5weaPRKFzXxczMDGZnZ9W65ubm1EgrzoHO5/NIJBIq8ZufmZ52rY/iMk1T9TL7vj80l5qfYSqVUs/hPGiWffu+j+XlZTVf2zRN9Ho9uK6rBDc/Lwrt0T5yvSS81+shk8moTRdBEARixCZy7/sCwlhk26xVEAThatgu97a7PvwqIvcf2TbrnRQm9molk8lxL+ECKCJZuquXWeuOpG3bKjDLtu2hMUSDwQCFQgGWZaFcLqsE5LW1NRQKBeRyOZw+fRpnzpxRTunFiMViSCaTcF1XOZ4cdUXhppdlc310NLc6P700mWvv9XqoVqvo9XqYmppSojiXyw0J0ouNkaLIpVAFNkUhhfalMAwDtm1v+bx2u416vY5EIoF2u602DwzDQK1Wg+M4cF0Xy8vLCIIAnU4HhmGg0+kMjZ3iRoPuyOvXjefBMm3TNJFIJIYc4Gg0qkq0wzBEo9FAuVzG/Pw8crkczp07B8dxkMvlsL6+roQw+595nfh7xvJsEolEkMvlEASB9DILgqA4/AVj7H1zV8Jdv/YKAODYB+7Ydi65IAjCpdgu92EAuPO/HwcAHP/2feidODnexWwTJlIwc0btpEFXkMLSsiwlPCmaKMjoEOr9w3w8lUphMBig3W6rkVSNRgOO46Df72NxcfGSYV+Eoo0jkOgq6+FfFKn6SCmupdvtKpecQVgUcTw+xVu1WkUsFsPMzAy63S4SiYQSzPl8/pIbHLpQ1qEQvh7i8Tji8ThmZ2fh+z46nQ5KpRLOnTuHbreLXC4H0zSVE9/tdlWZ+draGoDNcU4AlKNOB57rY/r4YDBQSdmdTkeFejWbTSXUeW69Xg/lchmu6yKdTmN9fR2+78N1XQBAtVpV/dh6Gf1o2T+w8bsXBAESiYSaly0IghCJxwFcvlpnkghdqZQRBGEbEw4L4+14HwaA0JF78ZUykSXZDz74IE6ePDnuZWxJr9eDaZoXjE5i+TNDqUbHA1EE2bat0rbb7TYcx0EymVRjko4fP47FxUU4jqPE18VgSTJDvyiY+TNg0zXW+5iBjVLnTCaj5jWzVJnp0brr2u/30Wg0VDn6YDBAsVjE9PQ0stmscqKvBwrF68FxHKTTaViWpcRsNptFPB5HKpVCsVhU4l4veabo5WfHDQOKZpaj6668fr58HsvMGazW7/dRq9Vw7tw55c5blgXTNFU5NzdR9FJ6Jmjr6A50LpeT0mxBEBBJpXDHX/a3jatB7vzvxxE7sG/cyxAEQbgm+q+9MfT9drwPA8Dhj55A9O7D417GtmDiHGbDMPCWt7wFX/3qV8e9lAugE0t3j72svu8rMWPbtpqtzPFEFECDwUCJt9XVVfi+j2w2C8uysLi4iDAMYZomPM9T/cSXQncgKYgp5lkiTBFPd5yimK/Re5bpkrPflkLf9/2hmcZ0lwuFAlKpFJrNJmzbRiqVuuZry7VcDq5LP1cdOvPFYhG9Xk/1XLNEutlsotvtwvd9OI6jBDFHR+ml2sDG58nrplcYMM2c78eZyplMRm1cBEGAwWCARqOBarWK6elptFottNttZLNZ5UJzE4KbGfo85tHr0+v1kEqlUKlUrqgKQRCEHQY3RbMZ7PuzALHI9W00CoIgCNdONJsBsH3DWA9/9ASOf9MsekvL417KRDORgvnQoUP4zGc+M+6lDEFxMyrqKHhYWh0EgerlpQjlvF9goze71+upXlfbthGLxeB5nkpS5tzji/UEE44mCoIAlmWpUl/f95WY1sO+9BJxvUSboWDtdntoRjNFHEuNXddVo5vS6bRKoY5EIsjn81d8LcMwRKvV2rKMm6XqYRhuKcA5tonl4sBmSbXneVheXlZhakyvHgwGqpQ5FoshkUigXq8jGo3CcZyh0nqGeA0GA+UGUyzz86BQZbK2bdvo9Xrqc4hEIio1m59ltVpFJpPB9PQ0Tp8+jX6/j3g8Dtd11TWka8w52BTo/K8eKpZKpdTvkCAItw/9108iWixg96fasCI783//0VwOiIxU2ZQrY1qNIAjCxZGNy9uDiRPMALCwsIC/+qu/GvcyhpiZmRkKaWI/K91OPaW61+sN9blSyFJ0BkGgyrFd10W73VYlz77vq/7tyzmu3W53qNebjjZHFekOsz6/mM/Vk5r1n41uDNAZp8DM5XLIZrMoFovYv3//kNN9JRiGgUQiseXPIpHIJfuhHcdRojYMQwRBAM/z1MaB67rI5XKIxWJYX1/H0tISarUaarUawjBU5c8MCaNry/WbpolCoaCuz2jpdhAEqt+YFQbAxsgtOs562jZD4mq1GlZXV7Fr1y7kcjk0m01VQs7Z3vzd0R3mrWZ6x2IxNdpLBLMg3F7EZmcw+/tNuNHt1y93KaLFAmBs3OOW/1seVmz4D9DMN4hgFgRhsojNzQKojnsZN5To1NQFj/XPZ/7czkykYE4mk6hUJuv/HBcWFpRTS4eRbiKwOa+X4pkjk/S+Yv7X8zxVEkxx2G631fN930cYhkgkEkOvH8X3fVQqFeVw66Fd+ixlli+zh5nCUQ8so0M+OuLIsiz1c8dxkM1m1RinbDY7FJhF2Md7KS7Xn030mdZbHYMCutvtwnEc1T8MQLm3zWYTnU5HbSLQLeZcYzrWqVQKsVhMufJ8Psvr9R5rvRycM6Iplun8c3OF17nRaKDb7SKfzyMIAvR6PSQSCWQyGVVirf8e6aX1FM79fh+e58F1XVVCfr3944IgbB+mf7+FlLl9y/+2IjozjVP/dQopd6PdyYK4NYIgTCZGb/P+NP8/Gzum0ic6Mw0jEsHp/5pHwh5u+Sv8PRO9c4tjWtlkMHGCmcFSk4ZeAktRBWDIsdXLoPWfsTcZgBJsvV4P2WwWyWRyyKVk+TOTn2dmZrC0tLTlmrrdLprNJjzPQ6fTUSKODrjek0wRSOFH95TOJs+B69ZFPgVzPB7H9PQ0du/ejV27dl3UJb6cWL6UCNZhuTv7iC+FLpSJbduYmppSGwjNZlOdB4Wo67pqzFc0Gt1yZFgikcBgMIDneYhGo2i1WkNuPIWvfr14jfl5GIaBVquF1dVV7Nu3DzMzM1hdXYVpmpiamkK73Val2RTm/Lx08c4NDwCYnp5WY8kEQRC2I7GFeZz4YAEZd/ImYwiCsL0IwxDGYABcRdXj1dJ74xQQhojt34uIsXrT3udW0p8v4NRPRJBJeEjgwnyc8q/HUfgHu9E7dWYMq5sMJk4w79mzB+vr6+NexgX0ej3lHPb7feUu68JUL5llmS9dX8MwkEwmVagXsJF27LquKolm/y4AZDIZ1cd7McHcbrdVz+6oM6z3HlOwx2IxNUqKa2MZMMOsdLHPc9Yd2UKhoOYucx7z1XKl7jLLoa8VhpYlEglks1lVhs3PzHVdtZmgbxJQ4FLAMpSLCdamacL3ffi+r96Hr9PFst4/zp7rRqOBVquFQqGAYrGori+T0vl8VgHoGxp6yTY3bq4kKE0QhJ1BsCuLCFbGvYwbRmz3Lrz+H3PIJkQsC4Jw/QwaDUSjEWC6eNPfa/a3yzumNab5bz1kBpfeZCj/io3CP9y7sWFwGzJxgvntb387zp49O+5lXABFMEURRaTe76u7g3RFKUY5ugjYKMnWXVM92KnT6ahjsOdZHwel0+v1sL6+jkqlokq8+V66w0lnkmKLpdd8b33d+vgp13VRr9cRBAEymYwqeU6n09cslm8lvV4PnU5HieRCoaBKs1k6DUCVv7daLQCbbjVL5+lyU+gz+RzY+Czp6AObs7bp4OuBa3pZNUvJm80m2u02UqkUqtXqUOAXy+r5OfLYJBqNIpVKIQiC6x7JJQjC9RGGISK9PsLYzdnE8vflsOtfHdsx5djBnjxO/3Af+WRr3EsRBEG4Iozuzii/1vH6Jq50IFb5l03kf+wA+sdP3NQ1TSITNYfZMAwcPXoUb7zxxuWffIuhmwhsJkizdJwzdynG9B5UYDNxmf3KdH6DIFDOst5vTFEeiUTUrOOL0W63sba2poKvKNAohPUSbK6VripdaSY000HVnd1Go6ESu9PpNKampq4qEXsc8LPpdDoqGI2p1ZyDnUwmVTl8MplU/dC2bcO2bbiuq+Y1UxxTMPOacZRWOp2GaZrqGuqzr7lhYRiGCgOj45zP59Xc6HQ6rX4/9LJ6fnb8miFzLFdnwJkgCONl0GhgsFYe6m+7kUz/9AnkrfZNOfY46PzkOvKZKxPLK2dyN3k1giAIl6d38jQQhojcf2Rbzl0epdWz8dXKHPqXcZd1Kv95oqTjLWPizvrOO+/ESy+9NO5lXADLqFlaS0FKd49iiiKa/3UcR4kfCmu6j51OB6VSSTnE+jxjurnsGb4YzWYTZ8+exfr6ukpw1vuVKZR1d5PrZd8u+3A5Xom9vEz5jkQiSKfT2LVrF3bv3n3ZHuVx0263ce7cOZw9e1aVQLdaLfi+PxQUxv7uZDIJ27aRTqeVk043PhaLwXVdlZJN8cvQLiZus7Qe2OxrHp2vHIvFVLp4IpFQYpx91fl8Xn0mAC4o9Qc2y7IBXPDZCoIwXgaNBsL12riXMfG0ejYG4ZW15gDA4R9+5iauRhCEnUTYH9y0jUsy9avnkIht7/yYVs/Gy+sz6F2FWCaR+4/chBVNNhMnmIvFIo4fPz7uZVxApVIZKqGmC6y7wbqg4j86mBQ1nU5H9TNPTU1hMBigVCqhWq0iDENks1nlOlI8z8zMXHTUku/7WFxcxOLiogr/0su36RozLIoCjOOVAKgSc72Hl4+xNLtYLKJYLA6FlI0yCSOOuG66y+w1pivMzQDHcdQcZEJRzGsThqFym23bVq/lc7jxwHJ79quHYYhut6uuOSsI4vE4isUiZmZm4LquOn4mk0E6nVZO9ej5cN38x3Ph55XNZkU0C8IOpnFPETFjZ6ThN7rONf+RJgiCcDlk4/LKeLU6fc334fK/G//f+7eaift/rFwuh9XVyUudK5VKQ0JZh+XNemm1Pr7JcRwkEglVCh2NRpHL5TA9Pa3EWSwWQ6FQQDweRzweRyqVQjKZhOu6yOfzOHDgwEWTotfW1vDKK6+gVCqpcUYAhkSxPvKKJctBEChhR8FF9zQajarxWezfbbVaQ3OfR7nUz24V/Gx4rZk4TqFLR5fusmmaSiiPbiSwOoBOMX9GAcvjUNRSWOtl2HqIGl3kRCIBy7KQyWSwd+9e7N69G4lEQolz/Vz0EDBWB+hzwAERzIIwSYS93g3tc6vfV8TCPz2Oot28YcccF42ug+O1qav+Iy18+/03aUWCIOxEbvR9WGfwjgdhbvMNzEbXuaoqn1EiRgi87b4buKLJZ+IE86SW+3qeh16vB9u2h4QMBRRFDB1HBnix9zmTySAejyvHV+9njUQiyOfzqjdaF7B0IHfv3n3R3mHP83D8+HG8/PLLqryb7icdSQrlUfeZSdl6qjYABEGgRKJlWfA8D5VKBdVq9aLXKJVK3aCrfe2wl7zVaqFer6PVaqHX6w2JY4pZ27bVZ8HPg9eMx9JnIOsp2Y7jwHXdIfHNMnoAQ0Ft/P1wXVeJZcuyUCgUsHv3bhw4cECVY+vu/egMZmBTRBM9VE4QhPEzaDQwKFVu2B9r+X90CrNO/YYca5zUui6O16bQ6V/95l79X0gwmCAIV86g0UBYb9zQYxreRgl24edPbfvwxdfrxeuu8ln/Fx4GX/PAjVnQNmCiBLPjOCqpeBJpNptq1rGO7iKyd5kChinLo6FbTN2ORqMoFApIJpNK7DWbTTQaDZXunMlkMD09jYMHD140nXplZQVPP/00jh07hmazOeRAcu4zBTCAITFH4cznAFABU3RGGZTFDYNJgQKy1+shCAJ4noe1tTUsLi6iVCqp3mVuQlAYUzTza4pn3aHWR4KNltpTJNNV5nHZH03xzM+BP+MGRDweV2XciUQCU1NT6jUU29zMAIZHcXGDQ/+Mk8nkZWdVC4Jwa7gZf6xtd0428tcklgVBEK6FsNOF0blxY59655aACfr791qpdd0b0hITMUJUf2LnBFFejomypXbv3n1JB3Pc+L4/lGSsp1Drydh0aylgms0mEokEMpmMKm+mIJ2dnUU2m1WzfSuVCprNjdI7z/NQLBZVf+vevXvheR5eeuklJcx1zp07h6efflqVf+fzeSX0RkdM0W0eDAbqMQrDbreLZrOJTqejgsf27duHYrGIVCp1xXOUbyS6O64nkuubDN1uF91uF6+//jrOnDmDVCqFfD6vUqUBDAlRnjOvQRAE6ppQTOufFV+rl2+zTz0IAiWKe70eWq2WmtnN3xUmdPP3IAxD+L6PaDQKx3GQyWRUjzjTsCmK9fXq/ejs106lUlhfXx9ynwVBGB9hp4tIp4vQmsyqqVtJpRO/qhRWQRCE62XQ2Ni0jBRy130fNloeEA7gf/PDcKMv34jljYVKJ46T9cINzZDovu8hmH/67A073qQyUYL57rvvxrlz58a9jIuSTqeVILEsC8BGGjYFmT6Oif/ljGMKPJZl+76PRCKhRC1HDQGbQU/AZllwMpnEgQMHkEql4Ps+Xn311QvW1+128dprrynRZZomUqnUBb25FPZ6AjPFKHuVK5UKwjDE9PQ0pqamUCgUhkqObyUUsQzw4j/2X9dqNSwtLSlxqlcC6DOK9U0Nur8UnRS3fIyfG4/DsmiKbT0B27ZtdLtd5R67rqtENn9f9FnclmUNOcl66T5/ro8k4++SXjHAddBdZ+q3IAiTwaDRgGFbQD477qWMFf6BJu6yIAi3mhslmntLK2j/zYex95++iow5/ryeq6XSiaMziGGxmbmh9+JoZIDaP6qj+Kc37JATy0QJ5kOHDuGVV14Z9zIuSiaTQaVSUaXUtm2j2WyqXmRd2HD8D4U1sDHuyHEcdDod9Pv9oRFCJJVKIZPJwDAMeJ6nxF0kEkE2m8XBgweRTCYRBAFOnjx5wRqbzSZeeOEFmKaJXq+HPXv2XOBsj/bAUnQ1Gg34vo9qtQrTNJFIJJBOp5FKpZQ4dRznlpf+ckOiWq0OlZt3Oh20222USiXU63V1vVkSDUC50MDmPGzd/ef1pTjmZ0tB3Ol0lDjVNxds21ZhXNzgoAjudruIRCLKKdaD2CiW9d+VbreLdruteq0ty1KfiX4NuF49CIzHoIgWBGFyCIPOdbnMRqN1XcEs46DWdeH1N893uZW+IX+gtb/trYh/4qnrPo4gCLcX1yOajUZroww7HCD3Y6e2XfgihfJyKw2/N1GSb9sxUVdv165dePrpp8e9jC2Jx+OIxWIIgo2mf92N1B0//XF9pFCn00Gv10Oz2US1WkUqlRpyaym82AsLQAk4irGpqSnMzMxgenoalmXhYx/7GM6ePXvBWmu1Gl544QUAG2Xkc3NzarYzxRhd1Ha7jVarhXa7jdXVVdRqNRiGgQMHDqBQKCCVSg0JOD0J/FZAwdlut5WzzNLpbreLRqMBz/OUE687syxb1v/pMKhND1kDoMQzS+35Oegl7HqSNpO4OQfbtm04jjO0Br3k3bZt1cOuh3jpP9cFuh4gpn89SiqVUmFtgiCMn+v9Q623vAJjMH8zlnbTWGqnUfXdyz/xKun+YBn4xA0/rCAItwG8FxtxF0bcRWheXP4YLQ/hecOit1YGBjd3pvPNotKJ43QjL0L5BjFRV3FmZgYvvzyZvQGzs7MIw3Co9JoiCIASQFuNlqLb3Ol0VGhXPB5XvcN637PjOEoQWZYF3/fVMROJBIrFImZnZ5FKpdDpdPDxj38cy8vLF6y3VCrhhRdeQDQaRaPRgOu6akyVXs5bKpWwtraGIAjQbrfR6XSQz+eViLRtGwDQarWGSoVvFv1+H91uV5Upt9tt1Ot1eJ6HIAhUAjiAC5xZikxeU342/Jzo9vP1FMSjwpXH4DXodDqwbRu9Xk9tYDAdmwLVdV3l4NNFHk245ufI0mz9mutjyfQE9dG0bT42+jmEYYh0Oo1KpSKCWRAmCP0Ptauht1ZG9Xvehv32he03k0qlE0fQv3l/VtS/621If+wLN+34giDsXAaNBtBoIJJKXfJ+vJVIrn7v23HInEx9shW3UixHDKDxnW9D6uM7+948UYI5l8vh2LFj417GlqTTaVXK2+l0hsb46GWywOaoHwovfRyVbdtIp9NIJpOIx+PKueVzAQwJHvbLUkQmEglks1nkcjn8nb/zdzAYDPCxj30MpVLpgjWXSiW8+OKL8H1fJXHTZbYsC51OB6VSCevr66rEl6KfQs/3fTSbTdVzrZcz30jCMFSjoGq1mnKOPc9Dq9Uamh+tJ3lTQDKASw83o+ik+B1lVFjqqdijTjrFNj8r13WVg6yvgSXr7XYbsVgM8Xgc7XYb/X5/aNwXe9k5rkwX+Xq1gj4fW9+A4e8c18hNA+ljFoTJg3+oXS3x713cNiOlKp04zjRz8Lo3L+ci/J4S8LGbdnhBEG4DruV+nPvfTm+bcuxSkMS5VuaWOcvRyADBd1cwiL0Nmd/auaJ5ogQze4InEb33VRc0hOFMFJ0UOhQ5DP4CoPqJKcJ834dlWUqM8X3CMEQsFkO324XneSiXy2i1WpiZmYFt2zhy5Aj+3t/7e+j3+/joRz+Kev3CP6yWl5fR7XZx6NChodTlwWCAIAiUc91ut9FoNIZEO0vI6TLrvb7XCxOiWe69srKClZUVrK2todVqqXVQUAJQo6N43fUgLrquDDfThSfPm+erf35Mt2aZOgCVQM7ncLODTrVpmmodLPXmZ8/PS3eS6UzzGne7XRU6xvXp46xY0k3nme/L5+qCXhf9+gxnQRCEW0klSNxUsSwIgiBcGorlW30vNqMDtP92DZnfuqVve0uZKME8yaFFdJYpjPV+Zv5cd/90kQNAic9Wq4VEIqEEFucyU/zoM38BKHHW6XSwvr6O06dPY2ZmBvl8HrFYDPfeey9+5Ed+BJVKBb/3e7+35bipcrmshPzCwoJ6XF8v18iSYJYoc8QUAFWGfClYYszEaJ7jqCtNd7jVamF1dRWvv/46qtUqPM9TrwGAIAhUaBZddiaTc3ay7s5T0I5+HvoYLQpbovec8zWjYpZjnvi56e+nXy++P9fA68AS81qthmq1qjZI6Dr7vo9Op6PmXuthcXwfOuD8Xl8rN2Xi8Tjq9fpEzcoWBGHnUunEUe+4aHWtyz/5BrD299+OqV/9/C15L0EQhO3CuMQyAHT7ETj/M3vL3/dWMlGCeZKxbVuN/tGTpilYdGdxq5JlCphOp4NkMqkcRr18m+XQtm0Puae2bauxSmfOnMHMzIzqRwaAu+66Cz/2Yz+GSqWCT3/601uuv1KpqP7g/fv3D4lYlnwDUKIwGo0qIciSYX1dhMfkOKpz586pnulsNotMJqM2CPhafePh5MmTeO2111TKtS4M2fcdBIEKJqOgjsfjQ33CPJ/RTQo6ytFoVDnrenI2HW5+Pizl9n3/giRrfk76RokeLKaPomKPsv4erBJYXV1V1QYA1LmxFN51XVVerYfH8R9/7/S+Z163XC6HZrM5VP0gCIJws6h24lhtJW/Z+1nfsoYV8xHM/JfP3bL3FARBmFQ++eUHgNDAvn2rY6vyGQwiKH5kZ29kToxg5nzhSYXOoi5YCL+mINR7mvUy2lwup8p7KZYpUDmWatTl5CgnBkytra3h2LFjKBQK2Ldvn3q/o0eP4sd//MfRbDbx13/911ueQ61Ww2uvvYYgCHD48GHkcrkLgqm4HrqknCtsWRay2awSkUEQwPM81d8MAI1GA+fOnVNzkBcWFtDtdhEEgSo15tilbreLarWK06dPq1FRAFTpOkUqR0fV63VUKhW0Wi1VssxwLcuyVPI0HXA9wVz/DPQZ1xS57BPXPzs+TiFKgcsy7Gg0il6vN7Rh4DgOfN9XgpnOuF7evr6+jmq1qpzofr+vera73a5yzUf7kbeax8yf6yPNEomE9DELwg7B+405LP+DxkT2MVc6cVSCxC1zlnWc969iyXwExgCY/aAIZ0EQbj/+4IX7kf9rC1nbQPXenrTE3GQmRjDv379/yx7cSYFupy6mdPE1mpjdarWUELNtG/l8HtPT06qnVXcH6eLyWBRSFJoMkgrDEI1GA8ePH8euXbuwa9cu1X8LAO94xzvw4z/+4wiCAM8+++yW59FqtfDGG2+g0WjgyJEjmJubg+u66HQ66HQ6SsjxXBzHQSqVQrFYRCqVQr1eRxAEqNVqaswTRW6tVsPKygo6nQ4ymQxqtZq6bnr6s+d5aLfbWFlZQavVGhKeeio4xWa/31fjuLrdLlzXVQKZ/yj8mTwOQB1jVCDrY6YolDnSCxgWpyyPp0vMnmd+ZrxmPK5pmvB9X7neemK2Lpoty0IikVAbAixl13ux+Ro98ftiYpgVCNxcEQRh+5P5rS+g+v0Lt0ww17ounlvZaNtJOQHeXDxz0ec2uzZK7cQtWddWxN+7im4/AnxwbEsQBEEYC3/w4n049Ktd9J0Qr39nDDN7KmNdTyw6wNL//gjmfnHnbmBOjGDeu3fvRAtmupW6KNbDmlgCS2Grpxvbto3p6WnMz88jFouhXC6rXlzHcVSwE93N0dFUerCT7/totVo4deoUDh8+jJmZGQAbgsmyLLzvfe9Dp9PBL/zCL+BLX/rSlucSBAHOnTsH3/exvr6O+fl5VSbu+74K1/J9H/V6HdFoFPF4XPUz859e7sy1UeQPBgPUajV0Oh3U63XlMHMToVarqd5ohmLRkaWLy77f9fV1FXjG8VipVEr1+urXmqXKumjWXWaKXV4z/b+68NWDxZhArZdfU9yOCvTRhHMKX/6uBEGAZrOJZDKp3pubFZ1OR20q6H3sW/3+6Y/ra75YIrggCMLFaHQdPLO8G71eFF51IzywZm2MQBwVzZ/8ygOIlUzsenDxlq9zlGgkxNmfegRGH1j49zv3DzVBEMZL87/uwuqPtzHtXP20gxvNH7x4Hw79Sg/4wvOo/uDbMbNnbdxLQjQyAL52HfjFca/k5jExgnlhYQHr6+vjXsZF0Ut6dVGij1li7+2oa8kSYpY09/t9VCoVNJtNOI6jUpt5HAo8vWc4DENVohyNRlGtVrG2toaZmRklzMIwRDKZxDd+4zciDEP8h//wH/Dcc89d9JzK5TLa7TbOnj2LO+64A7t27VIl457nKRe5VqupICm9j1YXj4PBAO12GwBUny/HQnEDQXc/Pc8bEqwU6d1uV70Pnezl5WWUy2WYpolMJoN8Po9UKjUU+sVrxM+Fx+AGBsvbdTG5VfI0w8bYi8y162FbdI51ccw+bz5PF7166TR7shuNhhLirDYYdcxH4Xnq5fr8nWSPNrBRFTCpafOCIEwWja6Dz53ZD79mDz0ediIoNRJAcfOxTz5/P+78ZQ+VN5nw7h1/+V/ECJH6mlX0BxGcjTyCXT8nolkQhBtP4veeQvMfzWIa4xXM/+ulN+Hwh7oIn3lhrOvYCjvWx9mf3Ln34YkRzLOzs1hbG/8uycVgXy/7RfVxQt1ud0ggUTwBG4FPurtJ4dxoNFCr1RCPx5FOp5UrS/Gsj5gCoNxWisRer4f19XW022010xfYEG6JRALf9E3fhMFggF/+5V/Gl770paGQKR2K2na7jXK5jD179iCZTCrxa9s2qtUqWq2WEmSc48zNAZaP0yFlmTbTqClMOfaImwOu66rXBkGg3jMajSqxXC6X4XkeTNNELpdDsVhELpdDIpFQApzOq+6+6inl3KRgzzg/Q32El95DzTJwfmb6aCq+j23bKumbide6aKaA5fvyvCnaGfIVj8fVezmOo2Zd8/eAr9PDv/REdZ4rnwdAfX6jc6YFQdh++L8wj+V/5t3wsuxWz8bnzu5Dvx9BULe3fE7vWApf/B9H1fd3nmyil7Wx9mgPMzd0NddHNDKA80gJZ/7ZI9j9b3fmH2uCIIyX+v+9G6V/5t/UecyVThxfXNw99Fg24eEtU6fxv15+Ew795y7CZ18EAHTfexTlt3Un5l680+/DEyOY5+fn8fTTT497GReFzp9e+soSYz3AiynXeok2XVNdGJmmibW1NaTTaaRSqaGUavYQ6+W6unPKEKrl5WWcOXMGCwsLqsQ3Go2i1WrBsix8y7d8C7LZLH7jN34Df/7nf45qtXrR82s0Gjhx4gTW1tawvr6O/fv3I5vNqjRqingKYl388Zp4nod6va6ENF1ylhlz3BT7e/WQLgpXilj29QIbAtC2bRQKBUxPTyOdTqvZ0HqwGq+PPk6KvcgUrbzGuvjUKwT4OY2OotJnOtP95aYBR4yx/5r/9LJqvSebJe4sJ9dLtvXkbz29m+hp3vw95O8LRbvMYxaEnYP9qWew0n8IsZ9+7bJ/qD33Lx+84uN24xG0v/7Safp2xUD8959S3/fe9Wa8/l1RzCxMXjWYGR3Ae3By27oEQdje2H/0DM71HkLkZ44hb7Vv2HEbXQefP7sPANDvR9BpDAcpek0bf/jVaRz6WAPhF78KAOj9jaN44wMGZuaqN2wdN4KdfB+eGMGcyWQm2mHWxyqx35aiVi/lBTZLfSmkWV7M/lwKR4onHp/ijKKJrjKfR6eUvcPr6+s4ceIEgiBAsVhUAtd1XXS7XSQSCbznPe/B1NQUisUinnzySZw4cQKe5215jt1uF5VKBe12GydPnsS9996L3bt3I5lMKueVIo+jnfTQMYpCXgOWkVNQhmEI13WVUGV5+eiMY5ae81zpyhcKBaTTabiuu+VsZYpHPW2cDm+321XrprhkIJc+5koXu/oxuFFhmuZQyT0FOROyW62Wek++1jRNNJtNdDod9Ho9RKNRpNNp1RfOKgReI91Z5qaCfjzd6WblAysSeB4S/CUIOwfr089iufNmLJoXjizUsf/0mSs+ZnxqCotfv/eq1tGesTCzULqq19xKXLuDU//67dj7L3f2eBNBEMaD9elnca57FGdjBu74Ny8jEQu2fN4gNPDiT92/5c+6iSiO/vMvwuub+MuTBzEYGOg2Lz5tIOxGkDgTUWK5/+4348T3GJiZrV73+dwMdup9+LKC2TCM/w7gmwCshmF47/nHfgbA3wNAhftTYRh+6vzPfhLADwDoA/jHYRhuPRh4hFQqNdGC2fM82LYN13VVuWssFlOCmIJKF9MUw4PBQJU1W5aFIAgQiUSQSqXgOM5QSa3eK60LJSY1MzyL4p0ly2fPnkU8HofjOCpgjGXPDz74IDKZDA4dOoTHH38cp0+fxsrKCkql0pDzSph+/dxzz2EwGGD//v0AoMq6dXFJl9UwDPU6jo+ik64niuuzkQEMlbdzA2JU7KbTaaTTaSQSCeXq6w4+/8vyaf0YoyXS/Jx4DMuyLkjMBjaca7rY+saILlQZ2BYEgZqXbZqmEr76Z8fXUOAyvIzPMU0TrVZLlfjrGyN0jmOxGCzLUp+DnqStu+UU8cLO4lbdi4XJJPoXX8KFyQbXeKx0Gm/8g0MAtm7V2YrBOx/E0mOTU/63FWZ0ANwl+Q3CzUPuw0Lss18EAJwMHkQYuYg5EQLm41tPq7FiMTwXPIjF7w0uKZQvRnvawszs5G5c7tT78JX8Zf0RAB8C8Jsjj/+nMAx/QX/AMIy7AXwngHsAzAP4c8MwDodheOm6L2w4zKXS5P4CvPbaazhy5AiSySR834fneXBdd0jgABgSusDmbGbP81CpVOA4jvpXKBSUAKTLqgdT0ammAKUjS9Hsui7q9TrOnTuHWCymSpVffvllzM/P46GHHkKhUIBlWTh8+DCmpqbwwAMP4M///M/x0ksvYWVlZaj0OZvNotPpqNJtwzCwurqKaDSKqakpBEGgBBtdU64H2BDa+hopIpPJJLLZrHoM2EjqBoaTpfUNB86fjsfjKBQKyOfzSCQSSgxSHOqv10PJGKiljwDTXWOWQVOYUpzH4/Eh51p3a+nkUvRzHBfdawpxVgLQJddL6vVe52g0inw+j8FgAM/z1GaK7/tDoXJ8Pc8LgLoO3W53yOHXe+yFHcdHcAvuxcL2JfjTferr3iCCxGMntn6iGYO/5yrE8jsexOvfF8HMdO06VygI256PQO7DAoDIX375ml4X9nqIf+4Yut926Kpf23/3m7H8DR1MX9M7C9fDZQVzGIZPGIax7wqP960APh6GYQDgDcMwXgPwMIDL+vKWZWFpaekK3+bWs7y8jEwmg5mZGTiOo8qqKcz4PcXyqNNXr9dRr9eRz+dVXy/7lYHhkmx9DjFFHgUmn09Xs9vtwjRNJSynp6dRLpdx/PhxtNtt3H333UgkEqqP+E1vehNisRjy+TxOnDiB9fV1Jf7n5uYAAGfOnMHy8vLQeKvV1VU1T5luJ8U8ZzSzRJvis9FoYGVlRZWAF4tFVdJNUUwByFJjOtF01OnEOo6jHFw6rNxsADaTxfWwrtEea37N53NDgmnYXA+rA/g6vqbX68G27aHPiPOluSZ9NBgrDfT0bKaJs6TddV21jkajoQLQKKpZxq/3T1Mkj4Z6bTV2Stg53Kp7sbC9MD67gIixscH2E7v/WD3eh4Ff/ou/ob5f912k3//6Nb1HJ2tienq8cz6vlIQb4MS/fzsO/F/yqy7ceOQ+LIyTIGdieqo87mXcllxP7eaPGobxvQCeBfB/hGG4DmABwBe055w9/9gFGIbxQwB+SC0kFpvoUTj9fh+vvfaamlm8uroKYHOED0uU9RJcfQZwv9/H2toaZmdnYZomUqmUKmOmq5pOp5XLSkGtl+TSKTUMA61WS71vNpvF/Pw87r77bszOzqLX6+HYsWN4/vnn8dJLL+HgwYOYnZ1FIpGAaZrYs2cPms0mFhYWEASB6lumI0sHu9FoKHGml/1S2MfjcSwsLGB2dhazs7PIZDJK0Hc6Hayvr+PMmTOoVCqqXFvve6YYHQ3toliOxWKqbJ3uLZ1ZPle/3sCmYNVL2yk6KXb1UC9+NnpIF9+Hnzt/ppd0s0xbT0nXx20BmyFi+vgszmGmYOaaeA3o+I+W4jORW3f0ib4RoJe1C7cNN+xe7CB+k5cq3Gh+ZPdfwDIuNKyiCPGPd39GfX+yU8TvY+qqjt3a38f633073Mr2McTM6ADGnhsXyCMIV4jchwUAQOrJ4kV/th7EEXvvafTrTdz1oTZe+dGd+Vmn4gFe/49vwx3/xxcu/+RtwrUK5l8B8G8AhOf/+x8BfP/VHCAMww8D+DAAGIYRuq478X/ks8SWpcF0d/VeZX7NEmoSjUZRr9dx8uRJzM/PK4FDp1h3o3WX07IseJ6nHGuWZzebTbRaLSSTSczMzGDPnj3YvXu3ev83velNmJqawokTJ1Cr1bCwsKB6cnft2oXp6WklrCgKyRtvvIFXX30VL7zwAk6dOoVGo6FGKbF313EczM7OYvfu3di7dy9mZmZUGBfd1Lm5ORSLRZw6dQqtVmvovXi+o+/NUmc9nVsfsaS7wDzGVi6yLqZ1gauLaI7yYhkzNwb0XnI+n+KXn61eOs/NjWq1qtx1XZTzc+Qx2u02KpUKGo2GEr/cFOl0Omi1WmojZbTkXO/d5mYK16FXBKRSKdRqtYn/35Rw3dzQe3HayMsvzA5l3lzH1OcO4WwzC/uxM1f0x1po9xHkonC3h7ksCONC7sMCAGDqc1l8z/STF/15J4ziv3zmPYi85wyMM8sADty6xd1CopEBIvNbBwxvV65JMIdhuMKvDcP4dQB/eP7bcwD0AWK7zj+2Y1hZWcGBAwfgui6CIBjqqQU2XU+OUWL6Mefrnjt3DpFIRPXK1ut15HI5ABtl25FIBMlkUonoVCoFYEOss5Tb930AQDweRyaTQSaTUe40MU0TCwsLmJqawtraGk6ePIlms4m5uTmVJH0x0uk0jhw5gve///144okn8Md//MeoVCqwLEs5zrlcDgsLC9i3bx92796t0q9Zrs0ycdd1Yds2FhcXUavVlHuul1/rwV0sNdbnTjPgDNgU2BTIfB2dWI7x0oPTdOFLgaqPf9KdbrrLLK/XA9kohAeDAVKpFBqNhjpfhsJRFFP8AhemprdaLayvr6NSqajPWRfI/X5ffU9nmefCUnMe82Kp5IlEAvV6XQTzDud2vhcLwK4vJLd0l7fCMvr4nunP4yupPfjsIAHj3BqAq0vJFgThQuQ+LJDvmb50tb1l9PGP9nwG//Wz7wa+fWcJyp3ONQlmwzDmwjBkw/HfBPDi+a//AMDHDMP4RWwEHBwCcEXDlXWxN8msrq5i//79Q6OSAAwFMkUiEfi+D8dxVB+yXla9traGwWAAx3FQLpfRaDSwb98+GIah3ORut4ter4daraYCvgAoB7PT6SCZTGLfvn04cuQIpqYuLLXjSKaFhQXMzMxsmaDc7/fRaDRUyBZnKJumiVwuh8ceewx33nknfu/3fg8vvfQSfN9HNBpFLpdDPp9HPp9XawMwlIodhiEsy0KhUFDXigKV/+U6dWHb6/XQbDZVmBYfZ7AW30d39vWSbn4mfH4Yhqo6gPOhR4876kDr6dd8nr5hwPYBfcOE5dK8BluVR1Potlot1Ot1tXnB43LcFhO79Q0C9o3r/fJ0yXkduJ7t8r8n4fq4GfdiYfvwzYXnxr2EiSST8nD8Q2/FoR996vJPFoTrRO7DAgAc+eKVSSrL6OPdU8fwl0N7KcKkcyVjpX4bwLsAFA3DOAvgpwG8yzCMB7BRfnISwA8DQBiGXzUM43cAvASgB+AfXmka4HYJK6KAoaAZXTfnJ1Ow6D2//Hmz2US/30c6nVZzevUE5pWVFcRiMeRyORV0xRLfXq+Her0O3/exd+9eHDhwAHNzc0OidRQGU21Fp9NBo9FQQpjOrmVZyOVymJ6exuHDh/Fd3/Vd+OhHP4qXXnoJiUQCxWIR6XRa9RfridN6r3N4fn51MpmE53lqTjGvHd1SXfTqzirhc3g++uOjfczAZvmy3qNMcayXYfMa9Ho9uK47lGCdSqXQarVU+bY+J1rvW+a4KH7OXAsFru5S8zNsNptYXl5Gv99HPB5XadhBEMDzvKHNF5736AgwfcOBTjcrG4Sdx626FwvCdicaGSCSk/ugcOOR+7BwMb4u8+Lln3SeI845nPmTPF7+yk1ckHBDuZKU7A9s8fB/u8Tz/y2Af3u1C9kugnl6evqC0CXd/aTYohjTk40pWjudDuLxOMIwxNzcHPbt2wfHcdDtdtVc5kQigWg0qkqdWabbbrdRrVaRSqVw6NAhHDly5IL1XA2WZalybo6XYkl0q9XC4uIipqensW/fPnz913+9ckYdx0EymVQJ0xSUej+y3k9M95Xf6wKZrisw3NNMwanPTebjwGavt14OD0CtieiJ1hTMDOBifzHLnnXBq4tqin8Gb+limYFilmXBcRx4nodEIqHel33MPId+v492u41SqTQ0O7rVaqFaraLZbA6N3eI//dqxWoH/HMdR14TXQNhZ3Kp7sSCQxGsW5n/9K2j/jXvGvRRBmAjkPizcCCyjj4dTr+OP8KZxL0W4Qq4nJfuGsl16LQ8ePKjcZZbSApvCmEKHfazRaFSJsdGEZ9M01ZgqCjn2+LLkmG52NBpVAhYAjh49ivvvv/+Swuj48eN49dVXsXv3bhw+fHhLF5qzkoFNp5KjmzqdDmq1GjqdDg4cOIC3vOUtqNVq+MIXvqB6kynsddeW85pH+23pROul1oTvrTu8FLP6uCjduR7tgeb763OtKVAp1nWRPSpm+VlyrZ1OR4WdDQYD5cLr72WaJlqtlkoX52xsPflbF+R6YFgYhirp3LbtocAvblqM9mGPpnUDGHL0uX79/QVBEK4FowcMWq1xL0MQBEHYZmTTbRz78Ftw+IeeGfdSbgiRyz/l1rBd/rhPpVLwfV+V8eq9y+xDpVAEoAKkKOIGg4Eqwab48X1fOYf8rz4nmKnYZ8+eRavVwj333IP3vOc9KBQKAIBarYa1tTW1xsFggM9+9rP4xCc+geeffx6f/vSn8fjjj6NarW55TuyRTqfTSKfTiMfjQ8KWYjIajeI973kPHnnkESSTSVWezPeMxWJKgOup0nrKNbBZXk0RybFaFMsUsXStKa5HXWoei46vLp75Hlw3fw5AubO2bSM8P+OaYloX1DwGhSv7sPVeaJZC66OhOPeaqeLRaBTpdBqFQkH1N/McKcb5Xq1WC+12e6jUm4KamwdM5WZfNgD4vq/mNwPiMAvC7cCH3/YwOuHV5xVEczm88jNXls7ae89RrP9A46rfQxAEQbg4bwTTuPtnVy/7vOQxC3O/8kUE738L6t9bvwUruzFEjBDRRO/yT9wmTIzDTOE1ySSTSVVOC2w4zBRfnuepZGgKL33EFLCRPk2BSBezVCrBcZyhQCc6qwBUT/HS0hIGgwGOHj2Kb/u2b8PevRvppkEQqPAoAKhUKvirv/orPP3002i32yoc6vjx41hYWEA8Hr+gn5mBVhRhvV4PuVxOJS17nodGo4FCoYBYLIa3ve1t+PznP49ms6kEIwWfLvQ9z1MbClxno9FQmwb683X3WO+D5lr5uL6hAGz2+HIN7C/WhTMAJZz5mC6AAQwFZ/Gc9DLobrcL13XVWunC8+fsG2a/uS7qCdOrWQqeyWSQzWbV7GWWY/O5Ojy+noytu+28Rvq5CYKws+mXr3HeU8RAaF9ZK+XAjCBuTf7/PwuCIGwnBjAQtq4gKXsAhEGAMGbAlXvx2JgYh1nvfZ1U3va2tyEMQ/i+r5xDAEoAskQ7Ho8P9asyxIoims5yPp9HMplU4jkejyOVSiGVSqlyYs/zUC6Xsb6+jre//e34kR/5EezevVsJIt/3Ua1W0ev18Prrr+N//s//iS9+8Yvq50zULpVKWFpaQhAEKujL87wLen3psDqOg0QiAc7H9jxPiblUKoW7774bpmmiWq2qOc10oikaed7dbhfNZhOlUgmVSgXtdhu1Wg2NRgPtdluJXArzRqOB1vkyQDq1juMMjWbiNXddd8jh1YUye4rZ08wQMPaUB0GgRDkFKd1x3a2lCNVdZ4pi3YGn+8018P1t21YbLaZpIplMIpvNDq2tUqmgVCqpBGyek+d58LzNGyqDwXgN9FJ4vV88Hr/0fFVBEHYGv3L/A+jjyiu0dlkVvPxvD172eYnXLCx8UAJ9BUEQroQP3X/0ip/7gr8LT789c9nnyX14cpgYh3k0AXgSyeVyKJVKCIJAjYxiGbFpmko0dTodOI6jynmBTTfasiyEYYhkMomZmRlkMhkMBgOsra0hmUwimUyqMm3OEz548CB+8Ad/EHfffTe63S7W19fVfGbP8+D7PlqtFk6ePIlyuaySqVmSTId3bW1NPbdararRURRXFMSWZSkh2+/3kc1m1bmSXbt24bXXXsPJkyfVe1F88nWGYaBer2N1dRUnT55EqVRS7i37plkKbZomfN9Hs9lUI7lY2s5rbFnW0JgnOspMpNbTqelE6wJad371EVsMGmOAl97vDECVglOccsMD2AwDi8fjaDab6jjcNOBzWTmQTCYRBAH6/T4sy0IqlUIsFlNuve5062Ou+F76pgAwPP+ba9XbBARB2NkM2m38l3vuxz/56nNX/qLYpTeo3TcsLPzfTyEcSKCvIAjClTBotfDBex647L34lWAOnzlaRBi0EU0kLvo8uQ9PFhMjmCd9buzBgwfR6/VUr3AsFlMhUBRXnGOs9yPHYjFYlqXcZbrPuVwOjuNgamoKtm3Dtm0sLy+jXq+j1+shnU7j3nvvxbve9S7Mzc3BMAxUq1V85StfwfT0tJq7XCqV8NRTTw0Fjo32CLMsme4kE7cBqOdQ3NEpbzQaKJfLsG0buVwOuVzugs8olUphfX0dzWYTiURC9UJT9LfbbaytreHEiRM4ffq0Eo2pVOqC1Gd9vnSj0VCCmuejlx/zfPTwq06nowS7ZVlD4pafQ6fTGSrf5sZGp9OB67pD5fCe5yn3WB/rpH/WvH58L4pZva9b7+Xu9/solUoAoDYpeE5M7db74fl+TEqnY8/NGb1MneKfzvR2yQQQBOH6CYMAHzxyP/7JZWaUvBLM4U/fdQj4uUsfzwgByB9pgiAIV8Xl7sUnO0X8yf1FhL2Lt83Z5yzs/VdPAeEAOP/3Y+frH0L7h6uYbKW0s5kYwTzpJdl33XWXKkvOZDIqqItCjyKMwoxuJoUMxSh7aVmiPD8/j0wmg2KxCMMwcOzYMUxPT+MDH/gA3vSmzbh5CrM777wTxWJRPW7bNprNJk6fPo3p6Wlks1lV5qv3tAIbPdjxeBxBECCXy6kQKjqz8Xgcg8EA9XodKysryvWsVquYm5tDIpFAMplU7+15HgzDwMrKCpaWlmBZFgqFAhKJBHq9Hmq1GtbX17G8vIx2u636doMgUA46nWP2KlcqFXQ6HeRyOeUCc1OC1xvYdFb1QDG9ZJqiVR9TRdeWJdl0qHlcXi/OoaYz7LquSv5m6TiFMK91NBodcsIZDKf/XvPndNQphhnYxdA4AEOuue4Y62O89OAzHpsBZboLLQjCzifsdvCfDt972edF08PziWMlEwd+8gs3a1mCIAi3FZe9Fw82KwD7a2u4+2csHPvRPcP34dHJQYaBaGSyddJOZ2IEM4XQpGKaJprNJuLxuBI2dBUZ9KU7iwBU3y3LsimmgiBAu92G4zhotVoolUrodruoVquIxWJot9s4duzYkGCmcFpZWUGr1cIdd9wBYGMu9Pvf/37UajWUy2WcPXt2KHGaZb6pVEqJ6VgshlwuB9u2h2YRA1DPp5Drdrtot9s4ffo0TNPEvn37kM1mAQCtVks5teVyGcvLy2p2NM9bD+gKwxCtVkuNqjJNUyU/JxIJdDod1Ot1JT7pLFPM6rOYdReVLiv7w9kTrfeLEzq5DEmjgGYfMIW0Ls71z28wGKjeY56jvjmhB5jpvdB6qFo0GlUl5ywV930f/X5fOeX6WChWA+hp5fyeLjqAoVFddN8FQbiNuAJXuF+t4chPvIGXf34/YhUTB37i87dgYYIgCLcRV1Gh0zt7Dgd+4txNXIxwI5gYwTzpJaS9Xg++76t1UoxQTFF4UejpjiPdSTqLwKb4OXXqlBJXjuNgenoavV4Pf/mXf4lnnnkGe/bswUMPPYSHH34YxWIRxWLxgv5cy7Kwf/9+HDlyBF/5ylfw5S9/WYlSikXXdVX5Nx1LfVwTxSyd4HQ6rdalB0+xVPrs2bM4deoUotEoZmdnsbq6ilKphHa7jVarpRxR27aVK62XjeuCMBKJqIToTqcD27ZVKFen01FzjPkaCkWKRzrUvNb6iCsK5tE0bj1AjNdA/xxZIWCaJjzPU44z19xoNJST3u/3Van7aN80+655PvoYqFarhXq9rq65bdvwPE8dh5+zHiTGc9L7nfXxX6Nl64IgCKP8/9v78zBJ7qvMH30jt4jcM2vvfe/WZrklS7JbBssWyGDAj83v4vsDM2a5/o0HhmEwA3cuM/O73JnhwsydC/gZGAYugxkMBgzYzGDAFjZCY2Ns7ZZaUu97d61ZlZV7Rq5x/6h6v3Uiq1rdkroqs7rO53nq6arMyMhvRJZC9cZ7zns68ws4/H8s9HsZiqIoyk0QefxZdMMPofHjb3AygvKmUcF8E9BJlP2lFF3sZaVwpgiVpcDsK5Vl2xQ3k5OTCAQCOHTokOn9DQQC2LZtG6rVKs6dO4cXXngBv/d7v4djx47hIx/5iBF+wJIgbbfbmJ2dxd13341v//ZvRyaTweOPP27GXwWDQWQyGYyOjq4aKUVYmtxut1Gr1Uy5sOM4yGQy8DwPpVIJ0WgUAPDyyy/jxRdfNGna7XbbzKimq0rXu9VqIR6PI5VKIRwOY2RkBOl0GtFo1Ig/BoXRjaUIZfmzdIEpdMPhMGzb9iWU8/OSwhpYceh5/rkvCm6WcdN1DwQCJohMOuZ8rRSnrDaoVqtmpjNL8cPhsG92N11m13WRy+VMtQLPY71e97nW8sYLb7bwBgd7sOmG88aAHImlKIqy1cgX4zjwg9/s9zIURVG2LLfbdVgF803AgC66gLLMluKN4oaimCW7FH7sc3YcB67rGoczmUwaUUexCsBsSyG1sLCAL37xizhz5gw+9KEP4d577wWwVPYtS7cB4IEHHkAymcTTTz+NyclJRKNRPPzwwxgfH191bCzfdl3XV7rMGcXz8/OwbRujo6PYuXOnef7FF19EuVxGqVRCpVLBwsKCEXSFQgGTk/7ykmq1akZFzczM4N577zXnlMfBZGmKwGq1asQ63Vqe32QyacRy77xrQvdX9nJLIUmBzTJsOf+Z2/GLji4/Z9u2TdiWfJzCmueV6wgEAkgkEibVnP3huVwOtm1jYmIChw4dQjweRy6XMyO/OD6L+5MJ4BTlspeZv0sUzYPe6qAoiqIoiqIog8zACOZBplarGWHMMmUAPuHCUUnsKeUoJ1mm67ouFhcXYVkW4vG4cSqHh4cBwLi6LBGWgWEU4efOncMv//Iv473vfS8+/OEPXzfc6ciRIzhy5MgNj42OZ71eN+4ohanruqZ/NpFIGKH5B3/wBzh16pRJmC4UCnBdF/l8HidOnLjhezJYbMeOHca9ZYgVj4fnkyOaHMcxpdc83zIMjKKRAhZYcZrpXMtUbjrSFKGtVsv0GdOF57HLEng5J5nhbqw0qNfrxi1n6Tmfb7VaaDabphKBTjJvjExMTGBoaMg43RxDJtO+ZZUDBbu8ycH34rGzFFxRFGUr0fUG9wa8oiiKsvkYGME8yKm+9XrdlPNSePUKs0gkYgQNHWXpVMqUZZYQU0gx/IoltTJkSvbixuNxZLNZzM3N4e/+7u8QDAbx3d/93UilUm/42GRwFEdgxWIxIxZ37tyJ8fFxI5a/8IUv4Pnnn0csFjOjrur1Oi5cuIB8/uZ7KxzHwdDQENLptK+XWbqkdNhjsZhxoCmi5TgvClR+LgzQkhUAAHyOsQwi400N9gKzJJo3QICl0mr2fcseY4pS9iTX63WUy2VTKcAgr15Rz95qim3eMKBYL5fLpmycnwudZv6eyC+ePym4B/m/KUVRlPWgUIniwIdf7PcyFEVRtjTd7u31N+jACOZBh86mdP8A+JxlwC/81wplAmDKiKWwOnDggHEF6SZL0cPEbc4qdl0XZ86cwcjICO655x6k02kz2/f1EAgEMDQ0ZAQ+AFPyLPE8D1/72tfwzDPPIBwOo1arYWpqClevXsXx48df13tGIhEcOHAA27dvNz3YUuyFw2GEw2Gk02kkk0nEYjGTpk1hz+07nQ7K5TLq9bpxoymI2YtMQc7Seb5nt9v13cSwbdv0GfPcUMDz85NCnedKzpCmeK9Wq+h0Oka4MziNAj8UCiEajRrxzGTvaDSK4eFhzM/Po16vm3Jx/p4xPE7OYZbrYz80AONyK4qibCWsUAheu33jDRVFUTYDlgVvEw1hLlSit1X/MjBAgnnQey3b7bbpiaVw5pxgzmSmAKtWqyYwiinJUkhTdDUaDRQKBTSbTSSTSZOAXSqVkEgkkEwmEY1GjYiLxWIm8GlhYQG5XA4XLlxAOp1GMBh8Q4KZsPyb0BH1PA+zs7P42te+hjNnzqDZbGJmZgZXr17FV77ylTf0PgcPHsT27dtNgFZvUBpdVzrLtm2bbWTSdbPZRLlc9s10pqhlgBlvPsgyevnZtcUfVQwA4/l2Xde8jmtjuT1/X5nKTXHLMny+nu/puq4Rus1mE/F43ISe8WYAR04lEgmMjIwY51q6ylw3+71lEJ28mcCqB6abK4qibAUyiTou/MHd2PcDL/V7KYqiKLeE5nc8gObHNCG7nwyMYL5eevOgIHuKXdeF4zhG/EgnmXONKWwcxzE90OxZlfN5h4eHceDAAZ/rGQqFUK/XEYvFTGAVZwdzznEqlUK1WsXU1BRGR0cxMjKCsbEx35rpwFJgynRtIkuFOZapXC5jenoaly5dwrVr1zA1NYVCoYBCoYCZmRm88MILKBaLb+g8Oo6Du+++G6lUyohXlpzL8UkUutFo1DircmSSnLXMUVd0dCmYKTDlTGXeyJAinedEOrJyLfyZfcssledrmVjNhPBqtWpGZHEtFLfdbhe2bZsxXyy3TqfTxlkPhULIZDIolUo+JxuA73dE9k0zZA6ArxJBURTlDRMIohvWfmBFUZS+sQmvw63mwMjLW8bAHBFF26BCwcSQKJb99oZBsQS3t6+Zr5cjphKJBMbHx7F9+3Yj8FhSK11Qij0K4EAggGQyaRzqhYUFLC4uolwum1LqbreLF198Ec899xxisRj279+Pffv2YWRkxCciJycnce3aNdTrdSwuLmJ+fh6zs7NYXFxEPp9HrVbD7Owsnn322TddBRCJRPD2t78dExMTZo08RjrDsVjMPM7jZb8whSDnM8sxUrKvWc5sBuCrDODj7Efu7UnmueH55/75XrzpIG80lMtlI7b5eZXLZQArpfSA36XmuK5EIgHbtpFKpcxamMS+sLCAfD5v1sr1U5zTSeYNBTlbmmXZiqIo18WyEBweWvnZ8+A1V4IC6996x6ab+9n1LLTcgfnTRlEU5eYJBBEcygDNFjqlEgCg/e6jm+o6XKo5OPiPbq9ybEAF800je5R7y5dlUrF8TgYyMYGZ7nMgEEAsFsPQ0JDvNUyM7nQ6aLVaqFQqpgyY+6RYSyQS6HQ6mJ+fx8mTJ2FZFg4fPoxIJIKXX34ZL7/8MsbHx3HPPfdgx44dsG0bc3NzuHjxIq5cuYJ8Po9cLoeFhQVUq1WUy2XMzc3h9OnTmJmZ8R3Xm8WyLIyPj2Pnzp2IRCKwbRuZTMaUXctRTwzt4mxjObrJ8zzUajXjLvOz6RWlPNcUtzLsi8KS20gnWgpibiNdaD7GsVvSOe52u6hWq0Zc83Ok8yx/VxKJhCm7T6fTyGQyZr/VahXNZhOZTAYLCwsolUq+0VZylrf8nWPbALdTFEW5LoEgOu96K/7mj37XPPRqs46PnviI2Ghx49f1JilVHRz+0ef7vQxFUZSbw7IQzGRg2RGU374HX/3N38bPTN+PU989vlT5Gb29wrM2KwMjmHtF6CDCMUTsS6ZI4mMUYtxWup2cLyxTjzOZDNLptM+5ZmgVRTWF8+LionGtWb7OsU7FYhGdTgeLi4s4ceIE2u02rl27hrGxMUxMTGBxcRHVahV79uzB+Pg4YrGYKbk+f/48rl27hjNnzmBxcXHdblyEQiHs2bPH9CSPjY0hk8kgHo8b8RcOh1EsFn03CNjDy/PBc9LrGHMGNN1gOrX8mSKYYpximq9hhQDfSyaiyxnOkUjEBLV1u11UKhUj5FmezxAxYEXMszRbJp+HQiGk02mMjY0hkUiYfXAutpzbLW8KMLxMCmX20XPtmpCtKMprETy41yeWFUVRlA3GshC49w5k/usM/mjf35iHf2XbC8ALwDONFv7l2Q8B7cHXSMBShY9bsfu9jHVhYD6BQXfE5DxbOYOXc3sp8GTKNV1QvjadTiOdTqPZbBrBLG8UUMCxn5mlxLZtIx6Po1wum37VcDiMbreLZDJpyn4ZyNXpdJBKpXDo0CGMjo6aMDK6s7Ozszh16hS+9rWv4YUXXtiQYKh4PI4dO3Zgx44dGB0dRSqVguM45hyyB5fHQqEq51kDMC4wBSkA340KClHZt8wSevl6YOWmBqGTTUGdSqVMWJcUvHw/CnD2EVP087j4uTebTV8YGH9nHMdBNBpFLBYzaw6FQuYzlfOeAaxyyPk4j1kGmKlgVhRlTQJBhPbugrszs+qpsNVF2nZRaUWQL8eRjmtooKIoyrqwLJaz/3UGf7TvyTU3ecgO498e+Dx+7vT/tsGLe2PUGhEc+pHbs8JnYASzFKSDCPtaKZSYpMwAL2DJTeRsXr6GbmgqlcL27duRTCaN8KW7DMD0tlI4UvzQcXYcx1duG4lEkEgkMDo6ikwmAwBYXFxEqVTC7t278cADDyAajaJcLsPzPKRSKbTbbfzFX/wF/vqv/xpf/OIXX9fc5DdDKBTCXXfdhf3792PHjh1GDMpRVhSAFJxyBFQkEvElYdNdlkFxPOcUzhSX/DwoSNm33NuzzNdzH9K9BrAqRAyAEdJSdHNtwWDQzGZmiTVfy+PmmuhAM1CMn20qlUI0GkU8HjfCXb5GrkX+POg3nxRF6R+h7RP466/9zzWfOxyO42/u/Cv886kH8c3/3/2Y+34L2WRtYxf4Juh6Fhr5aL+XoSiKckNC42NIv4ZYVgaLgRHMg9zDnEwmTTk2hR7LsSm2ZM8ry3spvEKhECYmJrBz504jkNmXypJhliZTKMrAKzm/t16vw/M8U849NjaGbdu2meTsbDZrBNzi4iKi0SgikQhOnTqFT3/60/jjP/5jzMzMbOj5Y+gYy8EpYCkigdUOPsUqy9QB+OYfAyu/M/I8cz8UpnSF5edHcczzJMuuARjhKYO1+DlT3DKlGwBqtRpc112VZM0e7MnJSbRaLQwNDZnSfb5Pp9NBoVAws6aZAM4bLVxfb0k3b970jiuTLraiKIqPQBDu4YkbbvZr25/FBz8+hMi/24nqz27Aum4RlbqNwz/2TL+XoSiK8tpYFloHtuGP9v3ea27W8FqY64xszJqU12RgBDP7PgfRaT506BDq9TpqtZpxP9mfHA6Hjespe4+l0xmPxzEyMmLmJdP1patMKJI8zzNOK0dVceSTnO3MpOUdO3YglUqtWnc2m0WlUsEXv/hF/MZv/AaeeuqpDZ/LGwgEsH//fuzatQvpdNqXMC3dWxm41Wg0TDAYhXM0GjUzr+V8ZN64YHl1r7sv082luKRr3bsOPidHR1GI1ut1I3ilEOc6WCLNz7Fer2NhYcEkZudyOVMi3+l0UKvVEAgE0Gg0fHOZZT+ydOC5TrrY0kmWIWZyTrSiKAoAIBBE69Gj+Lvf/+RNbZ6K1JEbCd94Q0VRFOV1ERzKovXvF2643XONIH713GMbsCLlRgyUYI7H4ygUCv1eyioSiYTPDZVOJ4WTLKVutVqmZJgltHQQu90uXNc1fcjcnwxq4qggCiS6iwyA4qxky7KQzWbNKCmuh/ssFAr4zGc+g1/5lV/BuXPnNux8SUKhEHbv3o2hoSEzexlYEX3XC6jijQOeM/Yt0+WVYWoUoJyBzdfzvMr+Z7qydJ9lDzWfp5vMkmtg6feTN0so0OVMZL4P3eFarYZSqYTFxZWU2Xa7jVKphGw26xsVVigUUKvVkEgkACyNqapUKr407N5RW8BKn3LvzQDZ360oigIAwWwaf3OTYvlau4LRSAVnfmR+nVd16+h6FtypeL+XoSiK8tpYFrr7dwAo9Hslt5SuZ6E2lej3MtaNgRHMlUoFQ0NDAymYKZDZd8peU5myLLeVpblMVuZc3Gq1ilKpBMdxfHOcuR/22LIHF1gKAePjruticXHRBEMlEglf+S2/z+fz+OQnP4lf//Vfx9WrV9f/JF2HsbEx7N27F8PDw6uCqWQvMYVpb19zb9kx3V+6uRTTa/X08jkZANY7Pqp3XJgUpOyRZo+6HPfF/fB92YPMUul2u2161QnX7bqu+Zel4xT+3W4XpVLJ9MgzFV32tcvzI3+W54k3BBRFUV4vv5x7N575/z4A/NDmEMxdz8L8uWEc+qmn+r0URVGU62NZwNvfgsLP15C88dabCrcVwqF/9nS/l7FuDEyUbrlcxtjYWL+XsSZyri3TsSmaKUoofCj8GNZFccW+10KhYNx027Z9/agAfEJZOsyhUMikZdOdHhsbQzabXbXemZkZ/MZv/Ab+83/+z30Vy8FgEAcPHsTExARisZhPwDHlu1fw8nGeS5lwzc/Btm0kEglEo1GEw2GzjQy8opjl98BKCjZFOAUp9w3AiFPeJOE6mEIut6UQbzabqNVqJg272+2iVqutefOn3W6jUqmYGx+zs7NoNBpoNBqo1+soFotYXFw0veq9s5aBlYAveWwSWSquKIryeqh1m3gwcRHdTSKWgeU/1FQsK4oy4AQSCRR//uaCFGvdJs42b5w5oWwMA+MwLy4uDqxgZvkvxRhdUAophknRmeTj7MvudruoVqsAlm4MJBIJZDIZ06NLZ5JjlWSgGOBPfA6Hw0ilUhgZGcHo6Kgp+QaWHMdLly7hE5/4BP7gD/4ApVKpD2drhdHRUdx7770YGxszQpT9vXSXKZYpeqVwpQhmWTIFqW3biMVipme3VquZUmjuj/uh00pHWX5mPM/AyvzmYDAI27ZNybXsgeZ6u90uGo2GeZ49zhT6ruuiUqkYZ1rCELdqtYp6vY5oNGrK72XJOMuqpSDu/ax5Prl+YKXCoVgsqsOsKMrrotZt4n9Ut+HXz7+n30tRFEW5vbAstO8/CGAp16bdDeB408W9EWfNzV9pWfitC+/awAUqr8XACOZcLofx8fF+L2NNKL4o6ijO1hI1UvjQgWTiMgDU63VkMhkzg5m90Sy55msoiKS44xrGxsZw5513Yvv27eY9u90uTpw4gV/4hV/An/3Zn23k6VmTZDKJ97znPThw4AASiYQ5dzw/vOnAtctyZ5k8Lp1mCsRarYZwOIxoNIpoNArLsszYKQA+p7m3N5n74fsweG0tZ5oCleefn4V0nzlfWfawV6tVE/S1Fp1Ox6RqRyIRX0K6TPtuNpu+pO/ekC/eCJBzqwH4gsMURVFuhs0qlrueBfdkpt/LUBRFuT6Whc6770Pl/75iZNVbYfz0uf8rfnH//4BjtXHUts1zla6Ll929fViocj0GRjDPz89jZGQwo9Nt24brukZcUcDRKZVl1FJQyQRsJmwHAgET0iUd6UgkYoKrAPhKbln2yzCw3bt345577jF90Z1OBydPnsQv/dIvDYRYDofDuPfee3HkyBFEo1GfW97pdNBut00KuAzMolMMwDjuLG2XQrbdbqNWq5mwLoaBsUd4rZsZsl+Y4pZCkzcobNs2gV4s9ZbztpvNpi+NWs7GZhUCw8J4g2QtGo2G6WMfHh5GKpUy++XvTblcNvOb+TvVm+Yte9flY7ypoyiKAstC55H74CZCAP5u1dMNr4WnG2Fcam7Hf73w7g1f3pul1Q5i37/6Rr+XoSiKsjaWhda33Y/avyiueqrajODjp/53pG0X/2bfX5nHzzb24r9d/JaNXOWboutZaLyS6fcy1pWBEcwzMzO47777+r2MNYlEIqhWq2i320bY9s7GZck2hZ6cwwysCF/HcUw5sRQ5vQKHYk7S7XYRj8dxxx13YPfu3ebxM2fO4Bd/8Rfxmc98Zp3PxI0JBoM4cOAAHnzwQQwPD5vSas6mDgQCZmY03eFms4lyuYxGo2HOq+xvBrBqVFKj0fDNR+5Nx5bhYHT56SoD/nA2YKWnWgZn8SZI77/S0XUcB7VazdwI4Pu81vguOtO2bZs+dtm/3el0TB8z17rW+Kve8WVMHl/rd0dRlK1H670PoBu04P7kIkKBLr5UW52gn+sMbzpXWVEUZbPQ/I4HUP/ni6+5TbHh4F+e+r4NWtGtp9UOYu//eXvfuBwYwXzx4kW85z2D/T/tYDCIcDjsE2ssE15L/DJQin2xrutieHjYuNIyJRqAT7D1JiBTCO3atQsHDx40a7p69Sp+67d+C5/97Gc35iTcgIMHD+Lbv/3bceTIESQSCd+c6kAggHg8br6AJeHbarXguq6vJJnnlaXbdIEJXVyWc/PmBUO3pONP4S3FNx1kikw5YoriU4pg3gQBlsrq5TxpuuXlctkX/PVaWJYF27aRTCYRi8WMU95sNlEqlczvlLwRsJZw58/clnOdtRxbUbY4gSBqHy+YH9vdAH7+zAf6t551IhDoovmdDyLy+LP9XoqiKIoP93seQuOf5vu9jHVnK1yHByYl+8KFC0ilUv1exmvCclsiXVCKMVk+TLHD0UBMx45Go+Y5imfuj6Kt1WqZYCmWfg8PD+PgwYMYHh5Gt9vF1atX8clPfhKf/vSnfevqF7t378Z73/te3HPPPRgeHkY0GjWlzxyvFYvFkEgkjLPa7XaNCyz7iHluZS8vsDIWiuXVtVoNlUrFlEZTsFYqFVSrVfN6Wf7O9+jtNZeP944Kk4/Jmx0ypIwC/GY/i2g0inQ6bRLTw+Gw+V2JRqNIJpPGRebvhmwL4A0W6chHIhGT4K4oytal9oEH+r2EDSEc7GLhxyr9XoaiKIqP+ge3hlgGtsZ1eGAEcy6XQzQa7fcyXhOKIbqlAHxlvhQzkUjEJCkDMG6m4zimJJvCjSnMFIcUZq1Wy4ybojAaHh5GOp1GvV7HlStX8OlPfxq/8zu/g3y+//9Bbt++HY8++igOHTpkAs1YGi37vumCShHI45QBVhSj/GJqtQy7ktvx3FHcypsN3EaWNkux2+vyyxsgLCmnUKcLzhsdFKwU0bZt35RYZQk2Z2wDMKFkLNXmKDLZNw3AV5rOY5DtAYVC4YYOt6IotyfV/8vbUf2+t6P1jxf6vZQNIxToov7Bh/q9DEVRFABA9fvejubH+v+3uXLrGJiS7FarBVskxA0SMrCKzijLgSUUexwz1Gw2TRk3BRL7VOUM516nkPN+KSy5rWVZKBQKyOVyeOKJJ/CpT30Kk5OTG3syegiFQti2bRve/e5346677kI2mzWjsSgu6Z7SWafwpBsM+AOsCEuje/uYeS4ogoGlUC8KWW4LwGxLUSv7f/k5sey7V7TLQDCWbbM8nJ8lhbrsMb4Zh1mKbdlzzCAzJmVT+MtzJBPUpZhmevtr9U8rinL7Uv7+d2yq+cm3ikiog/KPltANvh3xzz3d7+UoirKF2arX4VCgi+r3vR3xz96e1+CBEcwAfGWwg4QM+KJ46Z25TDHVW9JLl5X0jgfqHRnFnl6WCcuU6IWFBZw+fRovvfQSPvvZz+L8+fMbdQrWJJPJ4MCBA3jb296Ge+65x4jl3v7bRCJhksGBpXNSq9VQq9VM8jQA32gt2ZMLwJdGLV1mbsPHmY5N1549yrZtm35k2XPO95RBbnJmM0PcuG7pPsveZx6HnI38Wsh51CzlpnBmr7QsRed78nes93eIPdZaiq0oW5PiP3oH8OGt90cascNt5H+givjn+r0SRVG2Klv5OhwJdVD4cAXxwYhUuuUMlGAe1FE4FCkAjACjsJOjiSh+pGvKxwCgWq2i0WiYUUvSMQRW+mfZm9tqtZBMJhEOh+G6LqampvDSSy/hf/2v/9V3sTw+Po777rsPDzzwAHbt2oVsNusriZajl5iIzRFNrVYLpVLJjIbqTcOm88qy7l6hLAU24O955mdi27ZxsOVsbNlbzs+s9/PtFZ+cEc3H+BnxZgirDWS5942Qqd0U9lwXf242m77/JtZymvkvw8dki4CiKFuD/I8eQ/BDuX4vQ1EUZcui12EgHOyg+I/egfSnn+r3Um45AyWYy+UykskkyuVyv5fiQ4Z7yVLq3rJdACYVWpb2UvDRVWW4GQUcxVMgEDAjh/hzIBBAvV5HpVLB/Pw8vvnNb+LatWt9OxcAsG3bNjz00EN44IEHsGPHDsTjcVNOLsOoIpGIScqWYWalUgmVSgWtVsuIYilaAfiEbS8yOVy6/jyP3Ke8WSHnW7PPuLdvnEJVusTcF9fI/nKKapZmR6NR1Go1s48b0Wq1UKvV0Gg0TC8z37fT6cB1XTSbTZ8DzxnNXIscKUVnemZmRvuXFWULMf+xYwh/cGv/kaYoitIvFv7xMXgW9DqMpUqf0gcrSH+63yu59QyUYM7n8zh8+DCef/75fi9lTXodRCny1hoHxcAmKaDm5+d9M4i5P4pu13VNrzRLuSuVCmZmZvDqq6/2VSxHIhFs374dR48exf3334+dO3cikUiYkDMp4CKRiDkGOqCu66JaraJcLptzIh1pCtLe8mOWV8tzze2BlRsaFMDASvp4KBQybjDPMV8nRacsI+dNDsKbJdLBZl8zBTO/ehO2r0ez2UShUECr1TJhcBwHJXu8pXsOYNW5kk53JBJBrVZTh1lRtgC5Hz8GL2DB/u65fi9FURRlS5L78WMIfc88Apb+3UXscBsL/8cxDP/O7TWXeaAEcy6Xw7333juQgplOIwBfb23vTFyWHUshJlO1c7kcEokERkdHTQKzLEdut9tG/HB/pVIJ09PTmJqa2rgD7iGbzWL//v248847cejQIezcuROxWMwnPlmmLgO+KERbrRby+bzpy5UhWbJHmKKUs5h7hSLFI/uLpYDka2XpshSPdL/5OoaxSZFKkc/PguuTwl6um+uUQWE3I5jr9TpyuZwvqVveQAFWQs5CoZAJkOstL+c66OKrWFaU25/Zn3wYke/M6R9pPdjhNnI/dgyjv3V7/aGmKMpgMffPHkY3CNjv1etwL3a4jcJ7Kxj+nX6v5NYyUIL5woULuOuuu/q9jFU0m00jTqSYa7VaawoYAEZEyVm/nuehVCphcnIStm2b0mw5q5gp3AwDq9VqyOfzuHjxYt9mLY+MjOCOO+7A/fffjz179mBkZMSIZSke6XjSeQ0EAkYgV6tV1Ot1E7JFAUinVo5tAlbfaJCBW1I8s/dX9jNTaDIwjaFaUhjTtZa96PycuD6Kz2azaWZKh8NhnyDvLdmWJfuvRb1ex+zsrDkvPA9y9rRcA89z7zxpnp9gMIhcLndT5eCKomxeZj7+MCLfro7GWtjhNgqPVoDf6vdKFEW5XZn5qYcR+vZ52EFtf9tK3NAKsyxrl2VZT1qWdcKyrFcty/qp5ceHLMv6smVZZ5f/zS4/blmW9WuWZZ2zLOu4ZVn33+xiTp48iT179rzxo1knZPCSLPuVo4BkUrMcLxUIBEw/KkUOR0OxBJdOJUdQtdttlMtl1Go1LC4u4urVqygWixt/4AB27NiB++67D29729tw4MABjI2NmTnSFMZEPsb06FarhXq9jnq97nNfpWvL1wLwucVyjJR09CnOpVCWNyUoiimIZco535vI/nSZhE5RTgHKz5fPcR90gCmWr9d33Uu320W5XMbc3JzpfSaNRsMExMnSdR6LPF55TovFovYv38Zs5LVYGUymfnZJLAcD+t+5ovQDvQ5vbaZ/Zkksh1UsbzluZo5TG8DPeJ53F4B3APgJy7LuAvBzAJ7wPO8QgCeWfwaA9wE4tPz1MQC/ebOLOXHiBHbu3Pk6lr8xyLJgOapIlmbzeWBlfjCFVbPZRCgUMgFUrVYLuVwOuVzOuNd0TJvNpkldbjabmJ+f79us5fHxcdx111249957cfDgQQwNDZlScVlSzZ+BpT7nYDCIdrsN13XNV+88aVluDWBVmTbPIXt0KUqlkyz31+vs8nxyv3w9APM58LPke0rxT5efrj+FNUO4eDxSUMvxYDdDqVTClStXUC6XzXvJ3wOuXZb9986a9jwPtm2bVHXltmbDrsXK4DH1Lx9G5BEVy4rSZ/Q6vIVpHyupWN6i3FAwe5437XneC8vflwGcBLADwAcAfGp5s08B+ODy9x8A8PveEk8ByFiWte1mFlMsFpFOp1/fEWwAFCK982+lwKOQolii48jRRizn5evovNKxdl0XrVYL1WoV1WrV7HNhYQH1en0jDxcAMDExgTvuuANHjhzB3r17kclkTLk4sCI2e91WCkceE4XcWk6wvBHRO9dalroDS+fTcRwj2HtnPcv+Xf7L4CwSDAYRiUR8idhECmdWCrDHvDfkjQKc5dh8v14BfiMKhQIuXryIYrGIYrEI13VNgrjrur650XL/hDceQqEQ5ufn+1ayr2wMG3ktVgaPzkP6h5qi9Bu9Dm9drv2rhxG1mzfeULkteV09zJZl7QVwH4CnAYx7nje9/NQMgPHl73cAuCpedm35sWnxGCzL+hiW7rb5uNmS1o2k0WjAcRxf2S77TeUsXAqs3gRolhY7jgMAxjktlUpoNpurRBxHDbE0e6NJpVI4cOAAjhw5gn379iGVSsFxHCMM5Txi0lsqzGORvcM8d3xMimQKUxmgxV5jKbDXCtXi/npDsyi+6TTzMf5Lsc/jaDQaiEQipnxeztiW28nfA9nDLcu/bwaO2Mrn8+h0OsjlciiVSpiamkKpVDLnQ96kkT3c8rOo1+sa+LWFWK9rsYPY+i1aecNc/T8fRtSe7/cyFEUR6HV463DtXz8M+x0LetNyC3PTgtmyrASAzwH4uOd5pZ5yXM+yXl8Cied5vw3gt5f3bV47iD2Yrusik8n4yoUpAKXoYxkvsBI+ReHLcCpu02g0kMvlMD4+jmg0CmDFRfWWRzAtLi6iVqtt6LE6joPdu3dj//792LVrF4aHh80ILIpcikUK0d4bAVL8U8RJ95nnR7rV0qllWJecb82AMDq4/CzYq8z99JZ9S8HM9+NzcmYzxWm73UY0GjXvJT9P3gAgPAY5e/pmErIlrutiYWEBlUrFzN9eXFxEoVAw7rgMOZNimb9PxWJRw762EOt5LU5ZQ3rXZQDp3lvWP9QUZYDQ6/DWovWWKpIh/TtrK3NTf91blhXG0oXhDz3P+/Plh2dZVrL8L4dBTgLYJV6+c/mxm6JQKGD79u03u/mG4LquT+x2Oh0zPooCieXWsmw3HA6b8mQKMMdxzOgouoMUO5FIBNlsFrFYDJVKBblcDoVCYUOPdXh4GPv378fOnTsxMjLimwctA7UArCo/5rHKUmg+Lx1iIoWvDOSS4pBit1eQSoebnw1d6Xa77RP1gUDA5+LzcdkPLZ3hVqvlE/1ydNT13l+em9dz04fzmIvFIsrlMqrVqinPBrBKsMuybGt5/vLi4qKWY28RNvJarAwGl//9McSjjX4vQ1GUZfQ6vLW48v96GIm42+9lKH3mZlKyLQCfBHDS87xfFU99HsAPL3//wwD+Qjz+Q8vJgO8AUBRlKjfk7NmzeOtb33qzm78ppGB7LRYWFhAOh00ZNl3UtfbTm8BM8SdHKnmeh3g8jm3btiEWWyq9icViGBkZQTqdRq1WQ7lcRqlU2tAgp0QigZ07d2Lbtm0YHR01o5R4XPJcycRm3kQgskSaQpbuLJ1kjp2Srj33K8UhsJJkze169wf4Q8BkIjb3wxJrpl3THZbOuZyrLIO9OD5MBr1RHHM9dK9fTw8z183zC8CEvkm4XgaqyZsAlmVp2NcWYaOvxcpg0D1UU3dZUQYEvQ5vPVqH64iou7zluRmH+Z0APgLgUcuyXlz++i4A/xHAY5ZlnQXw7cs/A8AXAFwAcA7AfwPwT292MZ7n4eWXX96wWcy9KdfXo9VqGYHXm4BMscTyXTlXGIApYS6VSggGg6YXGgDGxsYwNDSEWCyGTCaDoaEhpNNpNJtNtFqtDS/HHhsbw86dOzE+Po5kMumbpQzAOM2RSGRVfy3d91arZUQrHXiKQik8OWNYurzAijCX7r0Uv9INliOhKNilCJautEzLlp+RLMfudcXD4TAcxzE3Ovi+FMpcEx1truf1uL0yiKzT6ZjAL9mPLJ1xx3EQDofR6XQQiURMcJyyJdiwa7EyGFz8pWNIJjY+9FFRlOui1+EtxKX/9zGkkhv7t7gymNywh9nzvK8BuJ5l9m1rbO8B+Ik3uqCXXnoJ3/u93/tGX/666A2uutG2siS32Wya3l1ZAtxqtUxPMsVgKBRCpVJBt9tFLBYzwVLdbhe2bWN4eBg7duxAJBJBOBxGJBJBp9PZUMFsWRZisRiGhoYwNDRkhJkUoywxpyCUqeCyB1jOP+YNA+lGs5RYlmnL3tze/mN5Y4Pvw/MtU7B7b35w37zhIV8vy7K5bim05VosyzKuL51dngMev7xJ8HqC2lhG3u124bouisWiOSY515trlD3jiUQCJ0+e9J0D5fZlo6/FSn+5+B+PIXFXXt1lRRkg9Dq8tWjtbqi7rAC4yR7mjeQb3/gGduzYsSHvJct+bwRLsqWwo4im2KGDKsuHWe7barVMyJTjOKjX66jVarAsCzt27MC2bdswMjKCHTt2mNFaHC+1EWQyGYyPjyObzcJxHNi27ROqDDiTCdhy/FOz2fQFclH0yrJ0Pi7DuWTytjx3PL+yt1ieU65J9g/zSyZty5Js3tCQfeeyjLt3PFbv7GdZzk0hy8f5O8FRWjcDX+e6LiqVCsrlMur1uilx7y1Jl259IpFAMpn0BaspinL70N6uf6gpiqIoyiDwusZKbQS1Ws309Q4atm0b0UUBzVFErVbLCMbecuNOp2OEVDqdhm3byOfzaDabiEajGB0dxfDwMEKhEKLRKHbs2IFoNLqhQiidTmNoaMiUYgPwOcYyeToYDJogNApp6c4C/pnB7L+l6JOiV4pUvk4KQ24vHWZZii1LuIEVR1imZQeDQRO8xsA1inw5wkqmnvMmB49RPidvBsj0c/5u3OxNmGQyieHhYXOO6/X6qpsOwEpfNz8Husz8WVEURVEURVGU9WHgHGYAqNcHr2er1WqZXlOKFPbjUrz19uMyjIniq1qtwrIsU1JbqVTQ6XSQzWYxMjKCiYkJpNNp3Hnnnat6pdebRCKBTCaDVCplBKUsO5bJ13KUkhSRPHYpantdUhme1jtfWPZ+UwzK0U48xywVl8KZ793bu8yUbzrIXDePQ4aTdTodhMNhXwk5E9GBFYcZWHHc6QRzP72BXa9FNBpFKpUyydj1et3nVvP3SI4l4/mLx+O4fPnyQP63oijKm+P8r7wD2Wyl38tQFEXZsuh1WJEMpGDO5/MbFvx1s9CNpOiiqARWhJR0PSm6WA7carXMbN1QKIRIJIJ8Po9qtYrx8XGkUikj7jZaMMdiMaTTaSQSCSPIZLkxe5fp0EpnlsfHMVt0p6W7LAUqy6RlgrbsWaZLy9Jq3oSQM5h5bhqNhhHTcuwTBScFMfdr2zZCoRBc1/UFc8mUafk+tm3Dtm0zs5lztVmuLRO9XdfF/Pw8Ll++fFPnfHR0FPfccw+CwSBqtZqpOOgdW8XfKd4kCIVCGB4ehuM4WFhY0P5lRbkN6Q63tHdZURSlj+h1WJEMnGD2PA9Xr17FI4880u+l+CiVSmuOPgLgS00GVtKcpTsaDAZRrVbR6XSQTCaRSCRQLBZRKpWQTqd9JcXpdBqpVGrDjo3vx3Ls3hnDdFClqOVaZWAWhSdnTMuyYrkvjuVi37PsGWZ5NFOt+SWd7OslYsv3Ylk8xazsxY7FYojH40b0UlhT7NNlBvyCn//Knmq53ezsLCqVm7sbGY1Gkc1m0Wq1TP87sCL8ZeAX4Xrj8TgajYaOk1IURVEURVGUdWbgBDMAvPLKK3jggQf6vQwfMzMzJt1Z9vZKJ5Zjl2S5sAwJazabaDQaiMViGB0dRSKRQKVSWTW3NxQKYWhoaMOOLR6PI5PJmBFKsv+XYpiPU1yyPFiGTtFtloFVa5VL82dg6ZywR5iP8/zJ+c48n+wV53YUzFLEchtZ8s2RV81mE5FIxHyOsqdaltXThQ4EAmYUmCxV54gnVg9QmN8MQ0NDuPPOO9Fut1GtVs3vAM+dLH0H/Onf0WgU6XQaU1NTWo6tKIqiKIqiKOvMwAlmz/Pw9NNPY//+/f1eio/Z2VmfsJVpyzIAisKOY6OAlR7gTqeDcrlsRkcFAgGUSqVV45DC4TAOHTq0YccWj8eRTqeNw8yybIpj3hhgqbYsG2bqdyQSgeM4xtntdZalawusJFEz3EwGggEr5xSAz6nudZ35GunQ8n2kuy+TuV3XRaPRWOVKy8oAin+JFO98L77f7Owsrl69elPneu/evYhGo2g2mygUCqYEvbeMXfZx84ZMKpWC4ziYn59Ho9G4qc9XURRFURRFuTnO/vrbMTJa6vcylAFi4FKyAWBqamrDRkvdLCzZZb+uDKdiiJUUcyxXtm3b9PeGQiGUy2XUajUT/FWr1VYJzGg0uqEOezgcRjweRzQahW3bAJb6gzlaqtFoIJVKGfczEAig0WisKrnmjQHZ6ytDsmRIFwO26O5KoSvPH9+DoVzS0Zb7702rZh8yPxeWmvPYmHTOdTIRW/ag872kO87zw/cFltzo2dlZzMzM3PBcx+Nx7Nu3D51OB7Ozs8ZhB7DKlZduPG86RKNRzMzMbOjIMUVRFEVRlK2CF+0gYOkUEmWFgXOYAazq3RwUKI6BJQEoy7Ip6CjQOCIKWBKStm0bhzmXy8FxHGQyGTQaDbz66qsAloRcPp9HMBjE+Pj4hhxTPB43pcUseeaaZU8yj59l2lJE0unsdru+RGp5o4ClzPKcyVFJcrYzy7ElfJ5l1fJcM5iMSDdb9klTeErx3ul0zGgwWQ7d26NOsWxZlhHc3G+1WsXk5OQNRzwNDQ3hHe94B8LhMGq1Gkqlkin7pkjmeCu6+bL0nNucPn0a5XL5jX/oiqIMLGd+8yF1NhRFURRlgBg8VbrM4uIixsbG+r0MHwsLCwBWQq2AldJiimRgqVybgkcKPZmmDACZTAatVgtPP/00gCXHlM8x3Gm9ofsajUZ9JeZS/HGGMkWqHAvV6XQQi8WMmw7AN0OaYlCWoVNwS0eZglqmPvMx13V9zvJaPc9cO9OsKdzpYnPGMbB0M4BrpovLzxFYCQzj/ridLEvnzQXXdbGwsIDTp0+/5nlOJBJ4y1vegmQyiW63i7m5OTQaDV+6N9+f54eCXIbHLSwsoFgs3vSsZ0VRNhnhrjobb5BSzcHeH3zta7GiKIqyftyu1+GBFMye5+H555/H2972tn4vxcfc3Bzi8bhxiwH43D+KyHg8btxYCjop+IrFIqrVKlKpFMLhMJ5//nkA8ImmeDyORx99dN2PiaOsKHgpaHlsdH8pRClu6Ro7jmMcZjk3mC4pxS2FJs+TFKgsgW61WqZ3WI7VksKW7rUU2Sy/BmDWKsVoMBg0Yp7rlC43e7ApvGXYFsu2uX/2PodCIdRqNVQqFeTz+Rs6vo7jYNeuXQiHwygUCqjVaiboi8dE1gpN47HNzMxo2JeiKMp18Fo6ak9RFKWf3I7X4YEUzN1uF//wD/8wcIJ5dnYW0WjUlGNTGEqxRrHHBGnphlIolkolLCwsIBqNYmhoCNPT0+Y9IpEIyuUyUqkUvud7vmfdj4liUiZGU2R6nmeSoWWPNkOqPM8zwpTbUMxJEStDuSicKZBZrs7SaWDlxgGTrYEVMS5FspwNLR1xOtoyHIxrYdk1n+NxAks3D1h6Lber1+vGreZ78jNdWFjAyy+//JrnOJvN4tu+7dvgOA4qlYqZnyxDx3heeY4kFO6lUgnT09M6e1lRFEVRFEVRNoiBFMye5+Fv//Zvcfjw4X4vxUe9Xvc5sFLU0amkeKTrStFMcUShODs7a0ZMFQoFfPWrX0UwGEQmk8Hly5cRi8Vw7NixdT8m2atL51SGaVEo8+YAe4UtyzJCm9u0Wi3Yto1arWZK0NvttnFfu92uGb0UDAYRiURg27bPtZYl2xz/RKeb78vzKnvK5euBlb5jvoY3MvjZUfjK88CwMgCmX5nincnnsj+70Wjg2rVreOWVV657fiORCPbs2QPHcVAqlTA1NYVKpWJcfZ5//g4B/jFSTB8PBAIoFAqajK0oiqIoiqIMHBXXxq4PvdrvZawLAymYAWB+fh7ZbLbfy1iFnBMMLAVMyXnBLL+mIJJimenS7XYbhUIB5XIZsVgMwWAQ//N//k8AS6W7cl8bcTwUj72jsbgGWV7ONGn2G0vhzH5bClSZFM5wMPZMUww3Gg2zH76H7EtmijUAI8BZNs3xVwB8a5ejoSThcBi2bZu5ykzHBmDSz+WsaCnmeXy8aVCv1zE7O4vjx4+/5vlNJBK477770G63MTc3h0KhYN5b9sLzPQgrExiaxpnNvWFoiqIoyvIfat93/ZuXiqIoygZwgwDczcrACmZgaVxP7yzcfkPRxDLjUqnk6+mV5ckUN81m0/TzUlC2220sLi4iFAohlUrh4sWLAJbKcbdv346ZmRlkMhm8733vW9fj4drlGCemXfN4+S9dW8uyTKgWRTTHTMneYgCrhLAUihTRFKhSqMsycD7Hvmeuk/u0bdsXxsWyapZl0x2u1Wq+45FhYZ7nIRaL+Uq8ZTo2KwMoqoPBIHK5HF566aXrnttEIoF3vvOd5sbA/Pw8arWaccZ5c0UeM9fLc+w4jglk6x3jpSiKogC1Zhg7/rfb09VQFEXZDNzu1+GBFcye52FmZgbvec97+r2UVbBsl04pYT+u7EcFlpKWKZJlCTMTsVOpFPL5PP7yL/8SgUAA2WwWjUYDo6Oj+OhHP7qux+K6LkqlkpmvLEuW6cCyVDkSiRhXt91u+9xPikz25srkcN5c4M+ytFs62b3J0NJZpoCUQWK9/eH8TAjLp+VzMmiM+5YVAhSnFLVy1BRd5mq1ilqthkuXLl33vFqWhWQyid27d6Ner2NqasqIZQDmPMuebgpiuWY60BxTNmg3kBRFURRFURTldmZgBXO328Wzzz6LRx55ZN3fqzel+LUIBAKIRqMm1EvODgZWHGimMvNnikoAJlSqVquhWCwikUggGAziS1/6EoCVAKhIJIIDBw6swxGv0Gg00Gg0jAikwyyFLp1OGfhFJ5fOMs8fe5Mp/mSZNt1hz/NQq9VMTzDdYymopYilMJYuNcuy+R4AjIiPRCI+QS+3oSiXo7Lo3rJk3LZt2LZtbnAAK045e9avXbtm0s3XIh6P41u+5Vvgui5mZmZQLBYBYNVILtJqtUwImkzr5rE6juNL81YURVGAejOMbR882e9lKIqibFkardBtfx0eaMH8D//wD3jrW9+67u+1Vr/r9ahWq8apDIVCvtJlurEAfL2xEulIUzADQDqdxqVLl3Dt2jXT93vu3Dlks1l8x3d8x5s+xuvBIC7OFLYsC81mc1V/Mo+LxyPTndlzTJFLB1mWrvPYWeZM51aObZJCXM4e5g0J9k27rmvKomUJPEveOTsZgBHBkUjEiHJZGk8Xna44S8SZtA3Adyye56FcLuOFF14wc7l7sW0bR48eRTqdxsLCAvL5PNrtts9NlmXYfA+GiXU6nVVJ2aFQCMlkEkNDQz4XXVEUZavS6gQwcZv/kaYoSh/oqjHxeuh4t//5GljBDACTk5MYHR3t9zJ8TE1NAYBxJfk9AONWSkFGWK4sy4hDoRByuRxqtRqSySTa7TY++clPIhgMYmRkBIVCAdFoFB/84AfX9ZgoQik2k8mkb7wUnWTZsyyPnYKZZcPSne52u3AcxxdutZZbzJsWsixZIsdGsaSd66OYlUnksidb9grzc+ntX+b8Znnjg8fOx4ElAX7lyhX8wz/8w3XPp23bOHLkCFzXxeLiIur1+qr+Y944CIfDcBzHd5yyj5qvY9WC4ziIxWI3XRGhKMomo3P7/+Fxq2h3tEVFUZRbz+GPPYu5uXS/l7EpaHUCt727DAy4YKabN0glqBcuXPCNKOp1pynKpIijQ5vJZHzist1uY35+HuVyGYlEAtFoFC+++KJ5j2g0itnZWezbtw+xWGzdjqlSqaBSqZgZytK1peMpRztR7PJ7Ka7b7Tbq9bpxgSnsuC85FqrXlZaBXkzBlg62PHcU7nTj+ZnI9GxuL9csU8B5DNyWY67osMvEbArncrmM5557zpyrXmzbxv333492u41KpYJisbhqvjJHQ/H4+TtCcSxvLnDNLH2Px+OmhF9RlNuPw/9E/1C7GWrNsLrLiqIofabZDt14o9uAgRfM586dw0MPPdTvpayi152UZcWyD5fupiz9leXG7XYbCwsLZgZzsVjEn/7pnyIcDmPv3r1wXRfxeBzf/d3fvW7HwpFFLF+Wa6Pg5A2AeDxuRmZRREoXlO4sz4O8gUAxSwEZDoeNcKUrDazMTeZrpLDmOWdiNfuMe4W57C+nCKUQbbVaRlQHg0EjQPl878gp7qPT6WBqagpPPfXUmucxGAxi3759GBkZQalUQrVaNYJXzpqWs64pyHmTQpaH81h4TtlbzrnMiqIoW5GuZ20JR0NRFGWQ6XQD2P69J/q9jA1hoP/q7nQ6ePrpp/Hoo4/2eyk+Op2OKYul6JG9zBRfLM2WglAmZVMMlUoldDodOI6DRCKB48eP4+WXX/bt493vfve6iaT5+Xnk83nUajXU63Uj0KQby++ZGC1df7rLjUbDuL0UnL2BX3IElZyhLN1n6WbLmdec3czzTBHK88gvOXaK46X4PhT+FN+RSMSkV8uxVrIUm9/Pzc3h61//+nXPo+M4uPvuuwEspY9zjBU/Q1myz+Oi8AdWAtMAmJA4roGCmcFgGv6lKIqiKIqi9IuqG7nxRrcJAy+YP//5z+PQoUP9XoqPUqkEYEU0MjiLj9VqNQAwQolOa7vdNqOBZE9ro9FAqVSC4zjIZrPI5XL4zGc+A9d1kc1mUa/X4TgOjh49ui7H02w2MTs7i0KhYHqZgZUgKjrGFKlM1qYwljcFKGJZDi1FKnuMZc+37FuW51MGffX2I1NMU4TLEmwpouV27JumcJXjs+iscz0cD0aB63keqtUqzp07h2eeeWbNcxiNRvGe97zHnA/XdU2AmOxr7+1N5hqkq89SdgpkhrJxZBcAdZkVRdmylGvOjTdSFEV5E3huEN0tEGb1RilWo9j1fa/0exkbxsD/xe26LhKJxEC5aXNzcz6BRucYWHExCftwAfhGKFHoAUuCdWZmBrVaDel0GvF4HMePH8df/dVfIRaLIRKJIBgM4qGHHlq33tW5uTlcvnwZ09PTqFQqRjRT8FMcS9EGLI1PkuOmpLssS9UpqmXyNkPQuI0ctUS3Vzrdch90qHtvPvD96dBStFKYr1UOzfXws+PxsCzcdV1cunQJf/d3f7fmuYtEInjkkUewb98+jI+PA1hKQOfoLMB/c6XRaPhuBMiecZ4HKZp5DliCLoPOFEW5DamG9A+167BYjmH3h17u9zIURbnNOfxjzyB3NavX4jXoetaWuw4PvGD2PA+FQgF79uzp91IMV69ehW3biEajRuCxvLhXsK0lhvgv3cxOp4PJyUksLi6i3W4jnU6j2WziC1/4Ak6ePGkCwfbv34/77rtvXY6pVCrh8uXLmJ+fR71eXyVcZfq053nGfabI5PfsgaY4lT3KMum60+msGtckBbMM+JLnlMKcr+FjPPcsfed7y0RvbkORTgEue5Qp4rnvdruNxcVFvPjiizh//vyq8xYMBnH48GE89NBDSKfTcBwHruuiWCz6bqLwPPBcAjDCncfMUWXsf280GkbYy3FlPHcqmBXl9uTQTzyN3OUsOt2B/1/0hrP3fz/e7yUoirJFOPxjzyBfSPR7GQNHobR+QcSDysD/37jT6eDkyZPrPlrp9ZJMJs0IJZmmzJLlTqdjnGFgyYWUIo5we9d1kcvl4HmeSUK+evUqPvvZz+LatWtIJBIYHR3Fgw8+6JvlfCtZWFjA3NwcqtWq6QsG4JtJTEfUcRzjkErXnP/yWLl9782DcDiMSCTi2y9FpRTWMsyL52ut5GsKSj7OtRCKaW7fu60s9aZQ7na7qFQqOHXqFB5//PFV58uyLOzbtw8f+chHkEwm0el0cOnSJfM59pafS/eYIlgmg3M9skxdCmxCl1lLshXl9uXwP30Gi8V4v5cxUMzrH66Koih9Z98PvNTvJWw4A/8Xd6fTwVe+8hU8+OCD/V7KKljeKwUNQ6Yo4ih82u22CbaicORrGXq1sLAA13URDocRj8cRjUbx/PPP46tf/SoKhQJCoRB2796Nd73rXesilnK5HC5cuIDLly9jdnYW9XrdlGAzrKtXoLbbbd/4JwpAOSqJ23PNFOCu65r+XD4uz6c8VwzPoiPN8na+p+xPpmCn6w/AuLZ09QGYoLXesm6KWdd1ceHCBXzlK19Z83xNTEzg53/+5zEyMgLbtjEzM4MrV64gn8/7BLIs0+e6KND5Jd1xwjXJ4C8K6t5kcUVRbj+6+Yi6zMvkFpI48OEX+70MRVG2GHod9jM3n+r3EvrCwP8GdLtdnD17FqOjo/1eio+ZmRkjmGXJLEWhdJYp4KRgJBShwWAQ1WoVs7Oz6Ha7SCaTGB4eRigUwksvvYTnnnsOzWYTqVQKBw8eNL2yt5qLFy/im9/8Js6ePYvp6WnTbyvLiyniWPosQ7vkscnzQ5G3VrmxTBOXwpvnle8txSLfS46Uku6xPOfcf+9rKealE97pdFCv181n8fd///e4cOHCqvO0Z88e/If/8B9Mj3kulzNimW44nWC5/9418hg6nY5vfJRMCacLz7A1vk6TshXl9ubQTz6NhYtamg0ABz/yzX4vQVGULcihn3waxXK038sYCObm0jj0Qy/0exl9YVP8X9jzPFQqFSQSg1OOdebMGZOmLBOYKZYohCiaZNI0ACOIesXh1NQUCoUCIpEIMpkMRkdHUa/X8corr+CVV15BpVJBNpvFW9/6Vti2fcuPq1qt4tSpUzh79iwuXLiAxcVFM2pKrlm6voBf/En3WTrD0oGmEOfjhMKZ++S/TMIGVlK05bgm6Q5zjXzMdV3jcHMfvT3TfC1vBuTzebzwwgt47rnnVp2jAwcO4JOf/CRGR0cRjUaRz+dx4sQJzM/Pm7FWvancXIPsDedz0iXnsUv3nHCbdruNUCiEaDSqZdmKcptz6CefxsKFLGanM1s2fGZ2OtPvJSiKsoXpzEa3/I3L2ekMDv3I8/1eRt/YFJ9+u93Gyy+/jPe///39Xoohn88bIShDsShqCEUby5Z73U35c7vdRrVaxfz8PJrNJsLhMDKZDEZGRrC4uIjjx4/j3LlzaLfbOHDgAO677z7jPt5KqtUqzpw5g7Nnz+LMmTPI5XKmdJrp1xSe1WoVrusacWhZFmKxmC84i0JQBlaxNJnbUOjKMnfZzyt7nAGYxGjCcy+FM11w2U/c20fdK7QbjQaKxSJeeeUVfPnLX151bu644w787u/+LkqlEqLRKCqVCp555hlMTk6iWq2aY+GsaemK8z16g+D4GN363p5mGSwnE9lZvaAoyu3NoX/+NA5/9DnMXxjacqJ59moWhz+6+saloijKRnHw408hf3YIs1s0NVuvw5tEMLdaLTzxxBM4duxYv5fio9VqmbRo2S/LwCvOApbPy/LtXhHFcuCZmRnk83kjPkdHR5FMJrG4uIizZ89ifn4ejuPg8OHDOHDgwLqIZrqmJ0+exKVLl5DP51EqlVCr1XzClY6qFLVS7FEchkIh35gnWQItz4WctdwrHGWZsuwPluXZfByArxSb+6CglV/tdhulUgkLCwu4fPkyXnzxRTzxxBM+QQ4Ad999Nz71qU8hl8shHA6j3W7jq1/9Kk6fPo1CoWCOUzrHLLWWjrEMA6PTzD5uimCGogHwJWnzJg3defl7pyjK7c2hn3waC2eHMXtpawjn2UtDOPxjW/uPNEVRBoODP/0UDv+TZ7FwZnhLXH8BYPbykF6Hl9kUf213Oh289NJL+Omf/ul+L8XH1NQUxsbGjLPoeR6azaavvLZer5vSadu2jZByXdeUMzM4iy5puVzG5OQkYrEYUqkUotEoxsbG0Gg0UCqVcOXKFQAwpdndbhcXL170pW/fCvL5PM6ePWtE3tDQEJLJJIaGhkywGUuuKUZlKjSwIlrl3GkZ8kWnHVgpU+d+2aMrhbF06aWA5uxnKaxDoRAajYbvmPgZsfS63W6jXC7j2rVruHz5Ms6ePYtTp07BdV3zmnA4jLe85S34L//lv+D06dOYmJjA0NAQPve5z+G5555DoVAwbjYrDmS5t6wm4PGx/3stB/l6rrHs7yaRSASu6/pmYyuKcvty8ONPAQDOfeIdGD0y3+fVrB9z54Zx+KeeAUTFlqIoSr85+NNP4fwvvwNDdy4gYN2e16e588OABxz++HNAt3PjF2wBNoVgBmBE5rZt2zA9Pd3v5QAArly5YuZDUxRS0IXDYSMMZc8tBWAkElkVFkbx2Wq1MD8/j1gsZlK3U6kUhoeHMTc3h2KxiOnpaQQCAYyMjODee+9Fp9PBlStXbrloXlhYwKuvvmrO/fbt203/cSQSQSQSMbOLeR7oqErnm8fPGwsyIEyKalnS3iuaZTCYdKilmOb57oWvbbVaqNVqAIBisYiZmRlcvHgRZ8+exdmzZ1edv9HRUTzwwAP4xV/8RVy6dAnj4+PYtWsXPv/5z+PrX/86FhYWfGXnwGoHnG40CYVCxpmXCdo8Xuk4y9Rxniv+PnU6HU3KVpQtCv9oG7lrc4vmuXPDsNqrr2GHfvZZFcuKogwkB372KVz4T8fQDXkYvWNzX4Mlc2dHYHWAw//qm/B6DKetzqYRzK1WCy+99BI+9KEP4dd+7df6vRwAS3200WgUsVhsldCiCKS7KF1B2btMd1b22wJArVbD1NQU0uk0duzYgVAohGw2i3a7jbm5OeTzeYRCIYyOjmJ0dBRHjx4FAFy9ehXNZvOWHmexWMSLL76I2dlZTE9PY+fOndixYwdGRkaQTqeRSqWQTqfNCCoZuCWPuTfQqze9minaMvVa0lt+LcPHeC7p1krhblkWXNdFs9lEoVAw5+/q1au4cOECzp8/v+q9AGDHjh34yEc+go985CO4dOkShoeHsX//fjz++OP4y7/8S8zNzZnPjWtgtQBFsexRp0hmGXWr1TKfv3SfZTJ4KBQy/e9EnluKbUVRth4HfvYpLP71oU3rcuROj+COf3cSnUKx30tRFEV5Xez/l98AAFz4j8eAADB8T67PK3rj5E6NINC2cOTfvoRurYbN+X+U9WXTCOZms4knn3wSP/VTPzUwghlYWpfjOKhWqz4HFFgRM+zPpVCSfblyG/a/cvtKpYLp6WmkUilEIhEEg0FkMhm4rovFxUXkcjl0Oh2Mj49jbGzMhIBdvnwZ9Xr9lh5nt9vF5OQkZmdnce3aNWzbtg3btm3Dzp07sWvXLpPcDPjdZGCppLnRaKDVaiEajfpuJMh+aCmC6ai+lnu81pguwl7lZrOJSqWCQqGAXC6Hy5cvY3p6GteuXcPCwsKaQhlYGhv1oz/6o3jf+96H8+fPI5vNYs+ePfjSl76EP/mTP8HMzIwRugwzkwncvSXjLGHnuvlZS/EvE9Z7hbCcHS1/b1i1oCjK1qT44gjYTjd0dPD/YMudGkHQXVrwkV85rWJZUZRNzf6fWxLOF3/p2Ka4BpP5EyMINJevxf/pBDqFIrS57/psGsHc6XRw7tw5JJNJEyA1CJw/fx7btm0zAlA6pO1221eeK2Fprex5XWsOcalUwvz8PLZv345AIIBEIoGJiQk0Gg1UKhUzt3nbtm0YHh7GW9/6VjiOg3PnzqFarV5XEL5R6HDPz89jamoKc3NzRpBms1lkMhlf+bnjOGi32z73ORAImD5i7pPnhOXd8qaC7AmWPdL8ohiVInlxcRGVSgW5XA6XLl3CzMwMpqenMTMzc8N+323btuHDH/4wHnroIczOziKbzWJ0dBSPP/44fv/3fx+5XM64w72jxNjL7TiOKa8GYPqVWU0AwBeSxt8RGQwmS9FlAJockRUOh1eVfCuKsnXY+2++Yb6//O+XgjEzDwzmH23zJ0ZwxyeuoD05BQDQzjhFUW4X9v3rbwz+NfjkCELVpb8lD//GeXRm5wDotfhm2DSCGVgSSZOTk7jjjjvwyiuv9Hs5AJYE8+7duxEKhUwpNN1Vijv+3JtsLNOb2asqZxcDMKI4mUwiGo0al7lSqaBSqaBWq2F+fh7tdhuZTAbRaBR33nknEokELl26hLm5uVteos2153I5FItFlEolTExMYPv27dixYwcymQwSiQQikYgpUaaQrNVqRuAx+AqAL1FclljLkVNyBjHLn6Ujz5FQCwsLuHDhAq5evYqZmRlMTk7edG/33r178f73vx8HDx5EuVzGnj170Ol08Id/+Id44oknMDU1hVarZZLQCddLwc+KAumM994gIHSk5c0T/ix/d/g7wjJu7ttxHJM+rijK1mXPzy+J56v/z4fhWUD6HXN9XtGSSI4Ul/5feOS/XzZiWVEU5XZDXoMBDNx1+MjvX0X78lUAKpJfL5tKMLuui6997Wv44Ac/ODCCGVgSe7FYzLio7J+lC8jHKJg4a5ilvAxxAlaSkCm2XNfF/Pw8EokEdu3aBdu2Yds2RkZGkMvl0Gw2Ua/XjeOcTqeRyWRw6NAhbNu2DefPn8fFixdRLBbXRVA1m02cO3cOMzMzuHz5snG6+ZVKpZBMJgHAiN1oNArbttFut2HbNizLQq1W87nzMuiq14Wl2G40GnBdF67rolqtolQq4cKFC5iamsLVq1eRz+eNw3szHDx4EO9///vx4IMPotPpIBaLYWZmBk888QSefPJJVCoV8xlTvMrPGFgpw++tHpCfMXube2d2y9dKR1mGo8m+d3mz4FZXEiiKsnnZ9QtfBywL137utUcxdhwg8+Ct+2Nu7uwIojP+lP/DfzaNzrmLAACtg1EUZSuw6xe+vvSNuA57ASD1zo0Tz/MnR2DPL12PeR3Wa/AbZ1MJ5kajgW984xv4wAc+0O+l+JicnMSOHTtQKBSM0KOAkb2pdAkppui89jqVFGIck9RsNjE/P494PI5kMgnLspBKpTAxMYFWq4V6vY52u41KpYJGo4FGo4FMJoOhoSE8/PDDOHDgAC5fvozJyUkTVHWroeM9NzeHWCyGZDKJTCaDbDaLbDZr3ORUKoVMJoNIJALP8xCJREw/r5wpzFJjimSek1qthmq1im63i0KhgPn5eRQKBSwuLqJQKGB6ehrVavV1rX337t0YGxvDt37rt+Kee+6B4zhwXRfHjx/Hc889h4sXL6JUKvmqBmS1AIUxy8Pb7TZarZaZqdxbqs/tZY8zRTVFsRw3xZstvAkjHXhg7VRwRVG2OJ6Hnf/h66+5SXB4CFc/eseqxxtZD0Nvfe2SwtkrQ0hc8P8JcfiJIrznX/U9pi6GoihbFnEdtsIRTP70A76nOzaQOuYX0YvlGCLPJ27qOiyZOzeM+LWV/JtDX8ij+8qppfd5o+tXDJtKMHueh3K5jFAohImJCczMzPR7SQCAa9eu4dChQ4hEImi1WqY/ly5hb0+q67pGHNJJlknLFEWyn5lO89DQkOkTZmk2y3FZogwsufG1Wg179+7FAw88gG/5lm/B5OQkzpw5gwsXLuDs2bO3PBgMgBHsi4uLuHLlCuLxOKLRKIClOdT82XEchEIh8xWLxQDAOLjyOQBmDFc+n8fi4iLa7Taq1SqKxSJqtdobmkU8Pj6O0dFRPPzww9i/f795j5mZGbz88st4/vnnkc/nfeXPcoyVDDqTpeK8CSIf49pYUs198TEZfMYebwpoGQom09bZBsAbK4qiKK+HzkIe2//TalEd2rcH1z644zVfe+CbLoL/6xnfY3oVUhRFWRuv1Vx1vQ2mUpis3ON7LDPvIfMHX7+p67Dk0DcqwFPHzc9qpdxabiiYLcvaBeD3AYxj6f+Hv+153n+2LOvfAvjHAHj74197nveF5df8KwAfxdJNjX/ued7f3KoFNxoNnDx5Et/xHd+BT33qU7dqt28Kuo2O46w5P7e3zJiuMQCfoAZWxlFRIFFUtVotMxLJcRw4joNoNIqhoSGUSiWUSiUjmprNpinVpni///778Z3f+Z143/vehwsXLuD48eM4ceIEvvGNb6BSqazbualWq2s6vtKdtSzLzFuWJc+8WQCspETXajU03sRsuHg8jh07dsBxHOzbt8+4y51OB7lcDqdPn8bc3BzOnj2LSqVixLEsp+bnKMvp+fnJEm3Zqyw/T5bc9waZSVHNfnb+LB1pCm4+/3rKzpXNyaBdh5Xbm/bFy5j4xOV+L0NRBg69Fiu3kk6phIlPrF0JpNfhweJmHOY2gJ/xPO8Fy7KSAJ63LOvLy899wvO8X5YbW5Z1F4DvB3A3gO0A/tayrMOe592Sv+prtRq+/OUv4wd/8AcHRjADwMzMDBzHQaVS8TmCvenHdA0BmLRvWVpLcUbhHIlEACwJrnq9jtnZWZ/oY5l2rVYzwVkUa+12G4uLizhx4gRarRZisRgefPBBHDx4EO9617tw5coVPPvsszh37pyZK7xRTiUTrsnrLaN+PViWhe3btyObzSKVSmHXrl0Ih8MIBoNoNBqYnJxEo9HwpWjX63WTuE1Bz331pp5L0dubcs3qAVYSsAogHA4jEon4qhH4ehlyJgV576iqXlGu3NYM1HVYURRli6LXYkXZgtxQMHueNw1gevn7smVZJwG8Vo3ABwB8xvO8BoCLlmWdA/AQgG+8xmtumlarhVOnTiEajRrBMQicO3cODz74ICKRCBqNhimxZk8rnVSKHgofimbppvaW6AIrQVDsEx4aGkI8HkckEkEmk0GxWDTuthRdzWYT+Xwex48fh+u68DwPjz32GEZGRjA8PIwHHngAlUoFb3/72zE3N4e//Mu/xKlTp1Aulze9c5lKpTA2NoahoSFs374d6XTaBIWVSiVzo4IztPP5vLnhwZscDBhjaXSvcJWwTFt+jnSe5Wgoltwz/A2A7/dD7k/eaJHjq9YKDVNuXwbtOqwoirIV0WuxomxNXlcPs2VZewHcB+BpAO8E8M8sy/ohAM9h6Y7bIpYuHE+Jl13DGhcTy7I+BuBjb2TRjUYDFy5cwLFjx/CVr3zljezillOpVIwjTCEGrIwMkmILWAlyomhi3y7HKMkxQtwPADNjeGFhAY7jIBAImCRqlhDLdGmK6G63izNnzhix+O53vxv79+8HAExMTOBHfuRHEAqF8I53vAPnz59HpVLB5z73OTz11FObTjgPDw9jZGQEo6OjGB4eRjKZNDcOKJibzaY537VazefW8jmWjctyaI7JkoFtvAEixTBheBvh74Ast5ZwP7Ztr5rzLJ/n74ucWa1sDW7ldXh5f+Za7CC2fgtXFEW5jVivv4n1Oqwog8dNC2bLshIAPgfg457nlSzL+k0Av4ClHo5fAPArAP5vN7s/z/N+G8BvL+/7ddlklUoFTzzxBN7//vcPjGAGltKyR0ZGjNCRTqIUVxRWFE3sfZahXQDM+CDpaDLsKpfLIZ1OIxwOw3EcZDIZLC4umn5kuo90SC3LQrVaxcWLF/Hnf/7nOHv2LN71rnfhscce8zmVjz76KB599FEAwNve9jY89dRT+LVf+zVMTk5u5Kl8Xdi2jdHRUWQyGXMu4vE4wuEwms0mZmdnUa/XTciZ7DPnaC85C5mhbEwyd13XNzeaNzvk2CemWrMXu9Vqrep/5k2TYDBogrpkBYHsUWZCuBTXrFYg8oaKsjW41ddhwH8tTllDWrKgKIpyA9bzb2K9DivK4HFTgtmyrDCWLgx/6HnenwOA53mz4vn/BuCvln+cBLBLvHzn8mO3jGaziQsXLmD37t23crdvmmvXrmH37t2r5ucCSyKoXq/7ymzlKCWKo94EZWBJTIXDYQAwIpgu88TEBAAgmUwim8363FO6k8Fg0IRpAUAul8NXv/pVXLx4EWfOnMHRo0fxyCOPmKRq8s53vhMPPfQQDhw4gFdffRVPPPEE/v7v/75vZcDs3Z6YmEA6nUY0GkUwGEQwGDTfdzoduK6LSqWCer2OVquFcrmMQqFg5j4zNIw3MhqNhrk5QYeeZfLAklCVApefjUyxphiWlQIsxacbTbjO16K3N3mt2cz8/VGHeWswaNdhRVGUrYheixVl63EzKdkWgE8COOl53q+Kx7ct93IAwPcCeGX5+88D+CPLsn4VSwEHhwD4Z0/cAprNJnK5HB566CE888wt3/0bguOdEomE6YWVgVAM86Iok64xBRudyt4SbmCljNvzPJ/LzDFNExMTKJfLWFxcXJW43Ivrurh48SLm5ubw7LPP4sUXX8Sdd96JRx55BJlMxry34zh47LHHcOTIERw6dAhHjx7F2bNncfbsWVy+fHldZjoDS677/v37cffdd+PgwYPYuXMnhoaGkM1mkU6n4TgOisUiZmdnUSqVUCwWsbi4iGKxiHq9Dtd14bouWq2WEdAM75Izj3m+ZfgaXd3eXnMGlXHcFY+dNzo8z0Or1TI/81/eFOF70dGmaObvhBTda/VJ9zrWcvSYcnszqNdhRVGUrYReixVla3IzDvM7AXwEwMuWZb24/Ni/BvADlmUdxVL5ySUA/wQAPM971bKsPwVwAktpgj+xHmmAxWIRf/M3f4Nv+7ZvGxjBDACnTp3CoUOHTBkvACPQ6BKzHFv2GAcCAeNE8nUUR3IckRxXVSgUsLCwgPHxcQSDQSQSCSQSCZTLZeOW8j1l8jK/2u02KpUKzp8/j9nZWTz11FN45plnsH37dhw5cgRvectbMDo6ikgkAs/zkEwmce+992JkZAQHDhzA1atXzRzoWq2GWq2GhYUF5HI3P2gdWCqpPnLkCPbt24c9e/Zg79692LVrF0ZGRjA+Po6RkRGk02mTVA0sCf4rV66g2Wyi3W6becwcY1Uul1Gv19Fut9FoNMzNBn5PEcvQOIpR6RgTnsdOp4NGo2EStvm5hEIhUy4tH6MLzFJuKbpl7zM/F27b60BTnMubJxTXvEmj3PYM5HVYURRli6HXYkXZgtxMSvbXAKxV8/mF13jNLwL4xTexrhviui5effVVPPLII+v5Nq+bubk5vOUtbzFl0NL98zzPVx5NASTFrUxCBlZ6W6WYo7hyXRe5XA6xWAzJZBIAEI1GYds2arUa2u22KTvmvoCVcl+WC8diMcRiMZNAfvbsWZw5cwavvPIKJiYmMD4+bsRwqVQyIj+dTiOdTptyZWBp7FexWITnecblLZVK5jgty0IqlUImk0E2m0Umk8HIyAj27NmD3bt348iRI9i5cycSicQNzzVvBlQqFeRyOczMzGBxcRG1Wg2VSsWEeXE2MsUzxTJFKW9MNJtN4zJTNLO8nefdcRzfuCgprvm9FMKytJs/SzdZ9phzHxKGv8nn+bujpdhbh0G9DiuKomwl9FqsKFuT15WSPUh4nodyuYz5+fmBKsv2PA/NZhPRaBSNRsOIMFlGC6yEcvW6kJ7n+cp9uc9e0UuXulQqIZ/PG4eaok2ma3c6HV9ydzAYhOM4iEQiSKfTGBoaQjKZhOM4JuQqEAhgamoKc3NziMfjZn+1Wg35fB5TU1OYnV1q2aHDGgqFEI1GTZk45xhznBXfd2RkBGNjYxgdHTVu8tjYGJLJJGKxmBHC16PVamF+fh6XLl3CiRMncP78eTNDuVKpmL5kOXKMvc48jzwn8mYEb1zwc+BnEwwGTTq2DPiiaJXC27Zt1Ot1XwAYS775+fFcsvxaCmA+x5sd0m2mey0/S3mzQlEURVEURVGUW8umFcwAUCgU8Pd///d417veNTCCGQBeffVVHD161PQxS/EqS3B7HUKGUrH8ei23mYKO4sp1XSwsLMC2bSSTSSPoKBZ73y8cDiMejyOVSiGVSiGbzSKZTMK2bUSjUViWZVxqAMYlBmAEeSKRQDabRTgcxuLiIjqdjnGRWa4cj8fNGikco9GocbO73S7K5TLm5uaMEKXol0FW3W4XtVoNhULB9GfPzc1hdnYWly9fxrVr1zA/P49isYhisQjXdX03DlqtlnGYeW6Ygs2bFLVazSeGZVl1KBQyrjTdXvY7yz50nt9IJOKbDy5vfvQmYssbKbJnmTdTZBl2b/+yFPQqmBVFURRFURRlfdjUgrnZbGJ6ehrj4+P9XoqPfD6PYDAI27bRbDZ9ycwUvbK0mg4mhZ0MCgNWEpEptKRL3e120Wg0UK1WzSgl6dDSCeW+bds2gjeTySCdThsns9lsmnVRuAJLopnlwxSRY2Nj2LlzJ+r1OsrlMqamptBoNEwJskz9tm0b4XAYkUgEsVjMHE+5XDaC//Tp08hkMkZMdzodU6Jer9dRKpVQq9VQKpUwNzdn+pTL5bLpn261WkZkRiIRI5p5fDx/sgy6d+SXTMFuNpsmJIyP97rAMsiNadsU1xTPdLyluyxL7QnPN88hg8nk7wHnQnNbPq4oiqIoiqIoyq1nUwtmYKlEeG5uDgcPHsS5c+f6vRwAS+4fnVMKLekCyp5aYGXeMvtbWT4MwAhA6UjyZwq/TqdjyoCTySSSySTK5bIZbSTfF4DpW3YcxzfGivOBKQC5LQU8XdNOp2NKgSORCJLJJHbt2oVms2mEq+zrjUajRsBGIhFfD7FlWabfuFKpIBwOm3FQwGqHuV6vo1AowHVdc17b7baZe8z3lKXVMp2ax0QobOXxU5Qy5MtxHCNiW62WOQZZDi0/F/Y6U/zLoDU5Vkyul6nawEpPdW9vtESWeqvDrCiKoiiKoijrw6YXzPl8Hk8++SQeeOCBgRHMwFJa9rFjx4yIpBi2bRvBYNA4zqR3jFBvIFhvarMMl2q320ZkRqNRZDIZ5PN5VKtVI77ohlarVdi2jZGREbMvvh+FcDgcNoI7kUiY0mYAZuYx31eKfDrc1WrVlC3TkQ6FQohEIrBtG6FQyBd8RtedLi0Fcr1eR71eR6VSQbFYRKVSMeKYbjfPEb8oRHsDznrHNnFtFPfSYea2LHmn8KWg5riter3u+wzlzQdWClB883neeJBOtW3bvvPAY5Ll2MDKzQv+y+NTwawoiqIoiqIo68OmF8yNRgOzs7M4evRov5fig720FHV0lKXIY8k2RY90HZnMHIlETKmznM3cW95br9dRLBYxNDRkgrN6ZzF3u120223k83lEIhHU63VfD7MU4t1u14R1hcNhNJtNk1zd6XRg27aZB831ci3yGPkYBSIFHx1VCr9Go2GOoVQqYXFx0QhkCnnpXMtZyrK8mULecRy4rmvGQMmZx3TmpYsty8Ap9hng1Ww2fa40w8xk77UceUUajYapDKAbzJsH/BxZtl2pVHxBbzKIjO9JpFjW0C9FURRFURRFWT82vWAGgGq1irm5OezduxeXLl3q93IMFy5cwMTEBGq1mikhJrIHlcgAKYqrXndZimqKRAo4ijv2CpdKJVNOLR1r13UxNTWFcrmMhYUFZDIZjI2NmdJpx3HgeR7i8bgRe71jjxhsFQqF4LouarUams2mKW1mzzLX6ziOL4iLa2YfMvuUFxcXUa1W0Wg04LquEa69Pb8U9sCKsytHefFfOdea29DRlSXQct/8bKQQ7p113Ft2Lh/vddGl8KUQ5uO9aeay/F7eHKGgl58/+9oVRVEURVEURVkfbgvBXCgUcPz4cRw9enSgBPOlS5ewa9cuI+4o7Ohw9gpfWcor+3BlCff1UpKDwSCq1Sqq1SpisRgymQwqlQpKpRK63a4RuNLxrFarqNfrJnl6eHjYjHsClkRxq9Uy/b8MF+NIr0qlYpLAK5WKEewA4DiOL22bY5b4vq1WC8ViEYuLi8aJpcvcO7eYbrt02IGVMUvASl8yR2I5jmPeU7r3MmSN55PClSKV54lOOoWvnJfNz0D2e/dWCdDRl2Fusj+ZJe10nCms2cMu98NzAcCcU0VRFEVRFEVR1pfbQjDTgctms/1eio92u425uTnEYjFUKhUjxuSIIjqRMtmZUOhSOFMwNhoNI7xYkst+23w+j3Q6jeHhYbiui2az6ZtHTGdUupeu6xqHN5fLIZvNIp1OIx6Pm7JmhnOxrJphXHSBZbKzLJdmv7Nt2+Y81Go1U0bOfVGwRqNRc+6ki8ubDFL4Aiu93r0CmO58q9Uy4pmOMdfG0msAPkdXfn69JfC9Ypmfk0wS57752dK1lyFvsv+bIp6iGlgJYeu9ORAOh00aN5O3FUVRFEVRFEVZH24LwQwA5XIZk5OTuOOOO3Dq1Kl+L8dw+vRpPPLIIyiVSsap5Fxf9idTGNN5lI4m4O8PloFUFFxSVM7Pz2NoaAi7d+/G+Pg4Wq0WSqWSb+QTaTQaRjzX63VYloW5uTnMz8+bAC8Ku0gkgna7bUZPUbhzfUyQZr8wS5kpiuv1unG0KcD5ejq8yWQS8XgcnU4H+XzenBt+yWC03lFQFLCy15hOLc8PRbQsf+YNB+naVioVBINBE/Ylw7bC4bBx3eVnwu95Y0OOl5KfY6vVMoJXpqP3jvJi7zZfx/PPdSiKoiiKoiiKsv4EbrzJ5qBQKODUqVPYt29fv5fig6I2lUrBcRwAK2nHDPyi6Go2m0aM0b2UCdfAykgnCj2KOJYel8tlLC4uotVqIZVKYefOnRgbG0Mmk8HQ0BCGh4d966AIlYKMfc6Li4u+0mvXdeG6LgAY15UhVgB8AlGWKfPYGo2G6XOWopE911Isyr5jrpVhYb2l2dxHb393KBQyvdPsK+b+ef4ptLntWqnZAHwimQQCAROwRtHNdcp+aPm+dKHl5xeJRHzuuRTe7MfmTQs5l1kTshVFURRFURRlfbltHGbP85DP5+G6Lt73vvfhi1/8Yr+XZDh16hTuuOMOFAoFn7hjqTR7WAEYd5lurex1pQvKbeg0SxHFfuJSqYRsNotsNmtSmOnoFotF5HI546RSQPI9KEYpOunIyrJn6WrzXzq8UvhRHLI/m2OZZA833WMK81ar5Stdl9vIf0xgJwAAK8NJREFUucby3Mkkbtu2EY1GTUkz076BlWA1iu5AIGAcfwBmzjI/g0QiAcdxTEk11yZvZPB4KH55PtiXzGNnDzeR7jjgD/GSFQZSgMtydBXLiqIoiqIoirK+3DaCGVgqpZ2ensaxY8f6vRQfMzMz2L9/vxGStVoNwErfLJ1C9rDKHliKTQZ8yRJj2c8rRaTruigWi0ilUojFYojFYkilUnBdF6FQCIlEwrjZLI+meJNp2r1hVQyiotspS4blDQApALk2vq437Ixin6JWjpoCsMp5lWJRpmDzXFLgcuwV+6y5Trq6fA37wOXNAM/zEIvF0G63jRPOGwYMFZPH7DiO+ZkVA/V63QhongMie5N7XWL2iMvPpDcMji59tVrVHmZFURRFURRFWUdum5Jscu3aNVy6dAnvf//7+70Ug+d5OHnyJJLJpOmjBWBKqwH4RCHdY4aEUazR6ZWziLl/7oOubKlU8jmg7MV1HAcjIyPYtWsXRkdHEY/HAQDRaBSxWMzsz7IsOI6DTCaDTCaDeDxuxLcUvzLATLqtFPutVsu4wFwfnW660aFQyJRMS0da9nJLR5hONkUnHeVMJoPh4WEkEgl0Oh3UajXUajUjiDnOir3Yvf3f7CVnUBldZRnQJUvB6TSztJuCmp8ZR20x6VyGfsl+ah6TnCXNMnnbtk0JPUeTceyXimVFURRFURRFWV9uK4cZWHKZp6amcOzYMYyOjiKXy/V7SQCAfD6P++67zwhkOpXSXaSYC4VCxgVlgJbsaWZYGIUwhRjd2W63i2KxiOnpaViWhVgsZoRbo9FAq9VCPB7Hrl27MD4+joWFBViWhXg8jsXFRZPqTGeajmm5XMbs7KwRmxRzrVYLruuaUmYeH+B3qulSJ5NJAPAlR7uua8qg6SJTJMvya4pJll6Hw2EkEgnE43EzBqtWq5kRVfLmRCAQQDQa9aVjU/g3m03E43Ff4jid4EajYT6HWCxmerFZ7s0Ubn6GAExpuzwHskqA21Iky5nQ7ImWY6f4HMvmZZq6oiiKoiiKoijrw20nmAHg3LlzyGaz+J7v+R789//+3/u9HMPk5CTGxsaQz+cBLM05pjDj7GCW+PaWYtOppXCjiOqdA0x313VdTE1NodFoIJvNGuFL0cVS40QiAQCo1WpIp9MYHR1FMBiEbdtwXdeUj9OlZSk3xW61WvU5pgBM6TdHHzWbTd84JBmCxUCrtcqx6bryeyk4pfvN8U3NZhOVSgXNZtOMeQLgG9ckw9NkejbTpyORCPL5PILBoBkJxbJopn1zHb09yixJ57YsC+d7yIA33hzoHW/FgDS66DL8LB6Pm98L7V9WFEVRFEVRlPXnthTMHEvkeR4OHjyIc+fO9XtJAJZGTI2NjZmfZSkzACOKZTCWLEXm9jIAi4JTBodRuLIsu1gswnEchMNhU17N/l6+t+u6KBQKSCQSaDQaiEQiKJfLKBaLPmFLJxhY6bdlaBb7g2XgF8uuWVZOYUgHVY6Eoohda1sAPjd5ZGQE2WwWnuehVCqhXq+jUqmgUCgY8S1nN9PxpdMOLI28Yop1Op0228j5zTxWiuNGo2HKx+WsZJ6f3hA2Ps/xWrz5wce5fx47KwhkyBrPBasH2u02yuXyqtRuRVEURVEURVFuLbelYAaAM2fOYGhoCN/6rd86MILZ8zycOXMGe/bsQaVSQbfb9TmMdF8bjYYJAQNgem3lbF+KK4oomZotxTX7eAlLrilwWe7NkuuFhQUTWMUxUnRP+T6cA8w10lkF4EvZloFWFItS5EnhLUutKXL5OPeXSCQwPj6O4eFhZDIZM2O6Wq0il8uZ3miZiM3XyjA1YMnxjUajZu2pVAqBQAClUsmIV5ZxS3HM88aQL/nZ9p5/vh9dfyni6cizRB9YEsbRaNS3HzmDmgSDQTOWSlEURVEURVGU9eO2FczAUjr13Xffjfe85z148skn+70cAEtr2rlzJ+LxuBF4cs6ynAMMrIhlusi9fc62bZtSXzk3mK/t/b5UKq0S1SyX5nvJMCyuj3OauS+6ygy+4vYMwKJLzsdlaXk4HDY9wBS1sp8XgJlN3G63TVDZnj17MDY2ZvqK8/k8pqenkc/nUa1WjSiVfdO2bZs1Sle32WwiGo2aXnD2KbPPuV6vm2PleXJd19xIiEajvlJuloBTSPNmhiyrlsnYa816XuvzZ6gYP2+eexXLiqIoiqIoirL+3NbJQZcuXcL58+exf//+fi/Fxze/+U2Ew2EjgHtn7MrgK7rM/JfbdbtdX7o0nVlZDszX8Xn2vtK57S2JpnCLxWLGTeWcYjqxTJvm9nJOM91S6cgCK06vFJLSQWW/s+zLBmASoXfu3Ik777wT+/btg23bqNVqmJqawtWrV5HP5837speYr6cTzjX0zmAGltLBAaBarRo3nWKda+Hao9GoEcWE/dJE9pTLVGw6wuzLlmX1co0yxE2KZiaOy7FSiqIoiqIoiqKsL7e1YAaW+oZrtRoee+yxfi/F0Ol0kMvlkEgkjGCliJTlxMDq0VG9jqTjOGYOsBSekUjEBFaxjJrimoKQvcx0e+lUc3wR10UByn0CMCOaLMsy44+i0ahPSDPFmkKe4lOmQgMrs4h7U7KHhoZw+PBh3HXXXdi2bRuazSYuXbqEkydP4sKFC1hYWPCNh+L6mJbN5/ivPJ8U9ewVpkilGGVSuOM4RgTzWHgOmGgNrIjiRqNh3Hvp/PM45TxrnhP5mVJUy/nTsvy794aIoiiKoiiKoijrx21dkg0A09PTmJmZwd69e32ltv1mamoK4+PjPqErS5ulu0jnWSZRM0hKhkxJoUUo2CjWAPhEIvfPMC8KTI6ioojncxwdJYWdLBenS8tRVBSvMlCL5d1STFJEs786k8lg37592LZtG6LRKEqlEi5cuICpqSlUq1WfM98rfuVYLq5RpmxT0PI1UrzzXPMGgxz/1Wq1zPFLZ5ifIW8i8DORnxkFPQATECZ72PlZ82aBHKclhTlD1nQGs6IoiqIoiqKsP7e9wwwATz31FAqFAr7ru76r30sxeJ6Hl19+2ecsyl5mjjii0ysdYAZ/UXg1Gg0jgBuNhgnrorMqy7spuCiCOSaKLif7ldlrDCwFeVHksvyY47D4s3RyeVOCYpzr4rHJ+dLsy6WLGo/HsWPHDtx55504ePAgYrEYZmZm8OKLL+Ls2bMoFotwXdccM8VoNBo1vcSy1FyKeXnuZRCX53mmNBtYKknnvmOxmFkjkTOcef54DuQcZgBmNBXPGUU0zwV7yOUc5t4gMJ5DnmPtYVYURVEURVGUjWFLCOZ6vY7p6WnEYjFMTEz0ezmGVquFfD6PRCLhcy4pomQZMAWgdGqBFZHG8mk595fvIccZUUzKucSc+SvLuCV0p+V7yhFNDNaiCOV7U4Tbtu3riZYuL3uLu90uEokExsbGsG/fPoyPj8N1XVy4cAEvv/wyrl275gsjY9k3xT3nI3Od/J6913SXebOB56lWqxmx6nke6vU6ut0uotGocd3D4bApn+4t7aaTLOdky95lfi4U7fLmRe8cavm5tdttU9LOsvhut4tKpeKbzawoiqIoiqIoyvqxJQQzsOQyF4tFPPLII/1eio/Z2VkzD5jCTaYk081kWbEs8ZXl23L7tcYRydArlgL3hk5RmNEdlc4pRS5/pgsLwLjHnU7HBIYBSy453Wt5I4Cl0jwGx3GQSqWwa9cuHDhwANlsFpVKBSdOnMC5c+fMqCcejyxhZ6k0bxr0biPDseT4KtlXzbWzPJs3FmQpfCQS8Tn6PC4KWpZJ03XnvGaKZsJ98zORNy74GXCdzWbT52bzxoQGfimKoiiKoijKxrBlBDOwNNIpmUziyJEj/V6KwfM8XLx40bi1soRZJlPTGZbupRTKUhADKw6sRPYMAytjjqQo537obMuxVjJwiqJNCnUGX8mZxXKf0vmVTmw6ncbOnTtx+PBhxONxzMzM4NSpU5idnUWtVvO5sDJAi4ndXEe73UY4HDYimq45j0OmV1OsstScpejSIW82myiXy6jVar7Z1xTTwFLoWiwWM+dPzsvm+0pRLNfBcyxnQ8twNJ5X7oujrxRFURRFURRF2Ri2lGB+5plnMD8/j7vuumtV2XE/6Xa7OH/+PIClMUcy7ImOI8WqFFx8nKKKjioFIAWl53mrxlNRgMkxRhRkdLzprMqSb8LSZenE0p3ltjJ0i/vlGrimdDptnOVIJIKLFy/i/PnzKJfLq0Zf0WXmuljiHQ6HjRhlebk8RwBMubp0lAOBgHHKe8uce91oee55PDJhnKFfPAecsW3btrmBQPFOl1h+pr2p2hTm3D+3095lRVEURVEURdk4Bkc1bhAvvvgiEokE7r///n4vxUen08G5c+d8pbxytjHdSzlzea19yD5mAGbeM8WmnMUsRxVRHEp3la/h+wMroi0Wi/n2Q4HZ61rLQDEZKuY4DsbGxnDXXXdh586daLfbOHXqFC5duoTZ2VksLCygWCyaUuRkMukrH+f+5JgqmZbNHuFekRsKheA4ji8pOx6P+0rEZXk19ycDyxgyxm1lHzgdeTlPW44L6+0tl6X4RM56ljOZc7mcb+azoiiKoiiKoijry5YTzJcuXUKr1cKePXsGbqZtp9Mx4VZyFBKRSdlyfJR0PCki6VpKB1oGbXE7/kux2+l0YNu2EefSgQbgG9kErAjj3rUAK+XadFOlU57JZLB7926MjIygWq3i1KlTOH/+PBYXF30ONQUr90ORKZOmmc4tS5opinmuHMeBZVkol8tGjHLElRTcPG6WdQMwNxISiYTPRec5azabppdZ9pfTIV/rHMoZy8FgEM1m07wPy+3l+VIURVEURVEUZePZcoIZAJ544gkEg0G87W1v6/dSVtFut3Hx4kU0Gg04jmMcSBkeRRHFnymAmXDNbeWoIgpl7o8Cjc8DMOOPZCiVFNytVsvMJW61WqZnGVhJm242m7Bt2ycS5Sxkx3Gwe/duHDlyBKOjo5iZmcHp06dx7do1I3y5NnlTgMdIESr7iGUYmZxfTGEt+7F57nrDtrhPimneOKBrzeOgs871dbtdtFotM6NZns9eAS37meV7y+d5k4D753EzaExRFEVRFEVRlI1jSwrmXC6HUqmEbDbrm787KHQ6HUxNTSGXy/nGGsnyaIovfl0vkIvijqFbkl4BKd1iijWWGcvnuCbP8xCJRMy/Mh2bUCgDS/ONs9ksDhw4gFQqhfn5eVy8eNGkYPP9ZM8yE7DpAtP5ZRAYRTH7mpnwzZFQct6y7OnmTYBarWZGQsnebznWCYB5X97E4M0Anms66yyB52ukQ87zKkUzxb0UzzLZnJ8zg8cURVEURVEURdk4tqRgBoAvf/nL2LNnDx599NGBCgAj3W4X+XweV65cQa1WQ7fbNYnMcr1SKMuQrN6SbulYyhFN8jn2x8pybekoA/CVM8sQqlAohEajgUgkAtd1jZtNsZ1MJrF7924cOnQIAHDx4kWcPn0a5XIZ9Xrd5xzTFY5GoyZ0TAZiyZAz3gzgYzLlGlgR7xSbFM6yLJp9xtKpBpZEtxTTfH/bts2NgnA47Lup0Huzgp+B7GMmdP1Zri5L6RmiJnutFUVRFEVRFEXZWEI33uT2pNPpYGZmBmNjY5iYmMDU1FS/l7Qm7XYbU1NTCIfDGBsbQzwe95XwyqAu2TPMUl7ZP9w71kjOHw4Gg0bAASsjjmQ4FUWc4zirBDP7ceX7c2xTMpnEtm3bsHv3briui0uXLmF6ehqlUsnXa0zxL+dNU8zL/uRoNOrr7WYPMcW1LDenqJZ9x+zBlvtnajgd7kAg4NuPFLA8557nmdAvHoMs45ahabLvnAnaPP+yhJvnXjrTlUrFjBhTFEVRFEVRFGXjGDxrdQP5/Oc/j0AggLe+9a2m/3RQabVamJycxIULFzA3N2dElwy/kiJS9jDL0CzpHlP0yRRrOeqIY5LonMr+aApclkw3m01fDy9nIkejUezevRs7d+6E67o4ffo0pqenjZvN/VCA062lAJeuaygUQiwWQzqdRiaTQTqdRiwWQywWQygUQrPZNI6wTJpmHzGFsUyelqXe/FeeU1kyLWdeU1jzNTx2OZdann/p/stzLJHjsfi5tFotVKtVLcdWFEVRFEVRlD6wZR1mcu7cOTz88MOYnJzE8ePH+72cG9Jut5HP55HP5xGJRJBOp2HbNhzHMcKWI44ouKSglgncwIqY5PN8nMKSYpZlyRTJdEdlKbIsaQ6FQhgbG8PY2BjS6TRyuRymp6cxNTXlE+2yvFymSveWgNMB5/xlGQwGwAh36ZCznJmiWB6XfE+6t9yeZdwU3wCMy87vXdc1a6HAZ+k6b0bI/mauicKfNx3kMUgBzp/r9bqWYyuKoiiKoihKn9jSDjMAfOUrX0EgEMAdd9wBx3H6vZzXRbPZRC6Xw7Vr10yJs+u6PidZ9vgCK32zFLsUer1up5y7LNOw5WsCgYAJJLNt2wjPYDCIbDaLPXv2IB6PY3Z2FlevXsXs7CyazaZPLMs0bY6SYvm0dJYpYpvNJqrVKiqVCqrVKsrlMiqViq8MXM5k5kxlOrvSAeZxep5nwrxarZYvgVsGrbFMWp4bnmueT4p+uQ3POx1mAKuc/V5Hm4K/Xq9rOraiKIqiKIqi9Ikt7zADS2OmHnvsMRw5cgQvvfRSv5fzhqhWq6hWqwCAeDyOTCbjc2fpArMXmKJQ/iznFrMcG4BxaGVwVW8ZN/fhOA7i8TjGx8fRbrcxNzeHyclJsz2db1lyDayEcsm18X34nLwZIEO8uC0Dv3pHR1HIruVoczwW186Qs95Z19wvzyEdbhm61iuG+b3sz6aYZ0m6XCudZ65bR0kpiqIoiqIoSn9RwQzgqaeewt133409e/bg/PnzqFQq/V7Sm4Li2bIsxGIxpFIp02PLsUvAiphjKTZFLIOsKBgpnvm9DBpj+TewUt7daDRQKBSQz+dRLpeNqyxHLlHEsmcZgE+c8nvpjNPdZq+yZVmwbdvn0sqy7N5UaoprmSLO/XS7XSNi6/W6GTfG/UrHeK2ecRnYBcCUZkv3XPbJy5sPa+273W6jUqmYXnVFURRFURRFUTYeFczLPP744/jYxz6Gubk5PP3006vE1mbE8zyf8xyLxZBIJBCJRNYMwQL8olI6wb09t3w9e3grlQoKhQJKpRIAYP/+/WYfFJ0sS5ZBW3xPCkXZ90s3WiZGSzHcO9MYWHGquWbpKNPFlaXeXAuTuWUiuDxWbi+Tr3njoNFomCRyYKVsnftstVq+2dQATAo3nW25T2Cp9JzjuRRFURRFURRF6Q8qmJeZnJzEyZMnsWvXLpw+fRqLi4v9XtItp1aroVarAYCZ6RyJRNDpdMw8YRlKxZnDUrRRSAIw7mw+n0e9Xjfb2LZtRLcsi+a/a4V+9QrfcDjs6xsG/K4sS8Hp6MqeZwr7XjdYPsZgLrlOWYLeOzdZ9hY3m01zXljSTXqDvPhaKcQlspxcpmNXq1V1lxVFURRFURSlz9ww9MuyLMeyrGcsy3rJsqxXLcv6d8uP77Ms62nLss5ZlvUnlmVFlh+3l38+t/z83nU+hlvGn/3Zn2HHjh246667VjmCtxuu6yKfz2NmZga5XA7z8/MoFAqoVCqo1+uoVqtGYDNkq1AooFgsYmZmBpcvX8bU1BQmJyd9YhkAstmsEZO985Xp6LIXWZZHAyvl2OwrlrOhuU2vc0ynmGJT9ixL4SqdcQBmbrIMH5P74ONMqqYQ5rbSbaZwl+O6+HOveJa948BKwjfX2Gg0VCwrq9hK12JFUZRBRK/DirI1uZmU7AaARz3PeyuAowC+07KsdwD4/wD4hOd5BwEsAvjo8vYfBbC4/PgnlrfbFHQ6HZw6dQr33nsvRkdH+72cDaXZbKJYLGJ+fh65XA75fB4LCwuYmprCzMwM5ubmkM/nUalUViVqS5iaDfjLqylE5bxj2bPM8m4iha90mHvnG1NU83H5HlKsXk9k8zkp4ntFNsUv18F+b5lELnuse0vNHccxQpvHxf30zl1ut9uo1WomcVxRBFvmWqwoijKg6HVYUbYgNxTM3hJMwQovf3kAHgXw2eXHPwXgg8vff2D5Zyw//22WVCADzuOPP45kMom77rpr042ZGgTi8bhPdMqPnkncdGFDoRAikYhxmKUrLUurZRJ2734pMplYLUU2y5wlrByg083QM+k6Ez4m068pcKWIX+tfbi+/l/vuFf/8t1qtrnLsFQXYetdiRVGUQUOvw4qyNbmpOcyWZQUty3oRwByALwM4D6DgeR7Toq4B2LH8/Q4AVwFg+fkigOE19vkxy7KesyzruTd1BOvA17/+dbz97W/H+Ph4v5ey6ZClxhw3tZarzP5pimM5/1gKbhlOxp+JLNHuFdTycdlTLPuNub0U7LKEXO6PAV9SWHc6HbOtLOGW66cDzm0k8rXctlar6Sgp5bqs97W4BW0FUBRFeS30OqwoW4+bEsye53U8zzsKYCeAhwDc8Wbf2PO83/Y87wHP8x54s/u61Xzta19DOBzGww8/jGQy2e/lbBromkpHlgKaXwB8idRMg5blyqFQyDjBsne4d98yQZs9xb2CWAaY0XGmSJYuN7eVoVxcM1/HY5E3BeR+egPOZCk3j0OOjeqdJV2tVs1oK0VZi/W+Fodh3/gFiqIoWxi9DivK1uOmBDPxPK8A4EkAxwBkLMti0+lOAJPL308C2AUAy8+nASzcisVuJI8//jgee+wxMx5JuTFyzrAsWbaWR0QxYEsKXEIxLF1jwt5gIsc39QpQ6dpKAfta46dYei/Hbcl9E+lAy6Ax6ULLtcqya9lfLUvF+XytVkOlUlm1TkVZi610LVYURRlE9DqsKFuHm0nJHrUsK7P8fRTAYwBOYuki8X3Lm/0wgL9Y/v7zyz9j+fm/8zbhMNmnnnoKV65cwaOPPopMJtPv5WwK2Jvc66bKlGs+5rqurwSb0L2l+Fwr5ZqilF9yZjR7l3t/5WR4GLAkohmsJXuJZS9y72t6g7i63S5arZZ5LzkKS74/102k287ty+WyusvKa7JVr8WKoiiDgl6HFWVrcjPW6TYAT1qWdRzAswC+7HneXwH4fwD4F5ZlncNSP8Ynl7f/JIDh5cf/BYCfu/XL3hj+9E//FPfeey/e9ra3wba1ROZmkEJQpmVLgUvoCktxKsO7gBURztJnKTwpRKUzzdfwZ24vnWfuX46HolPc6y5ToEsRD6zMTZbjpfh6utdco+zBlgnaXEOtVtOgL+Vm2LLXYkVRlAFBr8OKsgUJ3WgDz/OOA7hvjccvYKl3o/dxF8CHbsnq+syJEyfwx3/8xzh27BheeuklzM3N9XtJAw0FLN1lOf5JOq9SEEtkSTPLl4EVB7jXjeZj0lEOBoPGZeaYqt75zb3bUqz2CmYK9Xa7bZzvRqNhkr2BlX5lHiMf43tJR5zPtdttc65c10W5XH7NUV2KAmzta7GiKMogoNdhRdma3FAwb3W+9KUvodVqYWxsDIuLi5pgfAPYo7xWSbTs+fU8D+FweFXYFwCfa0vhTXG7FhTjAGDbthGuoVAIzWZz1egmWeYdCoXguq4RyFLcSgdbvq7T6ZhQMrrJPK5IJIJGo2GEce+ILZ4XHnO1WtWZy4qiKIqiKIoyoGia1U3w5JNPIpfLrRKBih8ZZCWd5OuNaQJWBDa3oyPL/ckS5rX6g7kv9i/3imoKV4pxCmPObwZgnuPaZV8018fn+DiDvwj32duDLQPH+LpwOIxwOIx2u41qtfpmTrmiKIqiKIqiKOuICuabZHZ21tebqqym2Wz6BKtlWeYx6S5TeLLXmMKYPcq9AVzyZyZss2Sb+5Gp1bK8meJ1raRzVgvwfaWwlgFisqSbs6M7nQ4cx0EkEkGn0zHHwtFWskeZ64hEIohEInAcB+12G/l8Xn+nFEVRFEVRFGWA0ZJs5ZYhw6+AlTnMFLjASogWBa5MviYUnNKZlcFZHB8lRbD8XpZ3ywRsPkdx67oums2mGYfVK+LpIrNUW7rbPE7ZP03RDayUc3POM9fDfS0sLKBWq92qU68oiqIoiqIoyjqgDrNyS2k0GgD84VtEil5uw3/lnGLZQ8zHWM5Mx1i6yxTcdK+5P5mQLQW5FOAM/ZKp3L1r5WMs+ZZzlOkkczs65LZtw3EcJBIJxGIxJBIJI8wXFxdRKpVu5WlXFEVRFEVRFGUdUMGs3FJKpZJvJJMcG0UnliKaorY3cRqA2Z6CVPYtS1EshbPsd2YZOB1j2a8sBbgM+mJptW3bpsdYJnNz9BUdbpmozf07jmO+EomE+Z5p23Nzc5iZmblugJmiKIqiKIqiKIODlmQrt5RGo2FcWIpECs52u23SpSk++T2RSdKyPJpilkKW2zLUS/7M9wJWgrr4etmjzHLrZrNphLR0k7lGCnKmWcvxWFyfbduIxWKmP5nvz75oz/NQKBQwOzurSeuKoiiKoiiKsklQh1m5pXieh0qlYoSv7GvmY3IUEwDTCyzdaM4qlqOl5Gu5XwZ2cT9yv3SCI5GIEbfSjeZ+WSodiUSMeOa+Go2GeZ38NxQKwXEcZDIZJJNJJJNJhEIhhMNh3w0CAHBdF7VaDdPT0yqWFUVRFEVRFGUToQ6zcsspFouIx+OIRqMA4HN4pZtMQQusOMsyNAxY6XOWY6F6U6gjkYjvNXyPTqdjXGHHcXz90XSG6TCz1JrroviVadmyHzsWiyESicDzPLiuax5vtVq+0nJgKT18dnbW9HcriqIoiqIoirI5UMGs3HLa7TZqtZpxbntdYorbVqvle4wjnNh/DPhHSvFfCm+WdLdaLTPOiY8xCEy61hwFRfEuS6fpOMsRVxTitm0bkS5TvYPBIFqtls/RlknbHEFVKBQwPz+/QWdfURRFURRFUZRbhQpmZV1YXFxEMBhENBr1lUxT0Pb2IjN8i8KU4lMKVYrncDgMy7Lguq4Ry3SKAfhSq9vttpmV3Gq10G63Ydu2TyBHIhFfKFg4HEYikfCNoAoEAub1XBt/JtwfnfVIJIL5+XlcuXJFQ74URVEURVEUZROigllZFzqdDgqFgm8GMeCfscwwLLq20n2mW9ybrk0xLHuXZUiY/JlIN7jVaiEajZrkagaB8f3j8TgajYavBJtueTgchuu6sG0brusaF5rCenh4GO12G5lMBsFgELOzs3j11VeNkFcURVEURVEUZXOhgllZN5rNJsrlsilPpjPrOA7q9bqv3Jq9yzI5m0KW2zqOY/YtnV32HAMrbrUUw5FIxLjMnIlsWRZarZbZliXdTPHm/tnHLOc8y/LxaDSKaDQKz/MwNjYGYMlZzuVyePrpp339zYqiKIqiKIqibC5UMCvrhud5KJVKRrDSiaX4ZAI1hTFFKgWqnN8s+5GbzaYR3iztZil1s9mE4zjGIaZIDofDZg3xeBzdbtfXz1yv1837hsNhVKtVADDCmPOZW60WksmkKfUeGxsz6d2cu3zu3Dl8/etfR61W6+fpVxRFURRFURTlTaKCWVlXut0uFhYWzGgn9g5TBLOMGliZ0yyTsen6yjnM7GF2HMeI3Hg8bsQzQ7qAJZebYjkQCKDRaKDb7ZrH2dsMwAST0eV2HMe4x+FwGNlsFqlUCtFoFJFIBI1GA4lEwvQ1h0IhtFotTE9Pa8iXoiiKoiiKotwGqGBW1p1Op4NSqYSdO3ea+cWe58FxHDQaDSNAKWqbzSaCwSBisRgAmNnGLImmI83yaQDmtQDMTOd2u22caT5GsUzhDKykeNu2DcuyEI/Hzc8jIyPmtbFYzBcuxpLvRCKBWCyG7du3o1Ao4Pnnn1/VR60oiqIoiqIoyuZDBbOyIVQqFSOao9GoEZ62bZtgMIpeit14PA4AxgFmGbScpdzpdIwIDgaDZr4ykXOdpXimA82y7GAwaEqqQ6EQkskkLMvCxMSEeR0JBAJIJpOYmJjAPffcg6NHj6JYLOKuu+5aNZdZURRFURRFUZTNiwpmZUPodruYnZ1FIpEwjjJnHstAMAre6yVgN5tNX39zt9tFq9Uy5dXhcNi42DKoiz9TmNu2jWg06gv6SiQSiMfjCAaDqNVqZv8cMWXbNrLZLO69914cPXoUgUAAtVoNu3fvhud5KJfL/Tm5iqIoiqIoiqKsCyqYlQ2j1Wrh3LlzcF0X4+PjAGAcYhn05bqumY9Md1cKaSZg82d+sf85EAgYtzoajSIejyMWi6FSqZiwL/ZCU7Sz1DsWi8GyLCOqd+zYgaNHj2LPnj2wbRvdbhfDw8PGofY8D5VKpZ+nVVEURVEURVGUdUIFs7KhtNttXLt2DZ7nYXh42NeH3FsuzceAJWHN4DCWWEciEViWZXqPHccxgV/JZNKI5mw2a/qQOUeZY6Wq1arZLpVKYfv27dizZw+OHDmC4eFhI9CTyaRZk6ZfK4qiKIqiKMrWQAWzsuG0221MTk7CsiyMjIwgFosZ95ajoSiEWb4dCoVMiTXLrukKh8NhU15NQc1063q9bkLCAoEAYrEYotEobNtGJpNBKBTC7t27ce+99yIajcKyLBP8JXuhG41GH8+YoiiKoiiKoij9QAWz0he63S5KpRJSqRQSiQQSiYQpp+52u77eYj7O52Q5NbCSkM3tWeZNER6NRpHNZpFMJo2DPDY25nOy9+zZg5mZGfMzg8YURVEURVEURdm6qGBW+kK320WxWEQkEjHjmkZHR2HbtinDtm0bwJIAZniX53mmx9jzPBPixf7jeDyOeDyO4eFh7N+/H2NjY2ZMFABYloW9e/fi2rVrvvVw9rKiKIqiKIqiKApRwaz0jU6ng5mZGZRKJRQKBbiui+3bt8NxHEQiEeMk8ysYDCISiSCRSCCbzWJoaAjDw8OYmJjA0NCQcZhfeeUVeJ6HiYkJxONx3H///Th+/Lh5X1lqrSiKoiiKoiiKcj1UMCsDQSaTwfj4OA4fPoz9+/dj9+7dGB0dRSaTQTQaheM4a77ux3/8x/Fbv/VbG7xaRVEURVEURVG2AiqYlb5Tq9Xw1FNP4Zvf/Cb27NmD3bt3Y9euXdi7dy9OnDiBP/mTP+n3EhVFURRFURRF2YJYnuf1ew2wLCsHoApgvt9reZOMQI9hENBjeH3s8TxvdIPeSxlgLMsqAzjd73W8SfS//8FAj+H1oddhBYD+TTxg6DEMBn2/Fg+EYAYAy7Ke8zzvgX6v482gxzAY6DEoyhvjdvi902MYDPQYFOWNczv87ukxDAZ6DLeGQD/fXFEURVEURVEURVEGFRXMiqIoiqIoiqIoirIGgySYf7vfC7gF6DEMBnoMivLGuB1+7/QYBgM9BkV549wOv3t6DIOBHsMtYGB6mBVFURRFURRFURRlkBgkh1lRFEVRFEVRFEVRBoa+C2bLsr7TsqzTlmWdsyzr5/q9npvFsqxLlmW9bFnWi5ZlPbf82JBlWV+2LOvs8r/Zfq9TYlnW71qWNWdZ1ivisTXXbC3xa8ufy3HLsu7v38pXuM4x/FvLsiaXP4sXLcv6LvHcv1o+htOWZX1Hf1btx7KsXZZlPWlZ1gnLsl61LOunlh/fVJ+Fcnuh1+KNQ6/F/b8W63VYGUT0Orxx6HW4/9dhYPNci/sqmC3LCgL4DQDvA3AXgB+wLOuufq7pdfIez/OOiqjznwPwhOd5hwA8sfzzIPF7AL6z57Hrrfl9AA4tf30MwG9u0BpvxO9h9TEAwCeWP4ujnud9AQCWf5e+H8Ddy6/5r8u/c/2mDeBnPM+7C8A7APzE8lo322eh3CbotXjD+T3otbjf12K9DisDhV6HN5zfg16H+30dBjbJtbjfDvNDAM55nnfB87wmgM8A+ECf1/Rm+ACATy1//ykAH+zfUlbjed5XAeR7Hr7emj8A4Pe9JZ4CkLEsa9uGLPQ1uM4xXI8PAPiM53kNz/MuAjiHpd+5vuJ53rTneS8sf18GcBLADmyyz0K5rdBr8Qai1+L+X4v1OqwMIHod3kD0Otz/6zCwea7F/RbMOwBcFT9fW35sM+AB+JJlWc9blvWx5cfGPc+bXv5+BsB4f5b2urjemjfbZ/PPlkszfleU/Qz8MViWtRfAfQCexu3zWSibj838O6bX4sFi012L9TqsDAib+XdMr8ODxaa7DgODfS3ut2DezHyL53n3Y6k04Ccsy3qXfNJbih/fVBHkm3HNy/wmgAMAjgKYBvArfV3NTWJZVgLA5wB83PO8knxuE38WirLR6LV4cNh012K9DivKLUGvw4PDprsOA4N/Le63YJ4EsEv8vHP5sYHH87zJ5X/nAPwPLJU1zLIsYPnfuf6t8Ka53po3zWfjed6s53kdz/O6AP4bVkpMBvYYLMsKY+nC8Iee5/358sOb/rNQNi2b9ndMr8WDw2a7Fut1WBkwNu3vmF6HB4fNdh0GNse1uN+C+VkAhyzL2mdZVgRLzeif7/OabohlWXHLspL8HsB7AbyCpbX/8PJmPwzgL/qzwtfF9db8eQA/tJxG9w4ARVEaMVD09C58L5Y+C2DpGL7fsizbsqx9WAoIeGaj19eLZVkWgE8COOl53q+Kpzb9Z6FsWvRa3H82/X//m+larNdhZQDR63D/2fT//W+m6zCwia7Fnuf19QvAdwE4A+A8gH/T7/Xc5Jr3A3hp+etVrhvAMJaS3M4C+FsAQ/1ea8+6/xhL5RktLNX8f/R6awZgYSmt8TyAlwE80O/1v8Yx/MHyGo9j6T+kbWL7f7N8DKcBvK/f619e07dgqbTkOIAXl7++a7N9Fvp1e33ptXhD163X4v6vX6/D+jVwX3od3tB163V4MI5hU1yLreU3VxRFURRFURRFURRF0O+SbEVRFEVRFEVRFEUZSFQwK4qiKIqiKIqiKMoaqGBWFEVRFEVRFEVRlDVQwawoiqIoiqIoiqIoa6CCWVEURVEURVEURVHWQAWzoiiKoiiKoiiKoqyBCmZFURRFURRFURRFWQMVzIqiKIqiKIqiKIqyBv9/7gR3olLzEWgAAAAASUVORK5CYII=",
       "text/plain": [
        "<Figure size 1296x432 with 3 Axes>"
       ]
@@ -812,8 +816,8 @@
     "    img_name = os.path.split(val_ds[case_num][\"image_meta_dict\"][\"filename_or_obj\"])[1]\n",
     "    img = val_ds[case_num][\"image\"]\n",
     "    label = val_ds[case_num][\"label\"]\n",
-    "    val_inputs = torch.unsqueeze(img, 1).cuda()\n",
-    "    val_labels = torch.unsqueeze(label, 1).cuda()\n",
+    "    val_inputs = torch.unsqueeze(img, 1).cuda(device=device)\n",
+    "    val_labels = torch.unsqueeze(label, 1).cuda(device=device)\n",
     "    val_outputs = sliding_window_inference(\n",
     "        val_inputs, (96, 96, 96), 4, model, overlap=0.8\n",
     "    )\n",
@@ -854,7 +858,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "Python 3",
+   "display_name": "Python 3 (ipykernel)",
    "language": "python",
    "name": "python3"
   },
@@ -868,7 +872,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.7.10"
+   "version": "3.8.0"
   }
  },
  "nbformat": 4,

From 9106ce52354b8ab273ca9e873dbf0a054a59e35c Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Sat, 9 Jul 2022 17:07:17 -0700
Subject: [PATCH 02/17] checkpoint, added amp, todeviced, threaddataloader

---
 .../swin_unetr_btcv_segmentation_3d.ipynb     | 265 +++++++++++-------
 1 file changed, 161 insertions(+), 104 deletions(-)

diff --git a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
index 82c9b0ffe4..2a13cefd46 100644
--- a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
+++ b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
@@ -81,23 +81,50 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [
     {
-     "data": {
-      "text/plain": [
-       "False"
-      ]
-     },
-     "execution_count": 5,
-     "metadata": {},
-     "output_type": "execute_result"
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.8/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"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Sat Jul  9 05:14:00 2022       \n",
+      "+-----------------------------------------------------------------------------+\n",
+      "| NVIDIA-SMI 450.119.04   Driver Version: 450.119.04   CUDA Version: 11.6     |\n",
+      "|-------------------------------+----------------------+----------------------+\n",
+      "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
+      "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
+      "|                               |                      |               MIG M. |\n",
+      "|===============================+======================+======================|\n",
+      "|   0  Tesla V100-SXM2...  On   | 00000000:0A:00.0 Off |                    0 |\n",
+      "| N/A   40C    P0    43W / 163W |      3MiB / 32510MiB |      0%      Default |\n",
+      "|                               |                      |                  N/A |\n",
+      "+-------------------------------+----------------------+----------------------+\n",
+      "                                                                               \n",
+      "+-----------------------------------------------------------------------------+\n",
+      "| Processes:                                                                  |\n",
+      "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
+      "|        ID   ID                                                   Usage      |\n",
+      "|=============================================================================|\n",
+      "|  No running processes found                                                 |\n",
+      "+-----------------------------------------------------------------------------+\n"
+     ]
     }
    ],
    "source": [
+    "# TEMP CELL PLEASE DELETE\n",
     "import torch\n",
-    "torch.cuda.is_available()"
+    "torch.cuda.is_available()\n",
+    "\n",
+    "!nvidia-smi"
    ]
   },
   {
@@ -109,8 +136,10 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 1,
-   "metadata": {},
+   "execution_count": 2,
+   "metadata": {
+    "scrolled": true
+   },
    "outputs": [
     {
      "name": "stdout",
@@ -119,9 +148,9 @@
       "Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
       "Collecting nibabel==3.1.1\n",
       "  Downloading nibabel-3.1.1-py3-none-any.whl (3.3 MB)\n",
-      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.3/3.3 MB\u001b[0m \u001b[31m63.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n",
-      "\u001b[?25hRequirement already satisfied: packaging>=14.3 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (21.3)\n",
-      "Requirement already satisfied: numpy>=1.13 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (1.22.3)\n",
+      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.3/3.3 MB\u001b[0m \u001b[31m106.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+      "\u001b[?25hRequirement already satisfied: numpy>=1.13 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (1.22.3)\n",
+      "Requirement already satisfied: packaging>=14.3 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (21.3)\n",
       "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.8/site-packages (from packaging>=14.3->nibabel==3.1.1) (3.0.8)\n",
       "Installing collected packages: nibabel\n",
       "  Attempting uninstall: nibabel\n",
@@ -133,7 +162,7 @@
       "\u001b[0mLooking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
       "Collecting tqdm==4.63.0\n",
       "  Downloading tqdm-4.63.0-py2.py3-none-any.whl (76 kB)\n",
-      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m76.6/76.6 kB\u001b[0m \u001b[31m18.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m76.6/76.6 kB\u001b[0m \u001b[31m32.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
       "\u001b[?25hInstalling collected packages: tqdm\n",
       "  Attempting uninstall: tqdm\n",
       "    Found existing installation: tqdm 4.64.0\n",
@@ -146,8 +175,8 @@
     }
    ],
    "source": [
-    "# FIXME: check version compatibility\n",
-    "# !pip install git+https://github.com/Project-MONAI/MONAI#egg.gitmonai@0.8.1+271.g07de215c \n",
+    "# FIXME: compatible with 0.9.0 but not with 0.9.1rc\n",
+    "# !pip install 'git+https://github.com/Project-MONAI/MONAI#egg.gitmonai@0.8.1+271.g07de215c'\n",
     "\n",
     "!pip install nibabel==3.1.1\n",
     "!pip install tqdm==4.63.0\n",
@@ -157,7 +186,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [
     {
@@ -210,6 +239,7 @@
     "import matplotlib.pyplot as plt\n",
     "import numpy as np\n",
     "from tqdm import tqdm\n",
+    "import time\n",
     "\n",
     "from monai.losses import DiceCELoss\n",
     "from monai.inferers import sliding_window_inference\n",
@@ -227,6 +257,8 @@
     "    Spacingd,\n",
     "    RandRotate90d,\n",
     "    ToTensord,\n",
+    "    EnsureTyped,\n",
+    "    ToDeviced\n",
     ")\n",
     "\n",
     "from monai.config import print_config\n",
@@ -234,6 +266,7 @@
     "from monai.networks.nets import SwinUNETR\n",
     "\n",
     "from monai.data import (\n",
+    "    ThreadDataLoader,\n",
     "    DataLoader,\n",
     "    CacheDataset,\n",
     "    load_decathlon_datalist,\n",
@@ -254,19 +287,19 @@
     "\n",
     "You can specify a directory with the `MONAI_DATA_DIRECTORY` environment variable.  \n",
     "This allows you to save results and reuse downloads.  \n",
-    "If not specified, a temporary directory will be used."
+    "If not specified a temporary directory will be used."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "/tmp/tmpo69cno4c\n"
+      "/tmp/tmpnv0d_576\n"
      ]
     }
    ],
@@ -286,13 +319,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 13,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [],
    "source": [
     "num_samples = 1\n",
+    "os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\"\n",
+    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
     "\n",
-    "# TODO: EnsureTyped + ToDeviced\n",
     "train_transforms = Compose(\n",
     "    [\n",
     "        LoadImaged(keys=[\"image\", \"label\"]),\n",
@@ -312,6 +346,8 @@
     "            clip=True,\n",
     "        ),\n",
     "        CropForegroundd(keys=[\"image\", \"label\"], source_key=\"image\"),\n",
+    "        EnsureTyped(keys=[\"image\", \"label\"]),\n",
+    "        ToDeviced(keys=[\"image\", \"label\"], device=device),\n",
     "        RandCropByPosNegLabeld(\n",
     "            keys=[\"image\", \"label\"],\n",
     "            label_key=\"label\",\n",
@@ -347,7 +383,7 @@
     "            offsets=0.10,\n",
     "            prob=0.50,\n",
     "        ),\n",
-    "        ToTensord(keys=[\"image\", \"label\"]),\n",
+    "#         ToTensord(keys=[\"image\", \"label\"]),\n",
     "    ]\n",
     ")\n",
     "val_transforms = Compose(\n",
@@ -364,7 +400,9 @@
     "            keys=[\"image\"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True\n",
     "        ),\n",
     "        CropForegroundd(keys=[\"image\", \"label\"], source_key=\"image\"),\n",
-    "        ToTensord(keys=[\"image\", \"label\"]),\n",
+    "#        ToTensord(keys=[\"image\", \"label\"]),\n",
+    "        EnsureTyped(keys=[\"image\", \"label\"]),\n",
+    "        ToDeviced(keys=[\"image\", \"label\"], device=device),\n",
     "    ]\n",
     ")"
    ]
@@ -382,30 +420,33 @@
     "\n",
     "3. Make a JSON file to define train/val split and other relevant parameters. Place the JSON file at `./data/dataset_0.json`.\n",
     "\n",
-    "    An example of the JSON file is provided in the current directory. If you would like to use this directly, please move it into the `./data` folder.   "
+    "   You can download an example of the JSON file [here](https://drive.google.com/file/d/1EF2By3k1NWDIIoH4r_3Xj3Q9yUDxoF0U/view?usp=sharing). If you would like to use this directly, please move it into the `./data` folder.   "
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "Loading dataset: 100%|███████████████████████████████████████████████████████████████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Loading dataset: 100%|█████████████████████████████████████████████████████████████████████| 6/6 [00:10<00:00,  1.81s/it]\n"
+      "Loading dataset: 100%|██████████| 24/24 [00:55<00:00,  2.29s/it]\n",
+      "Loading dataset: 100%|██████████| 6/6 [00:19<00:00,  3.17s/it]\n"
      ]
     }
    ],
    "source": [
+    "torch.cuda.empty_cache()\n",
     "data_dir = \"data/\"\n",
     "split_JSON = \"dataset_0.json\"\n",
     "\n",
     "datasets = data_dir + split_JSON\n",
     "datalist = load_decathlon_datalist(datasets, True, \"training\")\n",
     "val_files = load_decathlon_datalist(datasets, True, \"validation\")\n",
+    "\n",
+    "# TODO: try thread_workers\n",
     "train_ds = CacheDataset(\n",
     "    data=datalist,\n",
     "    transform=train_transforms,\n",
@@ -413,17 +454,13 @@
     "    cache_rate=1.0,\n",
     "    num_workers=8,\n",
     ")\n",
+    "# train_loader = ThreadDataLoader(train_ds, batch_size=1, shuffle=True, num_workers=0)\n",
+    "train_loader = ThreadDataLoader(train_ds, batch_size=4, shuffle=True, use_thread_workers=True, num_workers=1)\n",
     "\n",
-    "# TODO: try ThreadDataLoader (and num_workers)\n",
-    "train_loader = DataLoader(\n",
-    "    train_ds, batch_size=1, shuffle=True, num_workers=8, pin_memory=True\n",
-    ")\n",
     "val_ds = CacheDataset(\n",
     "    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4\n",
     ")\n",
-    "val_loader = DataLoader(\n",
-    "    val_ds, batch_size=1, shuffle=False, num_workers=4, pin_memory=True\n",
-    ")"
+    "val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)"
    ]
   },
   {
@@ -435,7 +472,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [
     {
@@ -495,13 +532,10 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [],
    "source": [
-    "os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\"\n",
-    "device = torch.device(\"cuda:1\" if torch.cuda.is_available() else \"cpu\")\n",
-    "\n",
     "model = SwinUNETR(\n",
     "    img_size=(96, 96, 96),\n",
     "    in_channels=1,\n",
@@ -522,7 +556,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 6,
    "metadata": {
     "scrolled": false
    },
@@ -534,7 +568,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
@@ -560,7 +594,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -578,61 +612,19 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": 8,
    "metadata": {
     "scrolled": true
    },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (X / X Steps) (loss=X.X):   0%|                                                          | 0/24 [00:00<?, ?it/s]Traceback (most recent call last):\n",
-      "  File \"/opt/conda/lib/python3.8/multiprocessing/queues.py\", line 245, in _feed\n",
-      "    send_bytes(obj)\n",
-      "  File \"/opt/conda/lib/python3.8/multiprocessing/connection.py\", line 200, in send_bytes\n",
-      "    self._send_bytes(m[offset:offset + size])\n",
-      "  File \"/opt/conda/lib/python3.8/multiprocessing/connection.py\", line 411, in _send_bytes\n",
-      "    self._send(header + buf)\n",
-      "  File \"/opt/conda/lib/python3.8/multiprocessing/connection.py\", line 368, in _send\n",
-      "    n = write(self._handle, buf)\n",
-      "BrokenPipeError: [Errno 32] Broken pipe\n",
-      "Training (X / X Steps) (loss=X.X):   0%|                                                          | 0/24 [00:01<?, ?it/s]\n"
-     ]
-    },
-    {
-     "ename": "RuntimeError",
-     "evalue": "CUDA out of memory. Tried to allocate 1.80 GiB (GPU 1; 15.75 GiB total capacity; 12.70 GiB already allocated; 1.45 GiB free; 13.13 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF",
-     "output_type": "error",
-     "traceback": [
-      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
-      "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
-      "Input \u001b[0;32mIn [15]\u001b[0m, in \u001b[0;36m<cell line: 88>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     87\u001b[0m metric_values \u001b[38;5;241m=\u001b[39m []\n\u001b[1;32m     88\u001b[0m \u001b[38;5;28;01mwhile\u001b[39;00m global_step \u001b[38;5;241m<\u001b[39m max_iterations:\n\u001b[0;32m---> 89\u001b[0m     global_step, dice_val_best, global_step_best \u001b[38;5;241m=\u001b[39m \u001b[43mtrain\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m     90\u001b[0m \u001b[43m        \u001b[49m\u001b[43mglobal_step\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtrain_loader\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdice_val_best\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mglobal_step_best\u001b[49m\n\u001b[1;32m     91\u001b[0m \u001b[43m    \u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     92\u001b[0m model\u001b[38;5;241m.\u001b[39mload_state_dict(torch\u001b[38;5;241m.\u001b[39mload(os\u001b[38;5;241m.\u001b[39mpath\u001b[38;5;241m.\u001b[39mjoin(root_dir, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mbest_metric_model.pth\u001b[39m\u001b[38;5;124m\"\u001b[39m)))\n",
-      "Input \u001b[0;32mIn [15]\u001b[0m, in \u001b[0;36mtrain\u001b[0;34m(global_step, train_loader, dice_val_best, global_step_best)\u001b[0m\n\u001b[1;32m     34\u001b[0m step \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m \u001b[38;5;241m1\u001b[39m\n\u001b[1;32m     35\u001b[0m x, y \u001b[38;5;241m=\u001b[39m (batch[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mimage\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mcuda(device\u001b[38;5;241m=\u001b[39mdevice), batch[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mlabel\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mcuda(device\u001b[38;5;241m=\u001b[39mdevice))\n\u001b[0;32m---> 36\u001b[0m logit_map \u001b[38;5;241m=\u001b[39m \u001b[43mmodel\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     37\u001b[0m loss \u001b[38;5;241m=\u001b[39m loss_function(logit_map, y)\n\u001b[1;32m     38\u001b[0m loss\u001b[38;5;241m.\u001b[39mbackward()\n",
-      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py:1111\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1107\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1108\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1110\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1111\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1112\u001b[0m \u001b[38;5;66;03m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1113\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[38;5;241m=\u001b[39m [], []\n",
-      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:280\u001b[0m, in \u001b[0;36mSwinUNETR.forward\u001b[0;34m(self, x_in)\u001b[0m\n\u001b[1;32m    279\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mforward\u001b[39m(\u001b[38;5;28mself\u001b[39m, x_in):\n\u001b[0;32m--> 280\u001b[0m     hidden_states_out \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mswinViT\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx_in\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mnormalize\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    281\u001b[0m     enc0 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mencoder1(x_in)\n\u001b[1;32m    282\u001b[0m     enc1 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mencoder2(hidden_states_out[\u001b[38;5;241m0\u001b[39m])\n",
-      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py:1111\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1107\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1108\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1110\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1111\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1112\u001b[0m \u001b[38;5;66;03m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1113\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[38;5;241m=\u001b[39m [], []\n",
-      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:971\u001b[0m, in \u001b[0;36mSwinTransformer.forward\u001b[0;34m(self, x, normalize)\u001b[0m\n\u001b[1;32m    969\u001b[0m x0 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mpos_drop(x0)\n\u001b[1;32m    970\u001b[0m x0_out \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mproj_out(x0, normalize)\n\u001b[0;32m--> 971\u001b[0m x1 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mlayers1\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m0\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx0\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcontiguous\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    972\u001b[0m x1_out \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mproj_out(x1, normalize)\n\u001b[1;32m    973\u001b[0m x2 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mlayers2[\u001b[38;5;241m0\u001b[39m](x1\u001b[38;5;241m.\u001b[39mcontiguous())\n",
-      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py:1111\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1107\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1108\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1110\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1111\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1112\u001b[0m \u001b[38;5;66;03m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1113\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[38;5;241m=\u001b[39m [], []\n",
-      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:841\u001b[0m, in \u001b[0;36mBasicLayer.forward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m    839\u001b[0m attn_mask \u001b[38;5;241m=\u001b[39m compute_mask([dp, hp, wp], window_size, shift_size, x\u001b[38;5;241m.\u001b[39mdevice)\n\u001b[1;32m    840\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m blk \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mblocks:\n\u001b[0;32m--> 841\u001b[0m     x \u001b[38;5;241m=\u001b[39m \u001b[43mblk\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mattn_mask\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    842\u001b[0m x \u001b[38;5;241m=\u001b[39m x\u001b[38;5;241m.\u001b[39mview(b, d, h, w, \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m)\n\u001b[1;32m    843\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdownsample \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
-      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py:1111\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1107\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1108\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1110\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1111\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1112\u001b[0m \u001b[38;5;66;03m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1113\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[38;5;241m=\u001b[39m [], []\n",
-      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:651\u001b[0m, in \u001b[0;36mSwinTransformerBlock.forward\u001b[0;34m(self, x, mask_matrix)\u001b[0m\n\u001b[1;32m    649\u001b[0m shortcut \u001b[38;5;241m=\u001b[39m x\n\u001b[1;32m    650\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39muse_checkpoint:\n\u001b[0;32m--> 651\u001b[0m     x \u001b[38;5;241m=\u001b[39m \u001b[43mcheckpoint\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcheckpoint\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mforward_part1\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mx\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mmask_matrix\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    652\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    653\u001b[0m     x \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mforward_part1(x, mask_matrix)\n",
-      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/utils/checkpoint.py:235\u001b[0m, in \u001b[0;36mcheckpoint\u001b[0;34m(function, use_reentrant, *args, **kwargs)\u001b[0m\n\u001b[1;32m    232\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mUnexpected keyword arguments: \u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;241m+\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m,\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;241m.\u001b[39mjoin(arg \u001b[38;5;28;01mfor\u001b[39;00m arg \u001b[38;5;129;01min\u001b[39;00m kwargs))\n\u001b[1;32m    234\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m use_reentrant:\n\u001b[0;32m--> 235\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mCheckpointFunction\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mapply\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfunction\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mpreserve\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    236\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    237\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m _checkpoint_without_reentrant(\n\u001b[1;32m    238\u001b[0m         function,\n\u001b[1;32m    239\u001b[0m         preserve,\n\u001b[1;32m    240\u001b[0m         \u001b[38;5;241m*\u001b[39margs\n\u001b[1;32m    241\u001b[0m     )\n",
-      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/utils/checkpoint.py:96\u001b[0m, in \u001b[0;36mCheckpointFunction.forward\u001b[0;34m(ctx, run_function, preserve_rng_state, *args)\u001b[0m\n\u001b[1;32m     93\u001b[0m ctx\u001b[38;5;241m.\u001b[39msave_for_backward(\u001b[38;5;241m*\u001b[39mtensor_inputs)\n\u001b[1;32m     95\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m torch\u001b[38;5;241m.\u001b[39mno_grad():\n\u001b[0;32m---> 96\u001b[0m     outputs \u001b[38;5;241m=\u001b[39m \u001b[43mrun_function\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m     97\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m outputs\n",
-      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:591\u001b[0m, in \u001b[0;36mSwinTransformerBlock.forward_part1\u001b[0;34m(self, x, mask_matrix)\u001b[0m\n\u001b[1;32m    589\u001b[0m     attn_mask \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[1;32m    590\u001b[0m x_windows \u001b[38;5;241m=\u001b[39m window_partition(shifted_x, window_size)\n\u001b[0;32m--> 591\u001b[0m attn_windows \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mattn\u001b[49m\u001b[43m(\u001b[49m\u001b[43mx_windows\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mmask\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mattn_mask\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    592\u001b[0m attn_windows \u001b[38;5;241m=\u001b[39m attn_windows\u001b[38;5;241m.\u001b[39mview(\u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m, \u001b[38;5;241m*\u001b[39m(window_size \u001b[38;5;241m+\u001b[39m (c,)))\n\u001b[1;32m    593\u001b[0m shifted_x \u001b[38;5;241m=\u001b[39m window_reverse(attn_windows, window_size, dims)\n",
-      "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py:1111\u001b[0m, in \u001b[0;36mModule._call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m   1107\u001b[0m \u001b[38;5;66;03m# If we don't have any hooks, we want to skip the rest of the logic in\u001b[39;00m\n\u001b[1;32m   1108\u001b[0m \u001b[38;5;66;03m# this function, and just call forward.\u001b[39;00m\n\u001b[1;32m   1109\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m (\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_backward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_forward_pre_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_backward_hooks\n\u001b[1;32m   1110\u001b[0m         \u001b[38;5;129;01mor\u001b[39;00m _global_forward_hooks \u001b[38;5;129;01mor\u001b[39;00m _global_forward_pre_hooks):\n\u001b[0;32m-> 1111\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mforward_call\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43minput\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1112\u001b[0m \u001b[38;5;66;03m# Do not call functions when jit is used\u001b[39;00m\n\u001b[1;32m   1113\u001b[0m full_backward_hooks, non_full_backward_hooks \u001b[38;5;241m=\u001b[39m [], []\n",
-      "File \u001b[0;32m/opt/monai/monai/networks/nets/swin_unetr.py:475\u001b[0m, in \u001b[0;36mWindowAttention.forward\u001b[0;34m(self, x, mask)\u001b[0m\n\u001b[1;32m    473\u001b[0m q, k, v \u001b[38;5;241m=\u001b[39m qkv[\u001b[38;5;241m0\u001b[39m], qkv[\u001b[38;5;241m1\u001b[39m], qkv[\u001b[38;5;241m2\u001b[39m]\n\u001b[1;32m    474\u001b[0m q \u001b[38;5;241m=\u001b[39m q \u001b[38;5;241m*\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mscale\n\u001b[0;32m--> 475\u001b[0m attn \u001b[38;5;241m=\u001b[39m \u001b[43mq\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m@\u001b[39;49m\u001b[43m \u001b[49m\u001b[43mk\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtranspose\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m-\u001b[39;49m\u001b[38;5;241;43m2\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m-\u001b[39;49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[1;32m    476\u001b[0m relative_position_bias \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mrelative_position_bias_table[\n\u001b[1;32m    477\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mrelative_position_index\u001b[38;5;241m.\u001b[39mclone()[:n, :n]\u001b[38;5;241m.\u001b[39mreshape(\u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m)\n\u001b[1;32m    478\u001b[0m ]\u001b[38;5;241m.\u001b[39mreshape(n, n, \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m)\n\u001b[1;32m    479\u001b[0m relative_position_bias \u001b[38;5;241m=\u001b[39m relative_position_bias\u001b[38;5;241m.\u001b[39mpermute(\u001b[38;5;241m2\u001b[39m, \u001b[38;5;241m0\u001b[39m, \u001b[38;5;241m1\u001b[39m)\u001b[38;5;241m.\u001b[39mcontiguous()\n",
-      "\u001b[0;31mRuntimeError\u001b[0m: CUDA out of memory. Tried to allocate 1.80 GiB (GPU 1; 15.75 GiB total capacity; 12.70 GiB already allocated; 1.45 GiB free; 13.13 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "def validation(epoch_iterator_val):\n",
     "    model.eval()\n",
     "    with torch.no_grad():\n",
     "        for step, batch in enumerate(epoch_iterator_val):\n",
     "            val_inputs, val_labels = (batch[\"image\"].cuda(device=device), batch[\"label\"].cuda(device=device))\n",
-    "            val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)\n",
+    "            with torch.cuda.amp.autocast():\n",
+    "                val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)\n",
     "            val_labels_list = decollate_batch(val_labels)\n",
     "            val_labels_convert = [\n",
     "                post_label(val_label_tensor) for val_label_tensor in val_labels_list\n",
@@ -658,12 +650,12 @@
     "        train_loader, desc=\"Training (X / X Steps) (loss=X.X)\", dynamic_ncols=True\n",
     "    )\n",
     "\n",
-    "    # TODO: use AMP\n",
     "    for step, batch in enumerate(epoch_iterator):\n",
     "        step += 1\n",
     "        x, y = (batch[\"image\"].cuda(device=device), batch[\"label\"].cuda(device=device))\n",
-    "        logit_map = model(x)\n",
-    "        loss = loss_function(logit_map, y)\n",
+    "        with torch.cuda.amp.autocast():\n",
+    "            logit_map = model(x)\n",
+    "            loss = loss_function(logit_map, y)\n",
     "        loss.backward()\n",
     "        epoch_loss += loss.item()\n",
     "        optimizer.step()\n",
@@ -701,36 +693,101 @@
     "                    )\n",
     "                )\n",
     "        global_step += 1\n",
-    "    return global_step, dice_val_best, global_step_best\n",
-    "\n",
-    "torch.cuda.empty_cache()\n",
-    "max_iterations = 30000\n",
-    "eval_num = 500\n",
+    "    return global_step, dice_val_best, global_step_best"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [],
+   "source": [
     "post_label = AsDiscrete(to_onehot=14)\n",
     "post_pred = AsDiscrete(argmax=True, to_onehot=14)\n",
-    "dice_metric = DiceMetric(include_background=True, reduction=\"mean\", get_not_nans=False)\n",
+    "dice_metric = DiceMetric(include_background=True, reduction=\"mean\", get_not_nans=False)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {
+    "scrolled": true
+   },
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (5 / 30 Steps) (loss=3.70081): 100%|██████████| 6/6 [00:21<00:00,  3.63s/it]\n",
+      "Training (11 / 30 Steps) (loss=3.63030): 100%|██████████| 6/6 [00:06<00:00,  1.00s/it]\n",
+      "Training (17 / 30 Steps) (loss=3.47021): 100%|██████████| 6/6 [00:05<00:00,  1.00it/s]\n",
+      "Training (23 / 30 Steps) (loss=3.55231): 100%|██████████| 6/6 [00:06<00:00,  1.01s/it]\n",
+      "Training (29 / 30 Steps) (loss=3.43838):  83%|████████▎ | 5/6 [00:05<00:00,  1.05it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[ANone of the inputs have requires_grad=True. Gradients will be None\n",
+      "\n",
+      "Validate (24 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]         \u001b[A\n",
+      "Validate (24 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:24,  4.96s/it]\u001b[A\n",
+      "Validate (24 / 10 Steps):  17%|█▋        | 1/6 [00:08<00:24,  4.96s/it]\u001b[A\n",
+      "Validate (24 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:16,  4.02s/it]\u001b[A\n",
+      "Validate (24 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:16,  4.02s/it]\u001b[A\n",
+      "Validate (24 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.33s/it]\u001b[A\n",
+      "Validate (24 / 10 Steps):  50%|█████     | 3/6 [00:16<00:09,  3.33s/it]\u001b[A\n",
+      "Validate (24 / 10 Steps):  67%|██████▋   | 4/6 [00:16<00:08,  4.21s/it]\u001b[A\n",
+      "Validate (24 / 10 Steps):  67%|██████▋   | 4/6 [00:19<00:08,  4.21s/it]\u001b[A\n",
+      "Validate (24 / 10 Steps):  83%|████████▎ | 5/6 [00:19<00:03,  3.83s/it]\u001b[A\n",
+      "Validate (24 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.83s/it]\u001b[A\n",
+      "Validate (24 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.81s/it]\u001b[A\n",
+      "positional arguments and argument \"destination\" are deprecated. nn.Module.state_dict will not accept them in the future. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.\n",
+      "Training (29 / 30 Steps) (loss=3.43838): 100%|██████████| 6/6 [00:29<00:00,  4.96s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Saved ! Current Best Avg. Dice: 0.04291130602359772 Current Avg. Dice: 0.04291130602359772\n",
+      "Total train time: 69.60 seconds\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "torch.cuda.empty_cache()\n",
+    "# max_iterations = 30000\n",
+    "# eval_num = 500\n",
+    "max_iterations = 30\n",
+    "eval_num = 29\n",
     "global_step = 0\n",
     "dice_val_best = 0.0\n",
     "global_step_best = 0\n",
     "epoch_loss_values = []\n",
     "metric_values = []\n",
+    "\n",
+    "begin = time.time()\n",
     "while global_step < max_iterations:\n",
     "    global_step, dice_val_best, global_step_best = train(\n",
     "        global_step, train_loader, dice_val_best, global_step_best\n",
     "    )\n",
-    "model.load_state_dict(torch.load(os.path.join(root_dir, \"best_metric_model.pth\")))"
+    "print(f\"Total train time: {time.time() - begin:.2f} seconds\")\n",
+    "# model.load_state_dict(torch.load(os.path.join(root_dir, \"best_metric_model.pth\")))"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 12,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "train completed, best_metric: 0.8430 at iteration: 29500\n"
+      "train completed, best_metric: 0.0808 at iteration: 59\n"
      ]
     }
    ],
@@ -872,7 +929,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.0"
+   "version": "3.8.13"
   }
  },
  "nbformat": 4,

From 5967dc349cfec74f988b0ffcb4f080599f2098bb Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Tue, 12 Jul 2022 08:00:26 -0700
Subject: [PATCH 03/17] checkpoint, debug and profiling code

---
 .../swin_unetr_btcv_segmentation_3d.ipynb     | 1949 ++++++++++++++++-
 3d_segmentation/swin_unetr_profiling.py       |  333 +++
 2 files changed, 2221 insertions(+), 61 deletions(-)
 create mode 100644 3d_segmentation/swin_unetr_profiling.py

diff --git a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
index 2a13cefd46..240f983ed2 100644
--- a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
+++ b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
@@ -96,7 +96,7 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Sat Jul  9 05:14:00 2022       \n",
+      "Tue Jul 12 02:08:22 2022       \n",
       "+-----------------------------------------------------------------------------+\n",
       "| NVIDIA-SMI 450.119.04   Driver Version: 450.119.04   CUDA Version: 11.6     |\n",
       "|-------------------------------+----------------------+----------------------+\n",
@@ -104,8 +104,8 @@
       "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
       "|                               |                      |               MIG M. |\n",
       "|===============================+======================+======================|\n",
-      "|   0  Tesla V100-SXM2...  On   | 00000000:0A:00.0 Off |                    0 |\n",
-      "| N/A   40C    P0    43W / 163W |      3MiB / 32510MiB |      0%      Default |\n",
+      "|   0  Tesla V100-SXM2...  On   | 00000000:07:00.0 Off |                    0 |\n",
+      "| N/A   35C    P0    43W / 163W |      3MiB / 32510MiB |      0%      Default |\n",
       "|                               |                      |                  N/A |\n",
       "+-------------------------------+----------------------+----------------------+\n",
       "                                                                               \n",
@@ -148,9 +148,9 @@
       "Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
       "Collecting nibabel==3.1.1\n",
       "  Downloading nibabel-3.1.1-py3-none-any.whl (3.3 MB)\n",
-      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.3/3.3 MB\u001b[0m \u001b[31m106.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
-      "\u001b[?25hRequirement already satisfied: numpy>=1.13 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (1.22.3)\n",
-      "Requirement already satisfied: packaging>=14.3 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (21.3)\n",
+      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.3/3.3 MB\u001b[0m \u001b[31m102.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+      "\u001b[?25hRequirement already satisfied: packaging>=14.3 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (21.3)\n",
+      "Requirement already satisfied: numpy>=1.13 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (1.22.3)\n",
       "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.8/site-packages (from packaging>=14.3->nibabel==3.1.1) (3.0.8)\n",
       "Installing collected packages: nibabel\n",
       "  Attempting uninstall: nibabel\n",
@@ -162,7 +162,7 @@
       "\u001b[0mLooking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
       "Collecting tqdm==4.63.0\n",
       "  Downloading tqdm-4.63.0-py2.py3-none-any.whl (76 kB)\n",
-      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m76.6/76.6 kB\u001b[0m \u001b[31m32.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m76.6/76.6 kB\u001b[0m \u001b[31m28.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
       "\u001b[?25hInstalling collected packages: tqdm\n",
       "  Attempting uninstall: tqdm\n",
       "    Found existing installation: tqdm 4.64.0\n",
@@ -186,17 +186,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "/opt/conda/lib/python3.8/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"
-     ]
-    },
     {
      "name": "stdout",
      "output_type": "stream",
@@ -216,7 +208,7 @@
       "Tensorboard version: 2.8.0\n",
       "gdown version: 4.4.0\n",
       "TorchVision version: 0.13.0a0\n",
-      "tqdm version: 4.63.0\n",
+      "tqdm version: 4.64.0\n",
       "lmdb version: 1.3.0\n",
       "psutil version: 5.9.0\n",
       "pandas version: 1.3.5\n",
@@ -260,6 +252,7 @@
     "    EnsureTyped,\n",
     "    ToDeviced\n",
     ")\n",
+    "from monai.utils import set_determinism\n",
     "\n",
     "from monai.config import print_config\n",
     "from monai.metrics import DiceMetric\n",
@@ -292,14 +285,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "/tmp/tmpnv0d_576\n"
+      "/tmp/tmp8tsvk5fk\n"
      ]
     }
    ],
@@ -319,11 +312,11 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [],
    "source": [
-    "num_samples = 1\n",
+    "num_samples = 4\n",
     "os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\"\n",
     "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
     "\n",
@@ -425,15 +418,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "Loading dataset: 100%|██████████| 24/24 [00:55<00:00,  2.29s/it]\n",
-      "Loading dataset: 100%|██████████| 6/6 [00:19<00:00,  3.17s/it]\n"
+      "Loading dataset: 100%|██████████| 24/24 [00:50<00:00,  2.11s/it]\n",
+      "Loading dataset: 100%|██████████| 6/6 [00:18<00:00,  3.10s/it]\n"
      ]
     }
    ],
@@ -454,9 +447,7 @@
     "    cache_rate=1.0,\n",
     "    num_workers=8,\n",
     ")\n",
-    "# train_loader = ThreadDataLoader(train_ds, batch_size=1, shuffle=True, num_workers=0)\n",
-    "train_loader = ThreadDataLoader(train_ds, batch_size=4, shuffle=True, use_thread_workers=True, num_workers=1)\n",
-    "\n",
+    "train_loader = ThreadDataLoader(train_ds, batch_size=1, shuffle=True, use_thread_workers=False, num_workers=0)\n",
     "val_ds = CacheDataset(\n",
     "    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4\n",
     ")\n",
@@ -532,7 +523,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -568,7 +559,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 9,
    "metadata": {},
    "outputs": [
     {
@@ -594,7 +585,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 10,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -612,7 +603,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 11,
    "metadata": {
     "scrolled": true
    },
@@ -698,7 +689,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 12,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -709,7 +700,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 13,
    "metadata": {
     "scrolled": true
    },
@@ -718,35 +709,1832 @@
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "Training (5 / 30 Steps) (loss=3.70081): 100%|██████████| 6/6 [00:21<00:00,  3.63s/it]\n",
-      "Training (11 / 30 Steps) (loss=3.63030): 100%|██████████| 6/6 [00:06<00:00,  1.00s/it]\n",
-      "Training (17 / 30 Steps) (loss=3.47021): 100%|██████████| 6/6 [00:05<00:00,  1.00it/s]\n",
-      "Training (23 / 30 Steps) (loss=3.55231): 100%|██████████| 6/6 [00:06<00:00,  1.01s/it]\n",
-      "Training (29 / 30 Steps) (loss=3.43838):  83%|████████▎ | 5/6 [00:05<00:00,  1.05it/s]\n",
+      "Training (23 / 30000 Steps) (loss=3.04756): 100%|██████████| 24/24 [00:24<00:00,  1.02s/it]\n",
+      "Training (47 / 30000 Steps) (loss=2.90349): 100%|██████████| 24/24 [00:22<00:00,  1.08it/s]\n",
+      "Training (71 / 30000 Steps) (loss=2.80009): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (95 / 30000 Steps) (loss=2.73891): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (119 / 30000 Steps) (loss=2.75600): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (143 / 30000 Steps) (loss=2.81679): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (167 / 30000 Steps) (loss=2.73382): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (191 / 30000 Steps) (loss=2.70652): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (215 / 30000 Steps) (loss=2.58161): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (239 / 30000 Steps) (loss=2.55389): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (263 / 30000 Steps) (loss=2.59991): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (287 / 30000 Steps) (loss=2.57285): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (311 / 30000 Steps) (loss=2.59925): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (335 / 30000 Steps) (loss=2.44729): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (359 / 30000 Steps) (loss=2.47251): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (383 / 30000 Steps) (loss=2.34368): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (407 / 30000 Steps) (loss=2.28056): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (431 / 30000 Steps) (loss=2.38724): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (455 / 30000 Steps) (loss=2.35536): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (479 / 30000 Steps) (loss=2.31131): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (500 / 30000 Steps) (loss=2.27539):  83%|████████▎ | 20/24 [00:19<00:03,  1.08it/s]\n",
       "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[ANone of the inputs have requires_grad=True. Gradients will be None\n",
       "\n",
-      "Validate (24 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]         \u001b[A\n",
-      "Validate (24 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:24,  4.96s/it]\u001b[A\n",
-      "Validate (24 / 10 Steps):  17%|█▋        | 1/6 [00:08<00:24,  4.96s/it]\u001b[A\n",
-      "Validate (24 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:16,  4.02s/it]\u001b[A\n",
-      "Validate (24 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:16,  4.02s/it]\u001b[A\n",
-      "Validate (24 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.33s/it]\u001b[A\n",
-      "Validate (24 / 10 Steps):  50%|█████     | 3/6 [00:16<00:09,  3.33s/it]\u001b[A\n",
-      "Validate (24 / 10 Steps):  67%|██████▋   | 4/6 [00:16<00:08,  4.21s/it]\u001b[A\n",
-      "Validate (24 / 10 Steps):  67%|██████▋   | 4/6 [00:19<00:08,  4.21s/it]\u001b[A\n",
-      "Validate (24 / 10 Steps):  83%|████████▎ | 5/6 [00:19<00:03,  3.83s/it]\u001b[A\n",
-      "Validate (24 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.83s/it]\u001b[A\n",
-      "Validate (24 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.81s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]        \u001b[A\n",
+      "Validate (480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.27s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:13,  3.27s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.97s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.97s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
       "positional arguments and argument \"destination\" are deprecated. nn.Module.state_dict will not accept them in the future. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.\n",
-      "Training (29 / 30 Steps) (loss=3.43838): 100%|██████████| 6/6 [00:29<00:00,  4.96s/it]"
+      "Training (500 / 30000 Steps) (loss=2.27539):  88%|████████▊ | 21/24 [00:41<00:22,  7.52s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.09675732254981995\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (503 / 30000 Steps) (loss=2.31343): 100%|██████████| 24/24 [00:43<00:00,  1.83s/it]\n",
+      "Training (527 / 30000 Steps) (loss=2.29317): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (551 / 30000 Steps) (loss=2.26501): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (575 / 30000 Steps) (loss=2.28498): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (599 / 30000 Steps) (loss=2.19028): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (623 / 30000 Steps) (loss=2.50455): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (647 / 30000 Steps) (loss=2.18401): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (671 / 30000 Steps) (loss=2.24277): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (695 / 30000 Steps) (loss=2.21543): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (719 / 30000 Steps) (loss=2.26871): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (743 / 30000 Steps) (loss=2.23642): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (767 / 30000 Steps) (loss=2.19016): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (791 / 30000 Steps) (loss=2.15235): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (815 / 30000 Steps) (loss=2.14225): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (839 / 30000 Steps) (loss=2.09030): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (863 / 30000 Steps) (loss=2.10239): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (887 / 30000 Steps) (loss=2.16179): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (911 / 30000 Steps) (loss=2.03907): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (935 / 30000 Steps) (loss=2.04785): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (959 / 30000 Steps) (loss=2.10109): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (983 / 30000 Steps) (loss=2.12406): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1000 / 30000 Steps) (loss=2.07229):  67%|██████▋   | 16/24 [00:15<00:07,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]        \u001b[A\n",
+      "Validate (984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (1000 / 30000 Steps) (loss=2.07229):  71%|███████   | 17/24 [00:36<00:50,  7.24s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.09206199645996094\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (1007 / 30000 Steps) (loss=1.96543): 100%|██████████| 24/24 [00:43<00:00,  1.79s/it]\n",
+      "Training (1031 / 30000 Steps) (loss=1.98335): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1055 / 30000 Steps) (loss=1.97615): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1079 / 30000 Steps) (loss=1.95188): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1103 / 30000 Steps) (loss=2.07248): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1127 / 30000 Steps) (loss=2.11488): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1151 / 30000 Steps) (loss=1.84554): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1175 / 30000 Steps) (loss=2.08018): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1199 / 30000 Steps) (loss=1.95251): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1223 / 30000 Steps) (loss=2.01169): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1247 / 30000 Steps) (loss=2.16851): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (1271 / 30000 Steps) (loss=2.17117): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1295 / 30000 Steps) (loss=1.99893): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1319 / 30000 Steps) (loss=2.00061): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1343 / 30000 Steps) (loss=1.94732): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1367 / 30000 Steps) (loss=1.99271): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1391 / 30000 Steps) (loss=2.02488): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1415 / 30000 Steps) (loss=1.95091): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1439 / 30000 Steps) (loss=2.07077): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1463 / 30000 Steps) (loss=1.87043): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1487 / 30000 Steps) (loss=2.07426): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1500 / 30000 Steps) (loss=2.03805):  50%|█████     | 12/24 [00:11<00:10,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (1488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (1488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (1500 / 30000 Steps) (loss=2.03805):  54%|█████▍    | 13/24 [00:32<01:20,  7.29s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059749215841293\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (1511 / 30000 Steps) (loss=1.97013): 100%|██████████| 24/24 [00:43<00:00,  1.79s/it]\n",
+      "Training (1535 / 30000 Steps) (loss=2.08701): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1559 / 30000 Steps) (loss=1.91518): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1583 / 30000 Steps) (loss=1.86013): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (1607 / 30000 Steps) (loss=1.92820): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1631 / 30000 Steps) (loss=1.95914): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1655 / 30000 Steps) (loss=2.31439): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (1679 / 30000 Steps) (loss=1.93297): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1703 / 30000 Steps) (loss=1.82793): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1727 / 30000 Steps) (loss=1.77602): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1751 / 30000 Steps) (loss=1.87049): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1775 / 30000 Steps) (loss=1.81319): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1799 / 30000 Steps) (loss=1.86025): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1823 / 30000 Steps) (loss=1.81045): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1847 / 30000 Steps) (loss=1.76598): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (1871 / 30000 Steps) (loss=1.93609): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1895 / 30000 Steps) (loss=1.85808): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1919 / 30000 Steps) (loss=1.87683): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1943 / 30000 Steps) (loss=1.79905): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (1967 / 30000 Steps) (loss=1.93705): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (1991 / 30000 Steps) (loss=1.87536): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2000 / 30000 Steps) (loss=1.84261):  33%|███▎      | 8/24 [00:08<00:14,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (1992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (1992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (1992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (1992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (1992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (1992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (1992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (1992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (1992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (1992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (1992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (1992 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (2000 / 30000 Steps) (loss=1.84261):  38%|███▊      | 9/24 [00:29<01:52,  7.49s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.08243301510810852\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (2015 / 30000 Steps) (loss=1.85733): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (2039 / 30000 Steps) (loss=1.97582): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (2063 / 30000 Steps) (loss=1.80809): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2087 / 30000 Steps) (loss=1.76823): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2111 / 30000 Steps) (loss=1.85991): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2135 / 30000 Steps) (loss=2.01875): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2159 / 30000 Steps) (loss=1.75839): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2183 / 30000 Steps) (loss=1.74449): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2207 / 30000 Steps) (loss=1.76575): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2231 / 30000 Steps) (loss=1.82603): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2255 / 30000 Steps) (loss=1.89225): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2279 / 30000 Steps) (loss=1.76695): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2303 / 30000 Steps) (loss=1.88593): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2327 / 30000 Steps) (loss=1.90981): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2351 / 30000 Steps) (loss=1.91762): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2375 / 30000 Steps) (loss=2.09532): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2399 / 30000 Steps) (loss=2.09043): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2423 / 30000 Steps) (loss=1.89411): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2447 / 30000 Steps) (loss=1.64777): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (2471 / 30000 Steps) (loss=1.70874): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2495 / 30000 Steps) (loss=1.97091): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2500 / 30000 Steps) (loss=1.70573):  17%|█▋        | 4/24 [00:04<00:18,  1.08it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (2496 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (2496 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.17s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.17s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.27s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:13,  3.27s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.97s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.97s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (2500 / 30000 Steps) (loss=1.70573):  21%|██        | 5/24 [00:25<02:41,  8.52s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07071128487586975\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (2519 / 30000 Steps) (loss=2.08651): 100%|██████████| 24/24 [00:43<00:00,  1.79s/it]\n",
+      "Training (2543 / 30000 Steps) (loss=1.68501): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2567 / 30000 Steps) (loss=1.68676): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2591 / 30000 Steps) (loss=1.98606): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2615 / 30000 Steps) (loss=1.88070): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (2639 / 30000 Steps) (loss=1.85338): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2663 / 30000 Steps) (loss=1.75137): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2687 / 30000 Steps) (loss=1.68586): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2711 / 30000 Steps) (loss=1.85745): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2735 / 30000 Steps) (loss=1.83789): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2759 / 30000 Steps) (loss=2.14781): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2783 / 30000 Steps) (loss=1.65742): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2807 / 30000 Steps) (loss=1.61873): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2831 / 30000 Steps) (loss=1.69797): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2855 / 30000 Steps) (loss=1.73362): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2879 / 30000 Steps) (loss=1.64782): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2903 / 30000 Steps) (loss=1.96968): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2927 / 30000 Steps) (loss=1.81555): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2951 / 30000 Steps) (loss=1.90872): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (2975 / 30000 Steps) (loss=1.74729): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (2999 / 30000 Steps) (loss=1.71663): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3000 / 30000 Steps) (loss=1.77936):   0%|          | 0/24 [00:00<?, ?it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (3000 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (3000 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (3000 / 30000 Steps) (loss=1.77936):   4%|▍         | 1/24 [00:21<08:25, 21.97s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07060686498880386\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (3023 / 30000 Steps) (loss=1.69800): 100%|██████████| 24/24 [00:43<00:00,  1.79s/it]\n",
+      "Training (3047 / 30000 Steps) (loss=1.95848): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3071 / 30000 Steps) (loss=1.87359): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3095 / 30000 Steps) (loss=1.77414): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3119 / 30000 Steps) (loss=1.63642): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3143 / 30000 Steps) (loss=1.91925): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3167 / 30000 Steps) (loss=1.59526): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (3191 / 30000 Steps) (loss=1.67294): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3215 / 30000 Steps) (loss=1.83990): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3239 / 30000 Steps) (loss=1.98160): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3263 / 30000 Steps) (loss=1.66625): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3287 / 30000 Steps) (loss=1.59335): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3311 / 30000 Steps) (loss=1.71648): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3335 / 30000 Steps) (loss=1.88376): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3359 / 30000 Steps) (loss=1.71496): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3383 / 30000 Steps) (loss=1.86785): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3407 / 30000 Steps) (loss=1.98294): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (3431 / 30000 Steps) (loss=1.75702): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (3455 / 30000 Steps) (loss=1.79603): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3479 / 30000 Steps) (loss=1.63082): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3500 / 30000 Steps) (loss=1.85942):  83%|████████▎ | 20/24 [00:19<00:03,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (3480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (3480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (3480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (3480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (3480 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (3480 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (3480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (3480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (3480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (3480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (3480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (3480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (3500 / 30000 Steps) (loss=1.85942):  88%|████████▊ | 21/24 [00:40<00:21,  7.21s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.0705963596701622\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (3503 / 30000 Steps) (loss=2.01710): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (3527 / 30000 Steps) (loss=1.84506): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3551 / 30000 Steps) (loss=1.62561): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3575 / 30000 Steps) (loss=1.63863): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3599 / 30000 Steps) (loss=1.87775): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3623 / 30000 Steps) (loss=1.74231): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3647 / 30000 Steps) (loss=1.68101): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3671 / 30000 Steps) (loss=1.91874): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3695 / 30000 Steps) (loss=1.53870): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3719 / 30000 Steps) (loss=1.74763): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3743 / 30000 Steps) (loss=1.55972): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (3767 / 30000 Steps) (loss=1.74384): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3791 / 30000 Steps) (loss=1.85877): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3815 / 30000 Steps) (loss=1.76684): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3839 / 30000 Steps) (loss=1.61158): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3863 / 30000 Steps) (loss=1.95114): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3887 / 30000 Steps) (loss=1.91905): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (3911 / 30000 Steps) (loss=1.72093): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (3935 / 30000 Steps) (loss=1.68328): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (3959 / 30000 Steps) (loss=2.00625): 100%|██████████| 24/24 [00:22<00:00,  1.08it/s]\n",
+      "Training (4055 / 30000 Steps) (loss=3.08001): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4079 / 30000 Steps) (loss=1.88033): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4103 / 30000 Steps) (loss=1.88240): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4127 / 30000 Steps) (loss=2.10009): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4151 / 30000 Steps) (loss=1.95376): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4175 / 30000 Steps) (loss=1.64926): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4199 / 30000 Steps) (loss=1.67962): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4223 / 30000 Steps) (loss=1.56884): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4247 / 30000 Steps) (loss=1.79777): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4271 / 30000 Steps) (loss=1.81146): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4295 / 30000 Steps) (loss=1.69637): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4319 / 30000 Steps) (loss=1.96981): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4343 / 30000 Steps) (loss=1.98609): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4367 / 30000 Steps) (loss=1.74551): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4391 / 30000 Steps) (loss=1.86034): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4415 / 30000 Steps) (loss=1.67974): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (4439 / 30000 Steps) (loss=1.91140): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4463 / 30000 Steps) (loss=1.49962): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4487 / 30000 Steps) (loss=1.69141): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4500 / 30000 Steps) (loss=1.59464):  50%|█████     | 12/24 [00:11<00:11,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (4488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (4488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (4488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (4488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.25s/it]\u001b[A\n",
+      "Validate (4488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.25s/it]\u001b[A\n",
+      "Validate (4488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.90s/it]\u001b[A\n",
+      "Validate (4488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.90s/it]\u001b[A\n",
+      "Validate (4488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (4488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (4488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (4488 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (4488 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
+      "Training (4500 / 30000 Steps) (loss=1.59464):  54%|█████▍    | 13/24 [00:32<01:20,  7.27s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07118804007768631\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (4511 / 30000 Steps) (loss=1.71029): 100%|██████████| 24/24 [00:43<00:00,  1.79s/it]\n",
+      "Training (4535 / 30000 Steps) (loss=1.62508): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4559 / 30000 Steps) (loss=2.17477): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4583 / 30000 Steps) (loss=1.88658): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4607 / 30000 Steps) (loss=1.80879): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4631 / 30000 Steps) (loss=1.79553): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4655 / 30000 Steps) (loss=1.68824): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4679 / 30000 Steps) (loss=3.31505): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4703 / 30000 Steps) (loss=1.69691): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4727 / 30000 Steps) (loss=1.76048): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4751 / 30000 Steps) (loss=1.83267): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4775 / 30000 Steps) (loss=1.76414): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4799 / 30000 Steps) (loss=2.53666): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4823 / 30000 Steps) (loss=2.14007): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4847 / 30000 Steps) (loss=1.74145): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4871 / 30000 Steps) (loss=2.61676): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4895 / 30000 Steps) (loss=1.81788): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4919 / 30000 Steps) (loss=1.57590): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4943 / 30000 Steps) (loss=2.00811): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (4967 / 30000 Steps) (loss=1.80384): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (4991 / 30000 Steps) (loss=1.90285): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (5000 / 30000 Steps) (loss=1.88040):  33%|███▎      | 8/24 [00:08<00:14,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (4992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (4992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.24s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.24s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.90s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.90s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.94s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.94s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.65s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.65s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.49s/it]\u001b[A\n",
+      "Training (5000 / 30000 Steps) (loss=1.88040):  38%|███▊      | 9/24 [00:29<01:51,  7.46s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.0705961063504219\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (5015 / 30000 Steps) (loss=1.90351): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (5039 / 30000 Steps) (loss=1.82040): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (5063 / 30000 Steps) (loss=1.98175): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (5087 / 30000 Steps) (loss=1.83518): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (5111 / 30000 Steps) (loss=1.94957): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (5135 / 30000 Steps) (loss=2.21675): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (5159 / 30000 Steps) (loss=1.65691): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (5183 / 30000 Steps) (loss=2.06594): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (5207 / 30000 Steps) (loss=1.71019): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (5231 / 30000 Steps) (loss=1.69909): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (5255 / 30000 Steps) (loss=1.78258): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (5279 / 30000 Steps) (loss=1.83680): 100%|██████████| 24/24 [00:22<00:00,  1.08it/s]\n",
+      "Training (5303 / 30000 Steps) (loss=1.72917): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (5327 / 30000 Steps) (loss=1.78574): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (5351 / 30000 Steps) (loss=1.66677): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (5375 / 30000 Steps) (loss=1.91206): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (5399 / 30000 Steps) (loss=1.62288): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (5423 / 30000 Steps) (loss=1.64085): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (5425 / 30000 Steps) (loss=2.11989):   8%|▊         | 2/24 [00:01<00:20,  1.08it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (6719 / 30000 Steps) (loss=2.10010): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6743 / 30000 Steps) (loss=2.14659): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6767 / 30000 Steps) (loss=2.25052): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6791 / 30000 Steps) (loss=2.32982): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6815 / 30000 Steps) (loss=2.88283): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6839 / 30000 Steps) (loss=2.16736): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6863 / 30000 Steps) (loss=2.30831): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6887 / 30000 Steps) (loss=2.07512): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6911 / 30000 Steps) (loss=2.20173): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6935 / 30000 Steps) (loss=2.10239): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6959 / 30000 Steps) (loss=2.10825): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (6983 / 30000 Steps) (loss=2.14991): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7000 / 30000 Steps) (loss=2.20176):  67%|██████▋   | 16/24 [00:15<00:07,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (6984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (6984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (6984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (6984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (6984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (6984 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (6984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (6984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (6984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (6984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (6984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (6984 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (7000 / 30000 Steps) (loss=2.20176):  71%|███████   | 17/24 [00:36<00:50,  7.23s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (7007 / 30000 Steps) (loss=2.22435): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (7031 / 30000 Steps) (loss=1.94915): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (7055 / 30000 Steps) (loss=2.10985): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7079 / 30000 Steps) (loss=3.61836): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7103 / 30000 Steps) (loss=2.17272): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7127 / 30000 Steps) (loss=2.11024): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7151 / 30000 Steps) (loss=2.18739): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7175 / 30000 Steps) (loss=1.98527): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7199 / 30000 Steps) (loss=2.06582): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7223 / 30000 Steps) (loss=1.85681): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7247 / 30000 Steps) (loss=2.20366): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7271 / 30000 Steps) (loss=2.27249): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7295 / 30000 Steps) (loss=2.02298): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7319 / 30000 Steps) (loss=2.12844): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7343 / 30000 Steps) (loss=2.19865): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7367 / 30000 Steps) (loss=1.98583): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7391 / 30000 Steps) (loss=1.90731): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7415 / 30000 Steps) (loss=2.00232): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7439 / 30000 Steps) (loss=2.11610): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7463 / 30000 Steps) (loss=2.11672): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7487 / 30000 Steps) (loss=1.97471): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7500 / 30000 Steps) (loss=2.09074):  50%|█████     | 12/24 [00:11<00:11,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (7488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (7488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (7500 / 30000 Steps) (loss=2.09074):  54%|█████▍    | 13/24 [00:32<01:20,  7.29s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (7511 / 30000 Steps) (loss=1.93107): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (7535 / 30000 Steps) (loss=2.07361): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7559 / 30000 Steps) (loss=1.73974): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (7583 / 30000 Steps) (loss=1.86634): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7607 / 30000 Steps) (loss=1.71773): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7631 / 30000 Steps) (loss=2.12695): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7655 / 30000 Steps) (loss=2.08464): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7679 / 30000 Steps) (loss=1.89307): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7703 / 30000 Steps) (loss=2.23980): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7727 / 30000 Steps) (loss=2.06015): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7751 / 30000 Steps) (loss=1.92099): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7775 / 30000 Steps) (loss=1.92985): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7799 / 30000 Steps) (loss=1.91172): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7823 / 30000 Steps) (loss=1.72619): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7847 / 30000 Steps) (loss=1.83092): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (7871 / 30000 Steps) (loss=1.85543): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7895 / 30000 Steps) (loss=2.39690): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7919 / 30000 Steps) (loss=2.06412): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7943 / 30000 Steps) (loss=2.02304): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7967 / 30000 Steps) (loss=1.82633): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (7991 / 30000 Steps) (loss=1.97773): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (8000 / 30000 Steps) (loss=1.75278):  33%|███▎      | 8/24 [00:08<00:14,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (7992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (7992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (8000 / 30000 Steps) (loss=1.75278):  38%|███▊      | 9/24 [00:29<01:52,  7.48s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (8015 / 30000 Steps) (loss=1.83981): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (8039 / 30000 Steps) (loss=1.89129): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (8061 / 30000 Steps) (loss=1.97249):  92%|█████████▏| 22/24 [00:20<00:01,  1.10it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (9191 / 30000 Steps) (loss=1.82559): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9215 / 30000 Steps) (loss=2.22664): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9239 / 30000 Steps) (loss=3.52945): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9263 / 30000 Steps) (loss=1.69937): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9287 / 30000 Steps) (loss=1.96457): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9311 / 30000 Steps) (loss=2.37476): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9335 / 30000 Steps) (loss=3.65335): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9359 / 30000 Steps) (loss=2.01223): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9383 / 30000 Steps) (loss=1.98225): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9407 / 30000 Steps) (loss=1.96237): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9431 / 30000 Steps) (loss=2.09754): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9455 / 30000 Steps) (loss=2.20548): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9479 / 30000 Steps) (loss=1.84057): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9500 / 30000 Steps) (loss=1.92208):  83%|████████▎ | 20/24 [00:19<00:03,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (9480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (9480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (9480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (9480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.25s/it]\u001b[A\n",
+      "Validate (9480 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.25s/it]\u001b[A\n",
+      "Validate (9480 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (9480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (9480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (9480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (9480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (9480 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (9480 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
+      "Training (9500 / 30000 Steps) (loss=1.92208):  88%|████████▊ | 21/24 [00:40<00:21,  7.21s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (9503 / 30000 Steps) (loss=2.06457): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (9527 / 30000 Steps) (loss=1.92062): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9551 / 30000 Steps) (loss=1.92944): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (9575 / 30000 Steps) (loss=1.87242): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (9599 / 30000 Steps) (loss=2.29003): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9623 / 30000 Steps) (loss=2.05942): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9647 / 30000 Steps) (loss=2.01496): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9671 / 30000 Steps) (loss=2.16334): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9695 / 30000 Steps) (loss=2.43831): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9719 / 30000 Steps) (loss=1.83332): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9743 / 30000 Steps) (loss=1.79377): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9767 / 30000 Steps) (loss=1.83020): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9791 / 30000 Steps) (loss=1.81070): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (9815 / 30000 Steps) (loss=1.98005): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9839 / 30000 Steps) (loss=1.87129): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9863 / 30000 Steps) (loss=1.97455): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9887 / 30000 Steps) (loss=1.85374): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9911 / 30000 Steps) (loss=2.03599): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9935 / 30000 Steps) (loss=1.99218): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9959 / 30000 Steps) (loss=1.84849): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (9983 / 30000 Steps) (loss=4.62641): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10000 / 30000 Steps) (loss=2.10157):  67%|██████▋   | 16/24 [00:15<00:07,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (9984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
+      "Validate (9984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (10000 / 30000 Steps) (loss=2.10157):  71%|███████   | 17/24 [00:36<00:50,  7.23s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059746980667114\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (10007 / 30000 Steps) (loss=2.04821): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (10031 / 30000 Steps) (loss=2.15589): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10055 / 30000 Steps) (loss=2.04477): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10079 / 30000 Steps) (loss=2.52525): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10103 / 30000 Steps) (loss=1.87341): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (10127 / 30000 Steps) (loss=2.04750): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10151 / 30000 Steps) (loss=1.67941): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10175 / 30000 Steps) (loss=2.09627): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10199 / 30000 Steps) (loss=2.10325): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10223 / 30000 Steps) (loss=1.72772): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10247 / 30000 Steps) (loss=1.73709): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10271 / 30000 Steps) (loss=2.13825): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10295 / 30000 Steps) (loss=2.38076): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (10319 / 30000 Steps) (loss=1.99312): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10343 / 30000 Steps) (loss=1.91718): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10367 / 30000 Steps) (loss=1.73751): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10391 / 30000 Steps) (loss=2.01234): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10415 / 30000 Steps) (loss=2.13904): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (10439 / 30000 Steps) (loss=1.90831): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10463 / 30000 Steps) (loss=1.90202): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10487 / 30000 Steps) (loss=1.94877): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (10500 / 30000 Steps) (loss=2.07265):  50%|█████     | 12/24 [00:11<00:10,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (10488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (10488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
+      "Training (10500 / 30000 Steps) (loss=2.07265):  54%|█████▍    | 13/24 [00:32<01:20,  7.29s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059737294912338\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (10511 / 30000 Steps) (loss=1.78768): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (10534 / 30000 Steps) (loss=1.97501):  96%|█████████▌| 23/24 [00:21<00:00,  1.10it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (11807 / 30000 Steps) (loss=1.89524): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (11831 / 30000 Steps) (loss=1.74966): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (11855 / 30000 Steps) (loss=2.43224): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (11879 / 30000 Steps) (loss=1.67055): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (11903 / 30000 Steps) (loss=1.77441): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (11927 / 30000 Steps) (loss=2.03059): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (11951 / 30000 Steps) (loss=1.83695): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (11975 / 30000 Steps) (loss=1.91604): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (11999 / 30000 Steps) (loss=2.13925): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12000 / 30000 Steps) (loss=3.23982):   0%|          | 0/24 [00:00<?, ?it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (12000 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (12000 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (12000 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (12000 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (12000 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (12000 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (12000 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (12000 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (12000 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (12000 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (12000 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (12000 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (12000 / 30000 Steps) (loss=3.23982):   4%|▍         | 1/24 [00:22<08:26, 22.00s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (12023 / 30000 Steps) (loss=1.82447): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (12047 / 30000 Steps) (loss=1.82887): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12071 / 30000 Steps) (loss=1.52439): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12095 / 30000 Steps) (loss=1.87138): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12119 / 30000 Steps) (loss=1.97862): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12143 / 30000 Steps) (loss=1.74948): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12167 / 30000 Steps) (loss=1.78102): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12191 / 30000 Steps) (loss=1.85116): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12215 / 30000 Steps) (loss=1.82716): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12239 / 30000 Steps) (loss=1.71390): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12263 / 30000 Steps) (loss=1.86355): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (12287 / 30000 Steps) (loss=1.94163): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12311 / 30000 Steps) (loss=1.95592): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12335 / 30000 Steps) (loss=1.90674): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12359 / 30000 Steps) (loss=1.51387): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12383 / 30000 Steps) (loss=1.79401): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12407 / 30000 Steps) (loss=2.12846): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12431 / 30000 Steps) (loss=1.88363): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12455 / 30000 Steps) (loss=1.55907): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12479 / 30000 Steps) (loss=1.92450): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12500 / 30000 Steps) (loss=1.80316):  83%|████████▎ | 20/24 [00:19<00:03,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (12480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (12480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (12500 / 30000 Steps) (loss=1.80316):  88%|████████▊ | 21/24 [00:40<00:21,  7.24s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (12503 / 30000 Steps) (loss=1.96900): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (12527 / 30000 Steps) (loss=1.71932): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12551 / 30000 Steps) (loss=2.44105): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12575 / 30000 Steps) (loss=1.82854): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12599 / 30000 Steps) (loss=1.87003): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12623 / 30000 Steps) (loss=1.98144): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12647 / 30000 Steps) (loss=1.74120): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (12671 / 30000 Steps) (loss=1.82985): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12695 / 30000 Steps) (loss=3.63615): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12719 / 30000 Steps) (loss=1.61349): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12743 / 30000 Steps) (loss=1.66895): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12767 / 30000 Steps) (loss=2.05730): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12791 / 30000 Steps) (loss=3.24769): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12815 / 30000 Steps) (loss=1.84073): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12839 / 30000 Steps) (loss=1.83791): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12863 / 30000 Steps) (loss=1.63371): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12887 / 30000 Steps) (loss=1.97341): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (12911 / 30000 Steps) (loss=1.69248): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12935 / 30000 Steps) (loss=1.78743): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (12959 / 30000 Steps) (loss=1.92369): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (12983 / 30000 Steps) (loss=1.73248): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (13000 / 30000 Steps) (loss=1.85713):  67%|██████▋   | 16/24 [00:15<00:07,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (12984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (12984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
+      "Training (13000 / 30000 Steps) (loss=1.85713):  71%|███████   | 17/24 [00:36<00:50,  7.22s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (13007 / 30000 Steps) (loss=1.91395): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (13031 / 30000 Steps) (loss=1.99260): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (13055 / 30000 Steps) (loss=1.97792): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (13079 / 30000 Steps) (loss=1.78542): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (13103 / 30000 Steps) (loss=1.85659): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (13127 / 30000 Steps) (loss=1.92632): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (13151 / 30000 Steps) (loss=1.81754): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (13152 / 30000 Steps) (loss=1.89086):   0%|          | 0/24 [00:00<?, ?it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (14279 / 30000 Steps) (loss=1.97302): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14303 / 30000 Steps) (loss=1.76956): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14327 / 30000 Steps) (loss=1.62204): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14351 / 30000 Steps) (loss=2.00075): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14375 / 30000 Steps) (loss=2.04387): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14399 / 30000 Steps) (loss=1.80620): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14423 / 30000 Steps) (loss=1.75732): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14447 / 30000 Steps) (loss=1.97194): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14471 / 30000 Steps) (loss=1.96930): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14495 / 30000 Steps) (loss=1.86121): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14500 / 30000 Steps) (loss=2.08860):  17%|█▋        | 4/24 [00:04<00:18,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (14496 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (14496 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (14496 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (14496 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (14496 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (14496 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (14496 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (14496 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (14496 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (14496 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (14496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (14496 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (14500 / 30000 Steps) (loss=2.08860):  21%|██        | 5/24 [00:25<02:41,  8.50s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07063417136669159\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (14519 / 30000 Steps) (loss=1.91196): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (14543 / 30000 Steps) (loss=2.08129): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14567 / 30000 Steps) (loss=2.49496): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14591 / 30000 Steps) (loss=2.31298): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14615 / 30000 Steps) (loss=1.64771): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14639 / 30000 Steps) (loss=2.52957): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (14663 / 30000 Steps) (loss=1.84937): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14687 / 30000 Steps) (loss=1.85760): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14711 / 30000 Steps) (loss=1.72808): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14735 / 30000 Steps) (loss=2.03447): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14759 / 30000 Steps) (loss=1.89378): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14783 / 30000 Steps) (loss=2.67441): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14807 / 30000 Steps) (loss=1.97396): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14831 / 30000 Steps) (loss=1.78090): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14855 / 30000 Steps) (loss=3.55123): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14879 / 30000 Steps) (loss=1.85744): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14903 / 30000 Steps) (loss=2.00595): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14927 / 30000 Steps) (loss=1.93430): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14951 / 30000 Steps) (loss=2.02877): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14975 / 30000 Steps) (loss=1.90165): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (14999 / 30000 Steps) (loss=2.20059): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15000 / 30000 Steps) (loss=1.80541):   0%|          | 0/24 [00:01<?, ?it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (15000 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (15000 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (15000 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (15000 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (15000 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (15000 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (15000 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (15000 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (15000 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (15000 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (15000 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (15000 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (15000 / 30000 Steps) (loss=1.80541):   4%|▍         | 1/24 [00:22<08:26, 22.04s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07060018181800842\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (15023 / 30000 Steps) (loss=1.99871): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (15047 / 30000 Steps) (loss=1.96666): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15071 / 30000 Steps) (loss=1.94317): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15095 / 30000 Steps) (loss=1.76293): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15119 / 30000 Steps) (loss=1.61266): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15143 / 30000 Steps) (loss=1.75006): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15167 / 30000 Steps) (loss=1.99979): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15191 / 30000 Steps) (loss=1.85684): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15215 / 30000 Steps) (loss=1.82103): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15239 / 30000 Steps) (loss=1.78026): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15263 / 30000 Steps) (loss=1.94322): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15287 / 30000 Steps) (loss=1.73688): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15311 / 30000 Steps) (loss=1.90172): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15335 / 30000 Steps) (loss=1.83768): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15359 / 30000 Steps) (loss=2.09235): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15383 / 30000 Steps) (loss=2.08885): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15407 / 30000 Steps) (loss=1.95160): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15431 / 30000 Steps) (loss=1.46536): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15455 / 30000 Steps) (loss=2.30151): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15479 / 30000 Steps) (loss=1.89452): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15500 / 30000 Steps) (loss=1.79847):  83%|████████▎ | 20/24 [00:19<00:03,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (15480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (15480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (15500 / 30000 Steps) (loss=1.79847):  88%|████████▊ | 21/24 [00:40<00:21,  7.23s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059748470783234\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (15503 / 30000 Steps) (loss=2.06165): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (15527 / 30000 Steps) (loss=1.82561): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15551 / 30000 Steps) (loss=1.95241): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15575 / 30000 Steps) (loss=1.67307): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15599 / 30000 Steps) (loss=2.27193): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (15614 / 30000 Steps) (loss=1.92388):  58%|█████▊    | 14/24 [00:13<00:09,  1.09it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (16871 / 30000 Steps) (loss=2.05397): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (16895 / 30000 Steps) (loss=1.79339): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (16919 / 30000 Steps) (loss=1.74023): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (16943 / 30000 Steps) (loss=1.65611): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (16967 / 30000 Steps) (loss=2.20335): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (16991 / 30000 Steps) (loss=1.91431): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (17000 / 30000 Steps) (loss=1.93411):  33%|███▎      | 8/24 [00:08<00:14,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (16992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (16992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (17000 / 30000 Steps) (loss=1.93411):  38%|███▊      | 9/24 [00:29<01:52,  7.48s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (17015 / 30000 Steps) (loss=1.51199): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (17039 / 30000 Steps) (loss=1.86803): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17063 / 30000 Steps) (loss=2.14775): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (17087 / 30000 Steps) (loss=1.87431): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17111 / 30000 Steps) (loss=2.15638): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17135 / 30000 Steps) (loss=1.84681): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17159 / 30000 Steps) (loss=1.77586): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17183 / 30000 Steps) (loss=2.58149): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17207 / 30000 Steps) (loss=2.62441): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17231 / 30000 Steps) (loss=2.00080): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (17255 / 30000 Steps) (loss=1.97522): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17279 / 30000 Steps) (loss=1.94671): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17303 / 30000 Steps) (loss=2.10150): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17327 / 30000 Steps) (loss=1.86351): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17351 / 30000 Steps) (loss=1.76649): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17375 / 30000 Steps) (loss=1.85833): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17399 / 30000 Steps) (loss=1.76396): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17423 / 30000 Steps) (loss=1.83632): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17447 / 30000 Steps) (loss=2.37074): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17471 / 30000 Steps) (loss=1.94557): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (17495 / 30000 Steps) (loss=1.82334): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17500 / 30000 Steps) (loss=2.04853):  17%|█▋        | 4/24 [00:04<00:18,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (17496 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (17496 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (17500 / 30000 Steps) (loss=2.04853):  21%|██        | 5/24 [00:25<02:41,  8.48s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (17519 / 30000 Steps) (loss=1.97825): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (17543 / 30000 Steps) (loss=1.77686): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17567 / 30000 Steps) (loss=2.08660): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17591 / 30000 Steps) (loss=1.69114): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17615 / 30000 Steps) (loss=3.15618): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17639 / 30000 Steps) (loss=2.46837): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17663 / 30000 Steps) (loss=1.90801): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17687 / 30000 Steps) (loss=1.86398): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17711 / 30000 Steps) (loss=2.52838): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17735 / 30000 Steps) (loss=1.85488): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17759 / 30000 Steps) (loss=1.79358): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17783 / 30000 Steps) (loss=1.85657): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17807 / 30000 Steps) (loss=1.57969): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (17831 / 30000 Steps) (loss=1.76049): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17855 / 30000 Steps) (loss=2.58910): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17879 / 30000 Steps) (loss=1.50078): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17903 / 30000 Steps) (loss=2.21724): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17927 / 30000 Steps) (loss=1.79426): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (17951 / 30000 Steps) (loss=1.57439): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17975 / 30000 Steps) (loss=3.44161): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (17999 / 30000 Steps) (loss=1.85103): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (18000 / 30000 Steps) (loss=2.06568):   0%|          | 0/24 [00:00<?, ?it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (18000 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (18000 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (18000 / 30000 Steps) (loss=2.06568):   4%|▍         | 1/24 [00:21<08:24, 21.94s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059751451015472\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (18023 / 30000 Steps) (loss=1.91046): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (18047 / 30000 Steps) (loss=1.77376): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (18071 / 30000 Steps) (loss=1.74509): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (18095 / 30000 Steps) (loss=1.92515): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (18119 / 30000 Steps) (loss=2.01573): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (18143 / 30000 Steps) (loss=1.78100): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (18167 / 30000 Steps) (loss=2.65431): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (18191 / 30000 Steps) (loss=3.30308): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (18215 / 30000 Steps) (loss=2.24336): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (18221 / 30000 Steps) (loss=2.80043):  25%|██▌       | 6/24 [00:05<00:16,  1.09it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (19487 / 30000 Steps) (loss=1.84788): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19500 / 30000 Steps) (loss=1.99503):  50%|█████     | 12/24 [00:11<00:10,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (19488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (19488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.25s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.25s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (19500 / 30000 Steps) (loss=1.99503):  54%|█████▍    | 13/24 [00:32<01:19,  7.27s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059739530086517\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (19511 / 30000 Steps) (loss=2.15546): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (19535 / 30000 Steps) (loss=2.16265): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19559 / 30000 Steps) (loss=2.11639): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19583 / 30000 Steps) (loss=2.04579): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19607 / 30000 Steps) (loss=1.67668): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19631 / 30000 Steps) (loss=2.25905): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19655 / 30000 Steps) (loss=1.94317): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19679 / 30000 Steps) (loss=1.69574): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19703 / 30000 Steps) (loss=1.79305): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19727 / 30000 Steps) (loss=1.96530): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (19751 / 30000 Steps) (loss=2.17987): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (19775 / 30000 Steps) (loss=2.07103): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (19799 / 30000 Steps) (loss=1.85445): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19823 / 30000 Steps) (loss=2.28405): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19847 / 30000 Steps) (loss=2.37703): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
+      "Training (19871 / 30000 Steps) (loss=1.48384): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19895 / 30000 Steps) (loss=1.98135): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19919 / 30000 Steps) (loss=1.64713): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (19943 / 30000 Steps) (loss=2.05469): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19967 / 30000 Steps) (loss=1.93943): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (19991 / 30000 Steps) (loss=1.64823): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20000 / 30000 Steps) (loss=1.65539):  33%|███▎      | 8/24 [00:08<00:14,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (19992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (19992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
+      "Training (20000 / 30000 Steps) (loss=1.65539):  38%|███▊      | 9/24 [00:29<01:52,  7.48s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (20015 / 30000 Steps) (loss=1.60423): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (20039 / 30000 Steps) (loss=1.54916): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20063 / 30000 Steps) (loss=2.57518): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20087 / 30000 Steps) (loss=1.74062): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20111 / 30000 Steps) (loss=1.84239): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (20135 / 30000 Steps) (loss=2.03068): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20159 / 30000 Steps) (loss=1.89472): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20183 / 30000 Steps) (loss=1.70734): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20207 / 30000 Steps) (loss=2.13889): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20231 / 30000 Steps) (loss=1.86293): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20255 / 30000 Steps) (loss=2.21442): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20279 / 30000 Steps) (loss=1.67967): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20303 / 30000 Steps) (loss=1.61414): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20327 / 30000 Steps) (loss=1.90312): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20351 / 30000 Steps) (loss=2.74694): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20375 / 30000 Steps) (loss=1.65601): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (20399 / 30000 Steps) (loss=1.87255): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20423 / 30000 Steps) (loss=1.80009): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20447 / 30000 Steps) (loss=2.14480): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20471 / 30000 Steps) (loss=1.98892): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20495 / 30000 Steps) (loss=1.86035): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20500 / 30000 Steps) (loss=1.84063):  17%|█▋        | 4/24 [00:04<00:18,  1.08it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (20496 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (20496 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
+      "Training (20500 / 30000 Steps) (loss=1.84063):  21%|██        | 5/24 [00:25<02:41,  8.49s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (20519 / 30000 Steps) (loss=2.33358): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
+      "Training (20543 / 30000 Steps) (loss=2.00746): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20567 / 30000 Steps) (loss=1.80294): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20591 / 30000 Steps) (loss=1.91813): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20615 / 30000 Steps) (loss=1.82094): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20639 / 30000 Steps) (loss=2.03983): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20663 / 30000 Steps) (loss=1.80291): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (20687 / 30000 Steps) (loss=2.07999): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (20711 / 30000 Steps) (loss=1.92146): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20735 / 30000 Steps) (loss=1.62822): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (20759 / 30000 Steps) (loss=2.19601): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20783 / 30000 Steps) (loss=1.59426): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (20807 / 30000 Steps) (loss=1.82663): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (20831 / 30000 Steps) (loss=1.95577): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (22079 / 30000 Steps) (loss=1.67489): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22103 / 30000 Steps) (loss=1.73300): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22127 / 30000 Steps) (loss=1.97787): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22151 / 30000 Steps) (loss=1.66715): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22175 / 30000 Steps) (loss=1.68178): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22199 / 30000 Steps) (loss=1.76948): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22223 / 30000 Steps) (loss=1.75758): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22247 / 30000 Steps) (loss=1.89182): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22271 / 30000 Steps) (loss=1.90641): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (22295 / 30000 Steps) (loss=2.13553): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22319 / 30000 Steps) (loss=2.31815): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22343 / 30000 Steps) (loss=1.51507): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22367 / 30000 Steps) (loss=1.76590): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22391 / 30000 Steps) (loss=2.07999): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22415 / 30000 Steps) (loss=4.97741): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22439 / 30000 Steps) (loss=1.50233): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22463 / 30000 Steps) (loss=1.63693): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22487 / 30000 Steps) (loss=2.01772): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22500 / 30000 Steps) (loss=1.91705):  50%|█████     | 12/24 [00:11<00:10,  1.11it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (22488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (22488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.16s/it]\u001b[A\n",
+      "Validate (22488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.16s/it]\u001b[A\n",
+      "Validate (22488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.27s/it]\u001b[A\n",
+      "Validate (22488 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:13,  3.27s/it]\u001b[A\n",
+      "Validate (22488 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (22488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (22488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (22488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (22488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (22488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (22488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (22500 / 30000 Steps) (loss=1.91705):  54%|█████▍    | 13/24 [00:32<01:20,  7.29s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07064937800168991\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (22511 / 30000 Steps) (loss=1.68698): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (22535 / 30000 Steps) (loss=1.63871): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22559 / 30000 Steps) (loss=2.25367): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22583 / 30000 Steps) (loss=2.06416): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22607 / 30000 Steps) (loss=1.83227): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22631 / 30000 Steps) (loss=5.85255): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22655 / 30000 Steps) (loss=1.60104): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (22679 / 30000 Steps) (loss=1.69582): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22703 / 30000 Steps) (loss=1.99134): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22727 / 30000 Steps) (loss=2.05481): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22751 / 30000 Steps) (loss=2.27661): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22775 / 30000 Steps) (loss=3.47190): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (22799 / 30000 Steps) (loss=1.81042): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22823 / 30000 Steps) (loss=1.93108): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22847 / 30000 Steps) (loss=2.04849): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22871 / 30000 Steps) (loss=1.81162): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22895 / 30000 Steps) (loss=1.57702): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22919 / 30000 Steps) (loss=1.56488): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22943 / 30000 Steps) (loss=1.66431): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22967 / 30000 Steps) (loss=1.77493): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (22991 / 30000 Steps) (loss=1.85323): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23000 / 30000 Steps) (loss=2.16792):  33%|███▎      | 8/24 [00:08<00:14,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (22992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (22992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (22992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
+      "Validate (22992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (22992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (22992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (22992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (22992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (22992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (22992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (22992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (22992 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (23000 / 30000 Steps) (loss=2.16792):  38%|███▊      | 9/24 [00:29<01:52,  7.48s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07060033828020096\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (23015 / 30000 Steps) (loss=2.04952): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (23039 / 30000 Steps) (loss=2.40743): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23063 / 30000 Steps) (loss=2.29741): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23087 / 30000 Steps) (loss=2.12759): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23111 / 30000 Steps) (loss=1.70525): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (23135 / 30000 Steps) (loss=1.76106): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23159 / 30000 Steps) (loss=1.72826): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23183 / 30000 Steps) (loss=1.72953): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23207 / 30000 Steps) (loss=1.72956): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23231 / 30000 Steps) (loss=2.05098): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23255 / 30000 Steps) (loss=1.78883): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23279 / 30000 Steps) (loss=1.72608): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23303 / 30000 Steps) (loss=1.86829): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23327 / 30000 Steps) (loss=1.71274): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23351 / 30000 Steps) (loss=1.66257): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23375 / 30000 Steps) (loss=1.82341): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23399 / 30000 Steps) (loss=2.08193): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23423 / 30000 Steps) (loss=1.92278): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (23445 / 30000 Steps) (loss=1.52153):  88%|████████▊ | 21/24 [00:19<00:02,  1.10it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (24695 / 30000 Steps) (loss=2.01804): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (24719 / 30000 Steps) (loss=2.03579): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (24743 / 30000 Steps) (loss=2.46138): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (24767 / 30000 Steps) (loss=1.75445): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (24791 / 30000 Steps) (loss=1.94900): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (24815 / 30000 Steps) (loss=1.49325): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (24839 / 30000 Steps) (loss=1.83015): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (24863 / 30000 Steps) (loss=2.09343): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (24887 / 30000 Steps) (loss=1.76656): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (24911 / 30000 Steps) (loss=1.70951): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (24935 / 30000 Steps) (loss=1.51334): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (24959 / 30000 Steps) (loss=1.63255): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (24983 / 30000 Steps) (loss=1.96455): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25000 / 30000 Steps) (loss=1.98859):  67%|██████▋   | 16/24 [00:15<00:07,  1.11it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (24984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (24984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (24984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (24984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (24984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (24984 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (24984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (24984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (24984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (24984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (24984 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (24984 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
+      "Training (25000 / 30000 Steps) (loss=1.98859):  71%|███████   | 17/24 [00:36<00:50,  7.21s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.0705978125333786\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (25007 / 30000 Steps) (loss=2.22888): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (25031 / 30000 Steps) (loss=2.12894): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25055 / 30000 Steps) (loss=1.67851): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25079 / 30000 Steps) (loss=4.27627): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25103 / 30000 Steps) (loss=2.11039): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25127 / 30000 Steps) (loss=1.83662): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25151 / 30000 Steps) (loss=1.89494): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25175 / 30000 Steps) (loss=2.07800): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (25199 / 30000 Steps) (loss=1.71037): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25223 / 30000 Steps) (loss=1.66549): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (25247 / 30000 Steps) (loss=1.62141): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25271 / 30000 Steps) (loss=1.73209): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25295 / 30000 Steps) (loss=2.35269): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (25319 / 30000 Steps) (loss=1.67006): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25343 / 30000 Steps) (loss=2.01230): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25367 / 30000 Steps) (loss=1.53763): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25391 / 30000 Steps) (loss=1.90676): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25415 / 30000 Steps) (loss=1.87406): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25439 / 30000 Steps) (loss=1.91836): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25463 / 30000 Steps) (loss=1.67455): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25487 / 30000 Steps) (loss=2.42163): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25500 / 30000 Steps) (loss=1.73544):  50%|█████     | 12/24 [00:11<00:10,  1.09it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (25488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (25488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (25488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (25488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (25488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
+      "Validate (25488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (25488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (25488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (25488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (25488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (25488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (25488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (25500 / 30000 Steps) (loss=1.73544):  54%|█████▍    | 13/24 [00:32<01:20,  7.28s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07065335661172867\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (25511 / 30000 Steps) (loss=1.98163): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (25535 / 30000 Steps) (loss=1.90531): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25559 / 30000 Steps) (loss=1.66735): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (25583 / 30000 Steps) (loss=1.74367): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25607 / 30000 Steps) (loss=2.38333): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25631 / 30000 Steps) (loss=1.92686): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25655 / 30000 Steps) (loss=2.20560): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25679 / 30000 Steps) (loss=2.16599): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25703 / 30000 Steps) (loss=1.76526): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25727 / 30000 Steps) (loss=2.01786): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25751 / 30000 Steps) (loss=1.74469): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25775 / 30000 Steps) (loss=2.09772): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25799 / 30000 Steps) (loss=1.73560): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25823 / 30000 Steps) (loss=1.76547): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25847 / 30000 Steps) (loss=1.88350): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25871 / 30000 Steps) (loss=1.72797): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25895 / 30000 Steps) (loss=1.92903): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (25919 / 30000 Steps) (loss=1.98125): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25943 / 30000 Steps) (loss=2.05416): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25967 / 30000 Steps) (loss=1.89355): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (25991 / 30000 Steps) (loss=1.78587): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (26000 / 30000 Steps) (loss=2.70790):  33%|███▎      | 8/24 [00:08<00:14,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (25992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (25992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (25992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (25992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.25s/it]\u001b[A\n",
+      "Validate (25992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.25s/it]\u001b[A\n",
+      "Validate (25992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (25992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (25992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (25992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (25992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (25992 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (25992 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
+      "Training (26000 / 30000 Steps) (loss=2.70790):  38%|███▊      | 9/24 [00:29<01:52,  7.47s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059632986783981\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (26015 / 30000 Steps) (loss=2.75751): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (26039 / 30000 Steps) (loss=2.22100): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (26042 / 30000 Steps) (loss=2.54282):  12%|█▎        | 3/24 [00:02<00:19,  1.09it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (27311 / 30000 Steps) (loss=2.06320): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27335 / 30000 Steps) (loss=1.96911): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27359 / 30000 Steps) (loss=2.08709): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27383 / 30000 Steps) (loss=2.45813): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27407 / 30000 Steps) (loss=1.77006): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (27431 / 30000 Steps) (loss=1.97804): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
+      "Training (27455 / 30000 Steps) (loss=1.70250): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27479 / 30000 Steps) (loss=1.85274): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27500 / 30000 Steps) (loss=2.18979):  83%|████████▎ | 20/24 [00:19<00:03,  1.11it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (27480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (27480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (27480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (27480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (27480 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (27480 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (27480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (27480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (27480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
+      "Validate (27480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (27480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (27480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (27500 / 30000 Steps) (loss=2.18979):  88%|████████▊ | 21/24 [00:40<00:21,  7.21s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07224365323781967\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (27503 / 30000 Steps) (loss=1.50000): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (27527 / 30000 Steps) (loss=2.27667): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27551 / 30000 Steps) (loss=2.03874): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27575 / 30000 Steps) (loss=2.35553): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27599 / 30000 Steps) (loss=2.26227): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27623 / 30000 Steps) (loss=2.12316): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27647 / 30000 Steps) (loss=1.58768): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (27671 / 30000 Steps) (loss=2.10832): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27695 / 30000 Steps) (loss=1.50807): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27719 / 30000 Steps) (loss=1.97166): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27743 / 30000 Steps) (loss=2.02111): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27767 / 30000 Steps) (loss=2.17290): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27791 / 30000 Steps) (loss=1.62545): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27815 / 30000 Steps) (loss=1.69387): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27839 / 30000 Steps) (loss=2.31924): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27863 / 30000 Steps) (loss=2.24957): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27887 / 30000 Steps) (loss=1.89561): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27911 / 30000 Steps) (loss=2.24539): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27935 / 30000 Steps) (loss=1.56361): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27959 / 30000 Steps) (loss=2.01321): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (27983 / 30000 Steps) (loss=1.92082): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28000 / 30000 Steps) (loss=2.47622):  67%|██████▋   | 16/24 [00:15<00:07,  1.11it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (27984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (27984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (27984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
+      "Validate (27984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (27984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
+      "Validate (27984 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (27984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
+      "Validate (27984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.97s/it]\u001b[A\n",
+      "Validate (27984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.97s/it]\u001b[A\n",
+      "Validate (27984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (27984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
+      "Validate (27984 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
+      "Training (28000 / 30000 Steps) (loss=2.47622):  71%|███████   | 17/24 [00:36<00:50,  7.25s/it]"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Model Was Saved ! Current Best Avg. Dice: 0.04291130602359772 Current Avg. Dice: 0.04291130602359772\n",
-      "Total train time: 69.60 seconds\n"
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.0713159367442131\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (28007 / 30000 Steps) (loss=1.80550): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (28031 / 30000 Steps) (loss=1.86145): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28055 / 30000 Steps) (loss=1.85558): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28079 / 30000 Steps) (loss=1.91898): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28103 / 30000 Steps) (loss=1.92812): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28127 / 30000 Steps) (loss=1.91062): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28151 / 30000 Steps) (loss=2.35851): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28175 / 30000 Steps) (loss=1.96296): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
+      "Training (28199 / 30000 Steps) (loss=2.03080): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28223 / 30000 Steps) (loss=1.88645): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28247 / 30000 Steps) (loss=1.96491): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28271 / 30000 Steps) (loss=2.02651): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28295 / 30000 Steps) (loss=1.71030): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28319 / 30000 Steps) (loss=1.89605): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28343 / 30000 Steps) (loss=1.70242): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28367 / 30000 Steps) (loss=1.84909): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28391 / 30000 Steps) (loss=2.02234): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28415 / 30000 Steps) (loss=1.91935): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28439 / 30000 Steps) (loss=1.55340): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28463 / 30000 Steps) (loss=1.59120): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28487 / 30000 Steps) (loss=1.56324): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28500 / 30000 Steps) (loss=1.99294):  50%|█████     | 12/24 [00:11<00:10,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (28488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (28488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (28488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
+      "Validate (28488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.25s/it]\u001b[A\n",
+      "Validate (28488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.25s/it]\u001b[A\n",
+      "Validate (28488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (28488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
+      "Validate (28488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (28488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
+      "Validate (28488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (28488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
+      "Validate (28488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
+      "Training (28500 / 30000 Steps) (loss=1.99294):  54%|█████▍    | 13/24 [00:32<01:19,  7.26s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07071000337600708\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (28511 / 30000 Steps) (loss=1.62969): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
+      "Training (28535 / 30000 Steps) (loss=2.18971): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28559 / 30000 Steps) (loss=2.10305): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28583 / 30000 Steps) (loss=1.82222): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28607 / 30000 Steps) (loss=2.32003): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28631 / 30000 Steps) (loss=1.76598): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28655 / 30000 Steps) (loss=1.68438): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (28663 / 30000 Steps) (loss=2.01976):  29%|██▉       | 7/24 [00:07<00:15,  1.10it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (29927 / 30000 Steps) (loss=1.82648): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (29951 / 30000 Steps) (loss=2.04575): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (29975 / 30000 Steps) (loss=1.79833): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
+      "Training (29999 / 30000 Steps) (loss=1.80398): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Total train time: 28567.65 seconds\n"
      ]
     },
     {
@@ -758,11 +2546,13 @@
     }
    ],
    "source": [
-    "torch.cuda.empty_cache()\n",
-    "# max_iterations = 30000\n",
-    "# eval_num = 500\n",
-    "max_iterations = 30\n",
-    "eval_num = 29\n",
+    "set_determinism(seed=0)\n",
+    "\n",
+    "# torch.cuda.empty_cache()\n",
+    "max_iterations = 30000\n",
+    "eval_num = 500\n",
+    "# max_iterations = 30\n",
+    "# eval_num = 29\n",
     "global_step = 0\n",
     "dice_val_best = 0.0\n",
     "global_step_best = 0\n",
@@ -780,14 +2570,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
-   "metadata": {},
+   "execution_count": 14,
+   "metadata": {
+    "scrolled": true
+   },
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "train completed, best_metric: 0.0808 at iteration: 59\n"
+      "train completed, best_metric: 0.0968 at iteration: 500\n"
      ]
     }
    ],
@@ -798,6 +2590,41 @@
     ")"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 864x432 with 2 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.figure(\"train\", (12, 6))\n",
+    "plt.subplot(1, 2, 1)\n",
+    "plt.title(\"Iteration Average Loss\")\n",
+    "x = [eval_num * (i + 1) for i in range(len(epoch_loss_values))]\n",
+    "y = epoch_loss_values\n",
+    "plt.xlabel(\"Iteration\")\n",
+    "plt.plot(x, y)\n",
+    "plt.subplot(1, 2, 2)\n",
+    "plt.title(\"Val Mean Dice\")\n",
+    "x = [eval_num * (i + 1) for i in range(len(metric_values))]\n",
+    "y = metric_values\n",
+    "plt.xlabel(\"Iteration\")\n",
+    "plt.plot(x, y)\n",
+    "plt.show()"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
diff --git a/3d_segmentation/swin_unetr_profiling.py b/3d_segmentation/swin_unetr_profiling.py
new file mode 100644
index 0000000000..f9ad62135c
--- /dev/null
+++ b/3d_segmentation/swin_unetr_profiling.py
@@ -0,0 +1,333 @@
+# command line: "pip install nibabel==3.1.1; pip install tqdm==4.63.0; nsys profile --output /results/test_output --force-overwrite true --trace-fork-before-exec true python3 swin_unetr_profiling.py --epochs 5 --val_epochs 5 --batch_size 1 --thread_workers False --num_workers 0"
+
+import os
+import shutil
+import tempfile
+
+import matplotlib.pyplot as plt
+import numpy as np
+from tqdm import tqdm
+import time
+
+from monai.losses import DiceCELoss
+from monai.inferers import sliding_window_inference
+from monai.transforms import (
+    AsDiscrete,
+    AddChanneld,
+    Compose,
+    CropForegroundd,
+    LoadImaged,
+    Orientationd,
+    RandFlipd,
+    RandCropByPosNegLabeld,
+    RandShiftIntensityd,
+    ScaleIntensityRanged,
+    Spacingd,
+    RandRotate90d,
+    ToTensord,
+    EnsureTyped,
+    ToDeviced
+)
+from monai.utils import set_determinism
+
+from monai.config import print_config
+from monai.metrics import DiceMetric
+from monai.networks.nets import SwinUNETR
+
+from monai.data import (
+    ThreadDataLoader,
+    DataLoader,
+    CacheDataset,
+    load_decathlon_datalist,
+    decollate_batch,
+)
+
+import nvtx
+from monai.utils.nvtx import Range
+
+import torch
+
+import argparse
+
+print_config()
+
+parser = argparse.ArgumentParser(description='Profiling Swin UNETR.')
+parser.add_argument('--epochs', default=5, type=int)
+parser.add_argument('--val_epochs', default=-1, type=int, help='validation every X epochs; if non-positive value entered, will perform validation only once')
+parser.add_argument('--batch_size', default=1, type=int)
+parser.add_argument('--thread_workers', default=False, type=bool)
+parser.add_argument('--num_workers', default=0, type=int)
+args = parser.parse_args()
+print(args)
+
+assert args.epochs > 0
+assert args.batch_size > 0
+assert args.num_workers >= 0
+
+if not args.thread_workers:
+    args.num_workers = 0
+
+max_iterations = args.epochs * args.batch_size
+eval_num = args.val_epochs * args.batch_size - 1
+
+
+
+directory = os.environ.get("MONAI_DATA_DIRECTORY")
+root_dir = tempfile.mkdtemp() if directory is None else directory
+print(root_dir)
+
+
+num_samples = 4
+os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+train_transforms = Compose(
+    [
+        Range()(LoadImaged(keys=["image", "label"])),
+        Range()(AddChanneld(keys=["image", "label"])),
+        Range()(Orientationd(keys=["image", "label"], axcodes="RAS")),
+        Range()(Spacingd(
+            keys=["image", "label"],
+            pixdim=(1.5, 1.5, 2.0),
+            mode=("bilinear", "nearest"),
+        )),
+        Range()(ScaleIntensityRanged(
+            keys=["image"],
+            a_min=-175,
+            a_max=250,
+            b_min=0.0,
+            b_max=1.0,
+            clip=True,
+        )),
+        Range()(CropForegroundd(keys=["image", "label"], source_key="image")),
+#         Range()(EnsureTyped(keys=["image", "label"])),
+#         Range()(ToDeviced(keys=["image", "label"], device=device)),
+        Range()(RandCropByPosNegLabeld(
+            keys=["image", "label"],
+            label_key="label",
+            spatial_size=(96, 96, 96),
+            pos=1,
+            neg=1,
+            num_samples=num_samples,
+            image_key="image",
+            image_threshold=0,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[0],
+            prob=0.10,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[1],
+            prob=0.10,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[2],
+            prob=0.10,
+        )),
+        Range()(RandRotate90d(
+            keys=["image", "label"],
+            prob=0.10,
+            max_k=3,
+        )),
+        Range()(RandShiftIntensityd(
+            keys=["image"],
+            offsets=0.10,
+            prob=0.50,
+        )),
+        Range()(ToTensord(keys=["image", "label"])),
+    ]
+)
+val_transforms = Compose(
+    [
+        LoadImaged(keys=["image", "label"]),
+        AddChanneld(keys=["image", "label"]),
+        Orientationd(keys=["image", "label"], axcodes="RAS"),
+        Spacingd(
+            keys=["image", "label"],
+            pixdim=(1.5, 1.5, 2.0),
+            mode=("bilinear", "nearest"),
+        ),
+        ScaleIntensityRanged(
+            keys=["image"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True
+        ),
+        CropForegroundd(keys=["image", "label"], source_key="image"),
+#        ToTensord(keys=["image", "label"]),
+        EnsureTyped(keys=["image", "label"]),
+        ToDeviced(keys=["image", "label"], device=device),
+    ]
+)
+
+
+torch.cuda.empty_cache()
+data_dir = "data/"
+split_JSON = "dataset_0.json"
+
+datasets = data_dir + split_JSON
+datalist = load_decathlon_datalist(datasets, True, "training")
+val_files = load_decathlon_datalist(datasets, True, "validation")
+
+# TODO: try thread_workers
+train_ds = CacheDataset(
+    data=datalist,
+    transform=train_transforms,
+    cache_num=24,
+    cache_rate=1.0,
+    num_workers=8,
+)
+# train_loader = ThreadDataLoader(train_ds, batch_size=1, shuffle=True, num_workers=0)
+train_loader = ThreadDataLoader(train_ds, batch_size=args.batch_size, shuffle=True, use_thread_workers=args.thread_workers, num_workers=args.num_workers)
+
+val_ds = CacheDataset(
+    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4
+)
+val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)
+
+
+model = SwinUNETR(
+    img_size=(96, 96, 96),
+    in_channels=1,
+    out_channels=14,
+    feature_size=48,
+    use_checkpoint=True,
+).to(device)
+
+
+weight = torch.load("./model_swinvit.pt")
+model.load_from(weights=weight)
+print("Using pretrained self-supervied Swin UNETR backbone weights !")
+
+
+torch.backends.cudnn.benchmark = True
+loss_function = DiceCELoss(to_onehot_y=True, softmax=True)
+optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)
+
+post_label = AsDiscrete(to_onehot=14)
+post_pred = AsDiscrete(argmax=True, to_onehot=14)
+dice_metric = DiceMetric(include_background=True, reduction="mean", get_not_nans=False)
+
+def validation(epoch_iterable_val):
+    model.eval()
+    epoch_iterator_val = iter(epoch_iterable_val)
+    with torch.no_grad():
+        for _ in range(len(epoch_iterable_val)):
+            with nvtx.annotate("val dataload", color="red"):
+                batch = next(epoch_iterator_val)
+                val_inputs, val_labels = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
+
+            with nvtx.annotate("sliding window", color="green"):
+#                 with torch.cuda.amp.autocast():
+                val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
+            
+            with nvtx.annotate("decollate batch", color="blue"):
+                val_labels_list = decollate_batch(val_labels)
+                val_labels_convert = [
+                    post_label(val_label_tensor) for val_label_tensor in val_labels_list
+                ]
+                val_outputs_list = decollate_batch(val_outputs)
+                val_output_convert = [
+                    post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list
+                ]
+
+            with nvtx.annotate("compute metric", color="yellow"):
+                dice_metric(y_pred=val_output_convert, y=val_labels_convert)
+            
+            epoch_iterable_val.set_description(
+                "Validate (%d / %d Steps)" % (global_step, 10.0)
+            )
+
+        mean_dice_val = dice_metric.aggregate().item()
+        dice_metric.reset()
+    return mean_dice_val
+
+
+def train(global_step, train_loader, dice_val_best, global_step_best):
+    model.train()
+    epoch_loss = 0
+    step = 0
+    epoch_iterable = tqdm(
+        train_loader, desc="Training (X / X Steps) (loss=X.X)", dynamic_ncols=True
+    )
+    epoch_iterator = iter(epoch_iterable)
+
+    for step in range(len(epoch_iterable)):
+        step += 1
+
+        with nvtx.annotate("dataload", color="red"):
+            batch = next(epoch_iterator)
+            x, y = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
+#         with torch.cuda.amp.autocast():
+        with nvtx.annotate("forward", color="green"):
+            logit_map = model(x)
+            loss = loss_function(logit_map, y)
+
+        with nvtx.annotate("backward", color="blue"):
+            loss.backward()
+            epoch_loss += loss.item()
+
+        with nvtx.annotate("update", color="yellow"):
+            optimizer.step()
+            optimizer.zero_grad()
+        
+        epoch_iterable.set_description(
+            "Training (%d / %d Steps) (loss=%2.5f)"
+            % (global_step, max_iterations, loss)
+        )
+        if (
+            global_step % eval_num == 0 and global_step != 0
+        ) or global_step == max_iterations:
+            epoch_iterable_val = tqdm(
+                val_loader, desc="Validate (X / X Steps) (dice=X.X)", dynamic_ncols=True
+            )
+            dice_val = validation(epoch_iterable_val)
+            # FIXME: epoch_loss is a running average at time of validation??
+            epoch_loss /= step
+            epoch_loss_values.append(epoch_loss)
+            metric_values.append(dice_val)
+            if dice_val > dice_val_best:
+                dice_val_best = dice_val
+                global_step_best = global_step
+                torch.save(
+                    model.state_dict(), os.path.join(root_dir, "best_metric_model.pth")
+                )
+                print(
+                    "Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
+                        dice_val_best, dice_val
+                    )
+                )
+            else:
+                print(
+                    "Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
+                        dice_val_best, dice_val
+                    )
+                )
+        global_step += 1
+    return global_step, dice_val_best, global_step_best
+
+# max_iterations = 30000
+# eval_num = 500
+set_determinism(seed=0)
+global_step = 0
+dice_val_best = 0.0
+global_step_best = 0
+epoch_loss_values = []
+metric_values = []
+
+begin = time.time()
+while global_step < max_iterations:
+    with nvtx.annotate("epoch", color="red"):
+        global_step, dice_val_best, global_step_best = train(
+            global_step, train_loader, dice_val_best, global_step_best
+        )
+print(f"Total train time: {time.time() - begin:.2f} seconds")
+
+
+print(
+    f"train completed, best_metric: {dice_val_best:.4f} "
+    f"at iteration: {global_step_best}"
+)
+
+if directory is None:
+    shutil.rmtree(root_dir)

From fc16f5a05dfa1de062fdfbb0ce7d2c76982d7cc6 Mon Sep 17 00:00:00 2001
From: "pre-commit-ci[bot]"
 <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Date: Tue, 12 Jul 2022 15:01:36 +0000
Subject: [PATCH 04/17] [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci
---
 3d_segmentation/swin_unetr_profiling.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/3d_segmentation/swin_unetr_profiling.py b/3d_segmentation/swin_unetr_profiling.py
index f9ad62135c..adac531bcd 100644
--- a/3d_segmentation/swin_unetr_profiling.py
+++ b/3d_segmentation/swin_unetr_profiling.py
@@ -220,7 +220,7 @@ def validation(epoch_iterable_val):
             with nvtx.annotate("sliding window", color="green"):
 #                 with torch.cuda.amp.autocast():
                 val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
-            
+
             with nvtx.annotate("decollate batch", color="blue"):
                 val_labels_list = decollate_batch(val_labels)
                 val_labels_convert = [
@@ -233,7 +233,7 @@ def validation(epoch_iterable_val):
 
             with nvtx.annotate("compute metric", color="yellow"):
                 dice_metric(y_pred=val_output_convert, y=val_labels_convert)
-            
+
             epoch_iterable_val.set_description(
                 "Validate (%d / %d Steps)" % (global_step, 10.0)
             )
@@ -270,7 +270,7 @@ def train(global_step, train_loader, dice_val_best, global_step_best):
         with nvtx.annotate("update", color="yellow"):
             optimizer.step()
             optimizer.zero_grad()
-        
+
         epoch_iterable.set_description(
             "Training (%d / %d Steps) (loss=%2.5f)"
             % (global_step, max_iterations, loss)

From 26063a7cb600921868de89a6bcc113b040fc620e Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Fri, 15 Jul 2022 19:58:53 -0700
Subject: [PATCH 05/17] checkpoint experiment script

---
 3d_segmentation/swin_unetr_orig-091.ipynb | 2308 +++++++++++++++++++++
 1 file changed, 2308 insertions(+)
 create mode 100644 3d_segmentation/swin_unetr_orig-091.ipynb

diff --git a/3d_segmentation/swin_unetr_orig-091.ipynb b/3d_segmentation/swin_unetr_orig-091.ipynb
new file mode 100644
index 0000000000..7e2e0e1928
--- /dev/null
+++ b/3d_segmentation/swin_unetr_orig-091.ipynb
@@ -0,0 +1,2308 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# 3D Multi-organ Segmentation with Swin UNETR  (BTCV Challenge)\n",
+    "\n",
+    "\n",
+    "This tutorial uses a Swin UNETR [1] model for the task of multi-organ segmentation task using the BTCV challenge dataset. The architecture of Swin UNETR is demonstrated as below\n",
+    "![image](https://lh3.googleusercontent.com/pw/AM-JKLVx_J2DKYA7DCo3F_gGbK2e1sI_yzjYwQt-EWCirNGKsUv1hi7qLMofkY0r5xVXJNzhr8qenBkUJJYXtj49xsWJgOgbkBpcN7rz9axkeN3tgJbWldtZhYcBgYOlklzUS34eMCL-gRkxyFydJQ_Y1HAx=w1322-h518-no?authuser=2)\n",
+    "\n",
+    "The following features are included in this tutorial:\n",
+    "1. Transforms for dictionary format data.\n",
+    "1. Define a new transform according to MONAI transform API.\n",
+    "1. Load Nifti image with metadata, load a list of images and stack them.\n",
+    "1. Randomly adjust intensity for data augmentation.\n",
+    "1. Cache IO and transforms to accelerate training and validation.\n",
+    "1. Swin UNETR model, DiceCE loss function, Mean Dice metric for multi-organ segmentation task.\n",
+    "\n",
+    "For this tutorial, the dataset needs to be downloaded from: https://www.synapse.org/#!Synapse:syn3193805/wiki/217752. \n",
+    "\n",
+    "In addition, the json file for data splits needs to be downloaded from this [link](https://drive.google.com/file/d/1t4fIQQkONv7ArTSZe4Nucwkk1KfdUDvW/view?usp=sharing). Once downloaded, place the json file in the same folder as the dataset. \n",
+    "\n",
+    "For BTCV dataset, under Institutional Review Board (IRB) supervision, 50 abdomen CT scans of were randomly selected from a combination of an ongoing colorectal cancer chemotherapy trial, and a retrospective ventral hernia study. The 50 scans were captured during portal venous contrast phase with variable volume sizes (512 x 512 x 85 - 512 x 512 x 198) and field of views (approx. 280 x 280 x 280 mm3 - 500 x 500 x 650 mm3). The in-plane resolution varies from 0.54 x 0.54 mm2 to 0.98 x 0.98 mm2, while the slice thickness ranges from 2.5 mm to 5.0 mm. \n",
+    "\n",
+    "Target: 13 abdominal organs including 1. Spleen 2. Right Kidney 3. Left Kideny 4.Gallbladder 5.Esophagus 6. Liver 7. Stomach 8.Aorta 9. IVC 10. Portal and Splenic Veins 11. Pancreas 12 Right adrenal gland 13 Left adrenal gland.\n",
+    "\n",
+    "Modality: CT\n",
+    "Size: 30 3D volumes (24 Training + 6 Testing)  \n",
+    "Challenge: BTCV MICCAI Challenge\n",
+    "\n",
+    "The following figure shows image patches with the organ sub-regions that are annotated in the CT (top left) and the final labels for the whole dataset (right).\n",
+    "\n",
+    "Data, figures and resources are taken from: \n",
+    "\n",
+    "\n",
+    "1. [Self-Supervised Pre-Training of Swin Transformers\n",
+    "for 3D Medical Image Analysis](https://arxiv.org/abs/2111.14791)\n",
+    "\n",
+    "2. [Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images](https://arxiv.org/abs/2201.01266)\n",
+    "\n",
+    "3. [High-resolution 3D abdominal segmentation with random patch network fusion (MIA)](https://www.sciencedirect.com/science/article/abs/pii/S1361841520302589)\n",
+    "\n",
+    "4. [Efficient multi-atlas abdominal segmentation on clinically acquired CT with SIMPLE context learning (MIA)](https://www.sciencedirect.com/science/article/abs/pii/S1361841515000766?via%3Dihub)\n",
+    "\n",
+    "\n",
+    "![image](https://lh3.googleusercontent.com/pw/AM-JKLX0svvlMdcrchGAgiWWNkg40lgXYjSHsAAuRc5Frakmz2pWzSzf87JQCRgYpqFR0qAjJWPzMQLc_mmvzNjfF9QWl_1OHZ8j4c9qrbR6zQaDJWaCLArRFh0uPvk97qAa11HtYbD6HpJ-wwTCUsaPcYvM=w1724-h522-no?authuser=0)\n",
+    "\n",
+    "\n",
+    "\n",
+    "The image patches show anatomies of a subject, including: \n",
+    "1. large organs: spleen, liver, stomach. \n",
+    "2. Smaller organs: gallbladder, esophagus, kidneys, pancreas. \n",
+    "3. Vascular tissues: aorta, IVC, P&S Veins. \n",
+    "4. Glands: left and right adrenal gland\n",
+    "\n",
+    "If you find this tutorial helpful, please consider citing [1] and [2]:\n",
+    "\n",
+    "[1]: Tang, Y., Yang, D., Li, W., Roth, H.R., Landman, B., Xu, D., Nath, V. and Hatamizadeh, A., 2022. Self-supervised pre-training of swin transformers for 3d medical image analysis. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 20730-20740).\n",
+    "\n",
+    "[2]: Hatamizadeh, A., Nath, V., Tang, Y., Yang, D., Roth, H. and Xu, D., 2022. Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images. arXiv preprint arXiv:2201.01266.\n",
+    "\n",
+    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/Project-MONAI/tutorials/blob/main/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb)\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Pre-trained Swin UNETR Encoder\n",
+    "\n",
+    "We use weights from self-supervised pre-training of Swin UNETR encoder (3D Swin Tranformer) on a cohort of 5050 CT scans from publicly available datasets. The encoder is pre-trained using reconstructin, rotation prediction and contrastive learning pre-text tasks as shown below. For more details, please refer to [1] (CVPR paper) and see this [repository](https://github.com/Project-MONAI/research-contributions/tree/main/SwinUNETR/Pretrain). \n",
+    "\n",
+    "![image](https://lh3.googleusercontent.com/pw/AM-JKLVLgduGZ9naCSasWg09U665NBdd3UD4eLTy15wJiwbmKLS_p5WSZ2MBcRePEJO2tv9X3TkC52MsbnomuPy5JT3vSVeCji1MOEuAzcsxily88TdbHuAt6PzccefwKupbXyOCumK5hzz5Ul38kZnlEQ84=w397-h410-no?authuser=2)\n",
+    "\n",
+    "Please download the pre-trained weights from this [link](https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/model_swinvit.pt) and place it in the root directory of this tutorial. \n",
+    "\n",
+    "If training from scratch is desired, please skip the step for initializing from pre-trained weights. "
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Setup environment"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.8/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"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Fri Jul 15 20:35:49 2022       \n",
+      "+-----------------------------------------------------------------------------+\n",
+      "| NVIDIA-SMI 450.119.04   Driver Version: 450.119.04   CUDA Version: 11.6     |\n",
+      "|-------------------------------+----------------------+----------------------+\n",
+      "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
+      "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
+      "|                               |                      |               MIG M. |\n",
+      "|===============================+======================+======================|\n",
+      "|   0  Tesla V100-SXM2...  On   | 00000000:89:00.0 Off |                    0 |\n",
+      "| N/A   36C    P0    43W / 163W |      0MiB / 32510MiB |      0%      Default |\n",
+      "|                               |                      |                  N/A |\n",
+      "+-------------------------------+----------------------+----------------------+\n",
+      "                                                                               \n",
+      "+-----------------------------------------------------------------------------+\n",
+      "| Processes:                                                                  |\n",
+      "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
+      "|        ID   ID                                                   Usage      |\n",
+      "|=============================================================================|\n",
+      "|  No running processes found                                                 |\n",
+      "+-----------------------------------------------------------------------------+\n"
+     ]
+    }
+   ],
+   "source": [
+    "import torch\n",
+    "torch.cuda.empty_cache()\n",
+    "\n",
+    "!nvidia-smi"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
+      "Requirement already satisfied: monai==0.9.1rc3 in /opt/conda/lib/python3.8/site-packages (0.9.1rc3)\n",
+      "Requirement already satisfied: numpy>=1.17 in /opt/conda/lib/python3.8/site-packages (from monai==0.9.1rc3) (1.22.3)\n",
+      "Requirement already satisfied: torch>=1.7 in /opt/conda/lib/python3.8/site-packages (from monai==0.9.1rc3) (1.12.0a0+bd13bc6)\n",
+      "Requirement already satisfied: typing_extensions in /opt/conda/lib/python3.8/site-packages (from torch>=1.7->monai==0.9.1rc3) (4.1.1)\n",
+      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
+      "\u001b[0mLooking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
+      "Requirement already satisfied: nibabel==3.1.1 in /opt/conda/lib/python3.8/site-packages (3.1.1)\n",
+      "Requirement already satisfied: packaging>=14.3 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (21.3)\n",
+      "Requirement already satisfied: numpy>=1.13 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (1.22.3)\n",
+      "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.8/site-packages (from packaging>=14.3->nibabel==3.1.1) (3.0.8)\n",
+      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
+      "\u001b[0mLooking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
+      "Requirement already satisfied: tqdm==4.63.0 in /opt/conda/lib/python3.8/site-packages (4.63.0)\n",
+      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
+      "\u001b[0m"
+     ]
+    }
+   ],
+   "source": [
+    "# !pip install 'git+https://github.com/Project-MONAI/MONAI#egg.gitmonai@0.8.1+271.g07de215c'\n",
+    "!pip install monai==0.9.1rc3\n",
+    "!pip install nibabel==3.1.1\n",
+    "!pip install tqdm==4.63.0\n",
+    "!python -c \"import matplotlib\" || pip install -q matplotlib\n",
+    "%matplotlib inline"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.8/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"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "MONAI version: 0.9.1rc3\n",
+      "Numpy version: 1.22.3\n",
+      "Pytorch version: 1.12.0a0+bd13bc6\n",
+      "MONAI flags: HAS_EXT = False, USE_COMPILED = False, USE_META_DICT = False\n",
+      "MONAI rev id: 7a5de8b7b9db101a431e70ae2aa8ea7ebb8dfffe\n",
+      "MONAI __file__: /opt/conda/lib/python3.8/site-packages/monai/__init__.py\n",
+      "\n",
+      "Optional dependencies:\n",
+      "Pytorch Ignite version: 0.4.8\n",
+      "Nibabel version: 3.1.1\n",
+      "scikit-image version: 0.19.3\n",
+      "Pillow version: 9.0.1\n",
+      "Tensorboard version: 2.8.0\n",
+      "gdown version: 4.4.0\n",
+      "TorchVision version: 0.13.0a0\n",
+      "tqdm version: 4.63.0\n",
+      "lmdb version: 1.3.0\n",
+      "psutil version: 5.9.0\n",
+      "pandas version: 1.3.5\n",
+      "einops version: 0.4.1\n",
+      "transformers version: 4.19.4\n",
+      "mlflow version: 1.26.1\n",
+      "pynrrd version: 0.4.3\n",
+      "\n",
+      "For details about installing the optional dependencies, please visit:\n",
+      "    https://docs.monai.io/en/latest/installation.html#installing-the-recommended-dependencies\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "import os\n",
+    "import shutil\n",
+    "import tempfile\n",
+    "\n",
+    "import matplotlib.pyplot as plt\n",
+    "import numpy as np\n",
+    "from tqdm import tqdm\n",
+    "import time\n",
+    "\n",
+    "from monai.losses import DiceCELoss\n",
+    "from monai.inferers import sliding_window_inference\n",
+    "from monai.transforms import (\n",
+    "    AsDiscrete,\n",
+    "    AddChanneld,\n",
+    "    Compose,\n",
+    "    CropForegroundd,\n",
+    "    LoadImaged,\n",
+    "    Orientationd,\n",
+    "    RandFlipd,\n",
+    "    RandCropByPosNegLabeld,\n",
+    "    RandShiftIntensityd,\n",
+    "    ScaleIntensityRanged,\n",
+    "    Spacingd,\n",
+    "    RandRotate90d,\n",
+    "    ToTensord,\n",
+    "    ToDeviced,\n",
+    ")\n",
+    "\n",
+    "from monai.config import print_config\n",
+    "from monai.metrics import DiceMetric\n",
+    "from monai.networks.nets import SwinUNETR\n",
+    "\n",
+    "from monai.data import (\n",
+    "    DataLoader,\n",
+    "    ThreadDataLoader,\n",
+    "    CacheDataset,\n",
+    "    load_decathlon_datalist,\n",
+    "    decollate_batch,\n",
+    ")\n",
+    "\n",
+    "\n",
+    "import torch\n",
+    "\n",
+    "print_config()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Setup data directory\n",
+    "\n",
+    "You can specify a directory with the `MONAI_DATA_DIRECTORY` environment variable.  \n",
+    "This allows you to save results and reuse downloads.  \n",
+    "If not specified a temporary directory will be used."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "/tmp/tmpt26y6vq7\n"
+     ]
+    }
+   ],
+   "source": [
+    "directory = os.environ.get(\"MONAI_DATA_DIRECTORY\")\n",
+    "root_dir = tempfile.mkdtemp() if directory is None else directory\n",
+    "print(root_dir)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Setup transforms for training and validation\n",
+    "To save on GPU memory utilization, the num_samples can be reduced to 2. "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "num_samples = 4\n",
+    "\n",
+    "os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\"\n",
+    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
+    "\n",
+    "train_transforms = Compose(\n",
+    "    [\n",
+    "        LoadImaged(keys=[\"image\", \"label\"]),\n",
+    "        AddChanneld(keys=[\"image\", \"label\"]),\n",
+    "        Orientationd(keys=[\"image\", \"label\"], axcodes=\"RAS\"),\n",
+    "        Spacingd(\n",
+    "            keys=[\"image\", \"label\"],\n",
+    "            pixdim=(1.5, 1.5, 2.0),\n",
+    "            mode=(\"bilinear\", \"nearest\"),\n",
+    "        ),\n",
+    "        ScaleIntensityRanged(\n",
+    "            keys=[\"image\"],\n",
+    "            a_min=-175,\n",
+    "            a_max=250,\n",
+    "            b_min=0.0,\n",
+    "            b_max=1.0,\n",
+    "            clip=True,\n",
+    "        ),\n",
+    "        CropForegroundd(keys=[\"image\", \"label\"], source_key=\"image\"),\n",
+    "        RandCropByPosNegLabeld(\n",
+    "            keys=[\"image\", \"label\"],\n",
+    "            label_key=\"label\",\n",
+    "            spatial_size=(96, 96, 96),\n",
+    "            pos=1,\n",
+    "            neg=1,\n",
+    "            num_samples=num_samples,\n",
+    "            image_key=\"image\",\n",
+    "            image_threshold=0,\n",
+    "        ),\n",
+    "        RandFlipd(\n",
+    "            keys=[\"image\", \"label\"],\n",
+    "            spatial_axis=[0],\n",
+    "            prob=0.10,\n",
+    "        ),\n",
+    "        RandFlipd(\n",
+    "            keys=[\"image\", \"label\"],\n",
+    "            spatial_axis=[1],\n",
+    "            prob=0.10,\n",
+    "        ),\n",
+    "        RandFlipd(\n",
+    "            keys=[\"image\", \"label\"],\n",
+    "            spatial_axis=[2],\n",
+    "            prob=0.10,\n",
+    "        ),\n",
+    "        RandRotate90d(\n",
+    "            keys=[\"image\", \"label\"],\n",
+    "            prob=0.10,\n",
+    "            max_k=3,\n",
+    "        ),\n",
+    "        RandShiftIntensityd(\n",
+    "            keys=[\"image\"],\n",
+    "            offsets=0.10,\n",
+    "            prob=0.50,\n",
+    "        ),\n",
+    "        ToTensord(keys=[\"image\", \"label\"]),\n",
+    "#        ToDeviced(keys=[\"image\", \"label\"], device=device),\n",
+    "    ]\n",
+    ")\n",
+    "val_transforms = Compose(\n",
+    "    [\n",
+    "        LoadImaged(keys=[\"image\", \"label\"]),\n",
+    "        AddChanneld(keys=[\"image\", \"label\"]),\n",
+    "        Orientationd(keys=[\"image\", \"label\"], axcodes=\"RAS\"),\n",
+    "        Spacingd(\n",
+    "            keys=[\"image\", \"label\"],\n",
+    "            pixdim=(1.5, 1.5, 2.0),\n",
+    "            mode=(\"bilinear\", \"nearest\"),\n",
+    "        ),\n",
+    "        ScaleIntensityRanged(\n",
+    "            keys=[\"image\"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True\n",
+    "        ),\n",
+    "        CropForegroundd(keys=[\"image\", \"label\"], source_key=\"image\"),\n",
+    "       # ToDeviced(keys=[\"image\", \"label\"], device=device),\n",
+    "        ToTensord(keys=[\"image\", \"label\"]),\n",
+    "    ]\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Loading dataset: 100%|██████████| 24/24 [01:33<00:00,  3.91s/it]\n",
+      "Loading dataset: 100%|██████████| 6/6 [00:28<00:00,  4.78s/it]\n"
+     ]
+    }
+   ],
+   "source": [
+    "data_dir = \"data/\"\n",
+    "split_JSON = \"dataset_0.json\"\n",
+    "\n",
+    "datasets = data_dir + split_JSON\n",
+    "datalist = load_decathlon_datalist(datasets, True, \"training\")\n",
+    "val_files = load_decathlon_datalist(datasets, True, \"validation\")\n",
+    "train_ds = CacheDataset(\n",
+    "    data=datalist,\n",
+    "    transform=train_transforms,\n",
+    "    cache_num=24,\n",
+    "    cache_rate=1.0,\n",
+    "    num_workers=8,\n",
+    ")\n",
+    "train_loader = DataLoader(train_ds, batch_size=1, shuffle=True, num_workers=8, pin_memory=True)\n",
+    "# train_loader = ThreadDataLoader(train_ds, num_workers=0, batch_size=1, shuffle=True)\n",
+    "val_ds = CacheDataset(\n",
+    "    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4\n",
+    ")\n",
+    "val_loader = DataLoader(val_ds, batch_size=1, shuffle=False, num_workers=4, pin_memory=True)\n",
+    "# val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Check data shape and visualize"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "image shape: torch.Size([1, 255, 223, 276]), label shape: torch.Size([1, 255, 223, 276])\n"
+     ]
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 1296x432 with 2 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "slice_map = {\n",
+    "    \"img0035.nii.gz\": 170,\n",
+    "    \"img0036.nii.gz\": 230,\n",
+    "    \"img0037.nii.gz\": 204,\n",
+    "    \"img0038.nii.gz\": 204,\n",
+    "    \"img0039.nii.gz\": 204,\n",
+    "    \"img0040.nii.gz\": 180,\n",
+    "}\n",
+    "case_num = 1\n",
+    "img_name = os.path.split(val_ds[case_num][\"image_meta_dict\"][\"filename_or_obj\"])[1]\n",
+    "img = val_ds[case_num][\"image\"]\n",
+    "label = val_ds[case_num][\"label\"]\n",
+    "img_shape = img.shape\n",
+    "label_shape = label.shape\n",
+    "print(f\"image shape: {img_shape}, label shape: {label_shape}\")\n",
+    "plt.figure(\"image\", (18, 6))\n",
+    "plt.subplot(1, 2, 1)\n",
+    "plt.title(\"image\")\n",
+    "plt.imshow(img[0, :, :, slice_map[img_name]].detach().cpu(), cmap=\"gray\")\n",
+    "plt.subplot(1, 2, 2)\n",
+    "plt.title(\"label\")\n",
+    "plt.imshow(label[0, :, :, slice_map[img_name]].detach().cpu())\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Create Swin UNETR model\n",
+    "\n",
+    "In this scetion, we create Swin UNETR model for the 14-class multi-organ segmentation. We use a feature size of 48 which is compatible with self-supervised pre-trained weights. We also use gradient checkpointing (use_checkpoint) for more memory-efficient training. "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model = SwinUNETR(\n",
+    "    img_size=(96, 96, 96),\n",
+    "    in_channels=1,\n",
+    "    out_channels=14,\n",
+    "    feature_size=48,\n",
+    "    use_checkpoint=True,\n",
+    ").to(device)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Initialize Swin UNETR encoder from self-supervised pre-trained weights\n",
+    "\n",
+    "In this section, we intialize the Swin UNETR encoder from weights downloaded from this [link](https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/model_swinvit.pt). If training from scratch is desired, please skip this section."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Using pretrained self-supervied Swin UNETR backbone weights !\n"
+     ]
+    }
+   ],
+   "source": [
+    "weight = torch.load(\"./model_swinvit.pt\")\n",
+    "model.load_from(weights=weight)\n",
+    "print(\"Using pretrained self-supervied Swin UNETR backbone weights !\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Optimizer and loss function"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "torch.backends.cudnn.benchmark = True\n",
+    "loss_function = DiceCELoss(to_onehot_y=True, softmax=True)\n",
+    "optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Execute a typical PyTorch training process"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {
+    "scrolled": true
+   },
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (23 / 30000 Steps) (loss=3.18401): 100%|██████████| 24/24 [00:38<00:00,  1.62s/it]\n",
+      "Training (47 / 30000 Steps) (loss=3.07850): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (71 / 30000 Steps) (loss=3.11205): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (95 / 30000 Steps) (loss=3.10495): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (119 / 30000 Steps) (loss=3.04469): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (143 / 30000 Steps) (loss=2.98636): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (167 / 30000 Steps) (loss=2.85411): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (191 / 30000 Steps) (loss=2.71353): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (215 / 30000 Steps) (loss=2.65432): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (239 / 30000 Steps) (loss=2.71864): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (263 / 30000 Steps) (loss=2.65412): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (287 / 30000 Steps) (loss=2.70267): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (311 / 30000 Steps) (loss=2.63068): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (335 / 30000 Steps) (loss=2.54479): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (359 / 30000 Steps) (loss=2.56929): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (383 / 30000 Steps) (loss=2.54598): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (407 / 30000 Steps) (loss=2.54486): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (431 / 30000 Steps) (loss=2.67341): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (455 / 30000 Steps) (loss=2.57942): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (479 / 30000 Steps) (loss=2.47799): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (500 / 30000 Steps) (loss=2.68251):  83%|████████▎ | 20/24 [00:20<00:03,  1.12it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[ANone of the inputs have requires_grad=True. Gradients will be None\n",
+      "\n",
+      "Validate (480 / 10 Steps):   0%|          | 0/6 [00:06<?, ?it/s]        \u001b[A\n",
+      "Validate (480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:31,  6.29s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  17%|█▋        | 1/6 [00:09<00:31,  6.29s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:18,  4.56s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  33%|███▎      | 2/6 [00:12<00:18,  4.56s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  50%|█████     | 3/6 [00:12<00:10,  3.63s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  50%|█████     | 3/6 [00:17<00:10,  3.63s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:08,  4.40s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  67%|██████▋   | 4/6 [00:20<00:08,  4.40s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.96s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps):  83%|████████▎ | 5/6 [00:24<00:03,  3.96s/it]\u001b[A\n",
+      "Validate (480 / 10 Steps): 100%|██████████| 6/6 [00:24<00:00,  4.07s/it]\u001b[A\n",
+      "positional arguments and argument \"destination\" are deprecated. nn.Module.state_dict will not accept them in the future. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.\n",
+      "Training (500 / 30000 Steps) (loss=2.68251):  88%|████████▊ | 21/24 [00:45<00:25,  8.48s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.10523679107427597\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (503 / 30000 Steps) (loss=2.53570): 100%|██████████| 24/24 [00:48<00:00,  2.01s/it]\n",
+      "Training (527 / 30000 Steps) (loss=2.46318): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (551 / 30000 Steps) (loss=2.37891): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (575 / 30000 Steps) (loss=2.55742): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (599 / 30000 Steps) (loss=2.47565): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (623 / 30000 Steps) (loss=2.52184): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (647 / 30000 Steps) (loss=2.34970): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (671 / 30000 Steps) (loss=2.38541): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (695 / 30000 Steps) (loss=2.32277): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (719 / 30000 Steps) (loss=2.29923): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (743 / 30000 Steps) (loss=2.22664): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (767 / 30000 Steps) (loss=2.17650): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (791 / 30000 Steps) (loss=2.33309): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (815 / 30000 Steps) (loss=2.35450): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (839 / 30000 Steps) (loss=2.33414): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (863 / 30000 Steps) (loss=2.23909): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (887 / 30000 Steps) (loss=2.25688): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (911 / 30000 Steps) (loss=2.09836): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (935 / 30000 Steps) (loss=2.14540): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (959 / 30000 Steps) (loss=2.18719): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (983 / 30000 Steps) (loss=2.17081): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (1000 / 30000 Steps) (loss=2.13638):  67%|██████▋   | 16/24 [00:16<00:07,  1.12it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (984 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]        \u001b[A\n",
+      "Validate (984 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.02s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.02s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (984 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (1000 / 30000 Steps) (loss=2.13638):  71%|███████   | 17/24 [00:38<00:52,  7.54s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.07569018751382828\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (1007 / 30000 Steps) (loss=2.15620): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
+      "Training (1031 / 30000 Steps) (loss=2.23309): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1055 / 30000 Steps) (loss=2.18655): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (1079 / 30000 Steps) (loss=2.23391): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1103 / 30000 Steps) (loss=2.04205): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1127 / 30000 Steps) (loss=2.06677): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1151 / 30000 Steps) (loss=2.01106): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (1175 / 30000 Steps) (loss=2.01657): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1199 / 30000 Steps) (loss=2.04089): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1223 / 30000 Steps) (loss=1.95371): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (1247 / 30000 Steps) (loss=1.93151): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (1271 / 30000 Steps) (loss=2.18730): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1295 / 30000 Steps) (loss=1.89861): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1319 / 30000 Steps) (loss=1.89109): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1343 / 30000 Steps) (loss=2.08866): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1367 / 30000 Steps) (loss=1.99767): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1391 / 30000 Steps) (loss=1.92629): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1415 / 30000 Steps) (loss=1.89440): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1439 / 30000 Steps) (loss=1.99090): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (1463 / 30000 Steps) (loss=2.22657): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1487 / 30000 Steps) (loss=1.92651): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1500 / 30000 Steps) (loss=1.90171):  50%|█████     | 12/24 [00:13<00:10,  1.11it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (1488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
+      "Validate (1488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.05s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.05s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.64s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.64s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.13s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.13s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (1488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.69s/it]\u001b[A\n",
+      "Training (1500 / 30000 Steps) (loss=1.90171):  54%|█████▍    | 13/24 [00:35<01:23,  7.60s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.08025763183832169\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (1511 / 30000 Steps) (loss=2.25799): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
+      "Training (1535 / 30000 Steps) (loss=2.16516): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (1559 / 30000 Steps) (loss=1.88683): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1583 / 30000 Steps) (loss=1.93661): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (1607 / 30000 Steps) (loss=2.15409): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1631 / 30000 Steps) (loss=2.03945): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1655 / 30000 Steps) (loss=2.16960): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (1679 / 30000 Steps) (loss=1.93416): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1703 / 30000 Steps) (loss=1.89512): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1727 / 30000 Steps) (loss=1.89541): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1751 / 30000 Steps) (loss=2.10734): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (1775 / 30000 Steps) (loss=1.77242): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1799 / 30000 Steps) (loss=2.02548): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1823 / 30000 Steps) (loss=1.97165): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (1847 / 30000 Steps) (loss=1.98719): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (1871 / 30000 Steps) (loss=1.89960): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (1895 / 30000 Steps) (loss=1.82791): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1919 / 30000 Steps) (loss=2.12694): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (1943 / 30000 Steps) (loss=1.93836): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (1967 / 30000 Steps) (loss=1.99752): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (1991 / 30000 Steps) (loss=2.00321): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (2111 / 30000 Steps) (loss=1.87203): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2135 / 30000 Steps) (loss=1.83229): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (2159 / 30000 Steps) (loss=1.97508): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2183 / 30000 Steps) (loss=1.93266): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2207 / 30000 Steps) (loss=1.96755): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2231 / 30000 Steps) (loss=2.01195): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (2255 / 30000 Steps) (loss=1.87510): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2279 / 30000 Steps) (loss=1.74624): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (2303 / 30000 Steps) (loss=1.99921): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2327 / 30000 Steps) (loss=1.96728): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (2351 / 30000 Steps) (loss=1.77553): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2375 / 30000 Steps) (loss=1.96980): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (2399 / 30000 Steps) (loss=1.96445): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (2423 / 30000 Steps) (loss=1.95579): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2447 / 30000 Steps) (loss=1.82860): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (2471 / 30000 Steps) (loss=1.88289): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (2495 / 30000 Steps) (loss=1.86248): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2500 / 30000 Steps) (loss=1.96714):  17%|█▋        | 4/24 [00:05<00:21,  1.08s/it]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (2496 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
+      "Validate (2496 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.00s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.00s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.62s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.62s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.11s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.11s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (2496 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (2500 / 30000 Steps) (loss=1.96714):  21%|██        | 5/24 [00:27<02:50,  8.97s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.07246538996696472\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (2519 / 30000 Steps) (loss=1.76536): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
+      "Training (2543 / 30000 Steps) (loss=1.87272): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (2567 / 30000 Steps) (loss=2.24276): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (2591 / 30000 Steps) (loss=1.79909): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (2615 / 30000 Steps) (loss=1.85001): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (2639 / 30000 Steps) (loss=1.79896): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (2663 / 30000 Steps) (loss=1.71249): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2687 / 30000 Steps) (loss=1.89300): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (2711 / 30000 Steps) (loss=1.88981): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (2735 / 30000 Steps) (loss=1.81299): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (2759 / 30000 Steps) (loss=1.86044): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2783 / 30000 Steps) (loss=1.80493): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (2807 / 30000 Steps) (loss=1.88155): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (2831 / 30000 Steps) (loss=2.06799): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (2855 / 30000 Steps) (loss=1.82452): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (2879 / 30000 Steps) (loss=1.90656): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2903 / 30000 Steps) (loss=1.62591): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2927 / 30000 Steps) (loss=1.92314): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (2951 / 30000 Steps) (loss=1.80026): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (2975 / 30000 Steps) (loss=1.80520): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (2999 / 30000 Steps) (loss=1.79177): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (3000 / 30000 Steps) (loss=1.91799):   0%|          | 0/24 [00:02<?, ?it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (3000 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
+      "Validate (3000 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.02s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.02s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.62s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.62s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (3000 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (3000 / 30000 Steps) (loss=1.91799):   4%|▍         | 1/24 [00:24<09:30, 24.79s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.09305248409509659\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (3023 / 30000 Steps) (loss=1.88577): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
+      "Training (3047 / 30000 Steps) (loss=1.74826): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (3071 / 30000 Steps) (loss=1.76287): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (3095 / 30000 Steps) (loss=1.66413): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (3119 / 30000 Steps) (loss=1.70303): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (3143 / 30000 Steps) (loss=1.70093): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (3167 / 30000 Steps) (loss=1.65261): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (3191 / 30000 Steps) (loss=1.73050): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (3215 / 30000 Steps) (loss=2.02838): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (3239 / 30000 Steps) (loss=1.80735): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (3263 / 30000 Steps) (loss=1.60768): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (3287 / 30000 Steps) (loss=1.59386): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (3311 / 30000 Steps) (loss=1.88477): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (3335 / 30000 Steps) (loss=1.74821): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (3359 / 30000 Steps) (loss=1.58187): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (3383 / 30000 Steps) (loss=1.72284): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (3407 / 30000 Steps) (loss=1.76967): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (3431 / 30000 Steps) (loss=1.58086): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (3455 / 30000 Steps) (loss=1.64616): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (3464 / 30000 Steps) (loss=1.96805):  33%|███▎      | 8/24 [00:09<00:15,  1.06it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (4511 / 30000 Steps) (loss=1.71779): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
+      "Training (4535 / 30000 Steps) (loss=1.59975): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (4559 / 30000 Steps) (loss=1.95123): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (4583 / 30000 Steps) (loss=1.68883): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (4607 / 30000 Steps) (loss=2.08812): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (4631 / 30000 Steps) (loss=1.75501): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (4655 / 30000 Steps) (loss=1.70849): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (4679 / 30000 Steps) (loss=1.89590): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (4703 / 30000 Steps) (loss=1.58810): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (4727 / 30000 Steps) (loss=1.58057): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (4751 / 30000 Steps) (loss=1.63159): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (4775 / 30000 Steps) (loss=1.61231): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (4799 / 30000 Steps) (loss=1.57463): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (4823 / 30000 Steps) (loss=1.65194): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (4847 / 30000 Steps) (loss=1.70500): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (4871 / 30000 Steps) (loss=1.84604): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (4895 / 30000 Steps) (loss=1.57213): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (4919 / 30000 Steps) (loss=1.68219): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (4943 / 30000 Steps) (loss=2.06159): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (4967 / 30000 Steps) (loss=1.72085): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (4991 / 30000 Steps) (loss=1.89126): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5000 / 30000 Steps) (loss=1.59737):  33%|███▎      | 8/24 [00:09<00:14,  1.07it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (4992 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
+      "Validate (4992 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.09s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.09s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.66s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.66s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.14s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.14s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.10s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.10s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.77s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.77s/it]\u001b[A\n",
+      "Validate (4992 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.70s/it]\u001b[A\n",
+      "Training (5000 / 30000 Steps) (loss=1.59737):  38%|███▊      | 9/24 [00:31<01:58,  7.87s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.07061345130205154\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (5015 / 30000 Steps) (loss=1.81268): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
+      "Training (5039 / 30000 Steps) (loss=1.55078): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5063 / 30000 Steps) (loss=1.44579): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5087 / 30000 Steps) (loss=1.62221): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (5111 / 30000 Steps) (loss=2.66676): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5135 / 30000 Steps) (loss=2.57928): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5159 / 30000 Steps) (loss=1.67426): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5183 / 30000 Steps) (loss=1.69841): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (5207 / 30000 Steps) (loss=1.53911): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5231 / 30000 Steps) (loss=1.62084): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5255 / 30000 Steps) (loss=1.49868): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5279 / 30000 Steps) (loss=1.70572): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5303 / 30000 Steps) (loss=1.66650): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5327 / 30000 Steps) (loss=2.02091): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (5351 / 30000 Steps) (loss=1.63479): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5375 / 30000 Steps) (loss=1.77253): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5399 / 30000 Steps) (loss=1.70641): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (5423 / 30000 Steps) (loss=3.40272): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5447 / 30000 Steps) (loss=1.70207): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (5471 / 30000 Steps) (loss=1.50105): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5495 / 30000 Steps) (loss=1.67170): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5500 / 30000 Steps) (loss=1.56272):  17%|█▋        | 4/24 [00:06<00:22,  1.12s/it]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (5496 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
+      "Validate (5496 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.03s/it]\u001b[A\n",
+      "Validate (5496 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.03s/it]\u001b[A\n",
+      "Validate (5496 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (5496 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (5496 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (5496 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (5496 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (5496 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (5496 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (5496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (5496 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (5500 / 30000 Steps) (loss=1.56272):  21%|██        | 5/24 [00:28<02:51,  9.02s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.07060054689645767\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (5519 / 30000 Steps) (loss=1.59257): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
+      "Training (5543 / 30000 Steps) (loss=1.86619): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5567 / 30000 Steps) (loss=1.56463): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (5591 / 30000 Steps) (loss=1.95694): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5615 / 30000 Steps) (loss=1.88110): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (5639 / 30000 Steps) (loss=1.79884): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5663 / 30000 Steps) (loss=1.66443): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5687 / 30000 Steps) (loss=1.47059): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5711 / 30000 Steps) (loss=1.85159): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5735 / 30000 Steps) (loss=1.79990): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (5759 / 30000 Steps) (loss=1.96519): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (5783 / 30000 Steps) (loss=1.55542): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5807 / 30000 Steps) (loss=1.43398): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (5831 / 30000 Steps) (loss=1.57149): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (5855 / 30000 Steps) (loss=1.49405): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (5879 / 30000 Steps) (loss=1.72335): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (5883 / 30000 Steps) (loss=1.76143):  17%|█▋        | 4/24 [00:05<00:22,  1.10s/it]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (7175 / 30000 Steps) (loss=1.97261): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (7199 / 30000 Steps) (loss=1.58691): 100%|██████████| 24/24 [00:23<00:00,  1.00it/s]\n",
+      "Training (7223 / 30000 Steps) (loss=1.79031): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (7247 / 30000 Steps) (loss=1.60217): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7271 / 30000 Steps) (loss=2.03104): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (7295 / 30000 Steps) (loss=1.54362): 100%|██████████| 24/24 [00:23<00:00,  1.00it/s]\n",
+      "Training (7319 / 30000 Steps) (loss=1.54586): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7343 / 30000 Steps) (loss=1.68927): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7367 / 30000 Steps) (loss=2.33010): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7391 / 30000 Steps) (loss=1.55624): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (7415 / 30000 Steps) (loss=1.61111): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (7439 / 30000 Steps) (loss=1.79105): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (7463 / 30000 Steps) (loss=1.40654): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (7487 / 30000 Steps) (loss=1.52006): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (7500 / 30000 Steps) (loss=1.72726):  50%|█████     | 12/24 [00:13<00:10,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (7488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
+      "Validate (7488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.18s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.18s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.69s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.69s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.16s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.16s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.11s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.11s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.77s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.77s/it]\u001b[A\n",
+      "Validate (7488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.71s/it]\u001b[A\n",
+      "Training (7500 / 30000 Steps) (loss=1.72726):  54%|█████▍    | 13/24 [00:35<01:24,  7.65s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.08777529001235962\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (7511 / 30000 Steps) (loss=1.54710): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
+      "Training (7535 / 30000 Steps) (loss=1.55861): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7559 / 30000 Steps) (loss=1.42543): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (7583 / 30000 Steps) (loss=1.59021): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7607 / 30000 Steps) (loss=1.79987): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7631 / 30000 Steps) (loss=1.86244): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (7655 / 30000 Steps) (loss=1.61107): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (7679 / 30000 Steps) (loss=1.57211): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7703 / 30000 Steps) (loss=1.50199): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7727 / 30000 Steps) (loss=2.25318): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (7751 / 30000 Steps) (loss=1.59236): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7775 / 30000 Steps) (loss=1.68014): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7799 / 30000 Steps) (loss=1.36067): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7823 / 30000 Steps) (loss=1.66846): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7847 / 30000 Steps) (loss=1.79751): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (7871 / 30000 Steps) (loss=1.57973): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (7895 / 30000 Steps) (loss=1.90628): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (7919 / 30000 Steps) (loss=3.27129): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (7943 / 30000 Steps) (loss=1.73865): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (7967 / 30000 Steps) (loss=2.42319): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (7991 / 30000 Steps) (loss=1.71041): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (8000 / 30000 Steps) (loss=1.91273):  33%|███▎      | 8/24 [00:10<00:15,  1.05it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (7992 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
+      "Validate (7992 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.15s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.15s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.68s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.68s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.15s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.15s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.10s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.10s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (7992 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.71s/it]\u001b[A\n",
+      "Training (8000 / 30000 Steps) (loss=1.91273):  38%|███▊      | 9/24 [00:32<01:58,  7.91s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.09312240034341812\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (8015 / 30000 Steps) (loss=1.65287): 100%|██████████| 24/24 [00:46<00:00,  1.92s/it]\n",
+      "Training (8039 / 30000 Steps) (loss=1.75467): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (8063 / 30000 Steps) (loss=1.52946): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (8087 / 30000 Steps) (loss=1.99682): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (8111 / 30000 Steps) (loss=1.69648): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (8135 / 30000 Steps) (loss=1.58250): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (8159 / 30000 Steps) (loss=1.71409): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (8183 / 30000 Steps) (loss=1.61727): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (8207 / 30000 Steps) (loss=1.50567): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (8231 / 30000 Steps) (loss=1.67341): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (8255 / 30000 Steps) (loss=2.14218): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (8279 / 30000 Steps) (loss=1.36899): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (8303 / 30000 Steps) (loss=1.73868): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (8327 / 30000 Steps) (loss=1.77028): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (8351 / 30000 Steps) (loss=1.72490): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (8375 / 30000 Steps) (loss=5.67763): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (8399 / 30000 Steps) (loss=1.56018): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (8423 / 30000 Steps) (loss=1.92986): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (8447 / 30000 Steps) (loss=1.67452): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (8471 / 30000 Steps) (loss=1.36784): 100%|██████████| 24/24 [00:23<00:00,  1.00it/s]\n",
+      "Training (8495 / 30000 Steps) (loss=1.71360): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (8500 / 30000 Steps) (loss=1.65908):  17%|█▋        | 4/24 [00:06<00:23,  1.18s/it]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (8496 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
+      "Validate (8496 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:21,  4.20s/it]\u001b[A\n",
+      "Validate (8496 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:21,  4.20s/it]\u001b[A\n",
+      "Validate (8496 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.70s/it]\u001b[A\n",
+      "Validate (8496 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.70s/it]\u001b[A\n",
+      "Validate (8496 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.16s/it]\u001b[A\n",
+      "Validate (8496 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.16s/it]\u001b[A\n",
+      "Validate (8496 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.10s/it]\u001b[A\n",
+      "Validate (8496 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.10s/it]\u001b[A\n",
+      "Validate (8496 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (8496 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (8496 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.71s/it]\u001b[A\n",
+      "Training (8500 / 30000 Steps) (loss=1.65908):  21%|██        | 5/24 [00:28<02:52,  9.10s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.08893509954214096\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (8517 / 30000 Steps) (loss=1.72599):  88%|████████▊ | 21/24 [00:44<00:02,  1.08it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (9839 / 30000 Steps) (loss=1.51064): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (9863 / 30000 Steps) (loss=1.97902): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (9887 / 30000 Steps) (loss=1.91444): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (9911 / 30000 Steps) (loss=1.44662): 100%|██████████| 24/24 [00:23<00:00,  1.00it/s]\n",
+      "Training (9935 / 30000 Steps) (loss=1.47294): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (9959 / 30000 Steps) (loss=1.47041): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (9983 / 30000 Steps) (loss=1.36296): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (10000 / 30000 Steps) (loss=2.28292):  67%|██████▋   | 16/24 [00:17<00:07,  1.11it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (9984 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
+      "Validate (9984 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.13s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.13s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.67s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.67s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.15s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.15s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.10s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.10s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (9984 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.70s/it]\u001b[A\n",
+      "Training (10000 / 30000 Steps) (loss=2.28292):  71%|███████   | 17/24 [00:40<00:54,  7.83s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Saved ! Current Best Avg. Dice: 0.10929516702890396 Current Avg. Dice: 0.10929516702890396\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (10007 / 30000 Steps) (loss=1.96494): 100%|██████████| 24/24 [00:46<00:00,  1.95s/it]\n",
+      "Training (10031 / 30000 Steps) (loss=1.57745): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10055 / 30000 Steps) (loss=1.45904): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (10079 / 30000 Steps) (loss=1.63818): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (10103 / 30000 Steps) (loss=2.50689): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (10127 / 30000 Steps) (loss=1.73940): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (10151 / 30000 Steps) (loss=1.46289): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10175 / 30000 Steps) (loss=1.45769): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (10199 / 30000 Steps) (loss=1.78349): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (10223 / 30000 Steps) (loss=1.69970): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10247 / 30000 Steps) (loss=1.50493): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10271 / 30000 Steps) (loss=1.51015): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10295 / 30000 Steps) (loss=1.54784): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (10319 / 30000 Steps) (loss=1.37233): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (10343 / 30000 Steps) (loss=1.48155): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (10367 / 30000 Steps) (loss=1.54143): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (10391 / 30000 Steps) (loss=1.70770): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (10415 / 30000 Steps) (loss=2.30014): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (10439 / 30000 Steps) (loss=1.79295): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (10463 / 30000 Steps) (loss=1.55343): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10487 / 30000 Steps) (loss=2.00099): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (10500 / 30000 Steps) (loss=1.65427):  50%|█████     | 12/24 [00:13<00:10,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (10488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (10488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.13s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.13s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.67s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.67s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.14s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.14s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (10488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.69s/it]\u001b[A\n",
+      "Training (10500 / 30000 Steps) (loss=1.65427):  54%|█████▍    | 13/24 [00:35<01:23,  7.62s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10929516702890396 Current Avg. Dice: 0.10640474408864975\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (10511 / 30000 Steps) (loss=1.52513): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
+      "Training (10535 / 30000 Steps) (loss=1.67685): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (10559 / 30000 Steps) (loss=1.44793): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (10583 / 30000 Steps) (loss=3.96370): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (10607 / 30000 Steps) (loss=1.67806): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10631 / 30000 Steps) (loss=1.59580): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10655 / 30000 Steps) (loss=1.57280): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10679 / 30000 Steps) (loss=1.75091): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (10703 / 30000 Steps) (loss=1.66555): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (10727 / 30000 Steps) (loss=1.67437): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (10751 / 30000 Steps) (loss=1.44212): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (10775 / 30000 Steps) (loss=1.55636): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10799 / 30000 Steps) (loss=1.54452): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (10823 / 30000 Steps) (loss=1.62608): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (10847 / 30000 Steps) (loss=1.61273): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (10871 / 30000 Steps) (loss=1.66387): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (10895 / 30000 Steps) (loss=1.68224): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (10919 / 30000 Steps) (loss=1.69885): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (10943 / 30000 Steps) (loss=2.09697): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (10967 / 30000 Steps) (loss=5.33873): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (10991 / 30000 Steps) (loss=1.53754): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (11000 / 30000 Steps) (loss=1.61410):  33%|███▎      | 8/24 [00:10<00:15,  1.05it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (10992 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (10992 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.00s/it]\u001b[A\n",
+      "Validate (10992 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.00s/it]\u001b[A\n",
+      "Validate (10992 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.62s/it]\u001b[A\n",
+      "Validate (10992 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.62s/it]\u001b[A\n",
+      "Validate (10992 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.11s/it]\u001b[A\n",
+      "Validate (10992 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.11s/it]\u001b[A\n",
+      "Validate (10992 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (10992 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (10992 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (10992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (10992 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (11000 / 30000 Steps) (loss=1.61410):  38%|███▊      | 9/24 [00:32<01:57,  7.84s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.10929516702890396 Current Avg. Dice: 0.10596194118261337\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (11015 / 30000 Steps) (loss=2.03147): 100%|██████████| 24/24 [00:45<00:00,  1.91s/it]\n",
+      "Training (11039 / 30000 Steps) (loss=1.42184): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (11063 / 30000 Steps) (loss=1.60865): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (11087 / 30000 Steps) (loss=1.88724): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (11111 / 30000 Steps) (loss=1.37930): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (11135 / 30000 Steps) (loss=2.38266): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (11159 / 30000 Steps) (loss=1.86418): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (11173 / 30000 Steps) (loss=1.72598):  58%|█████▊    | 14/24 [00:14<00:08,  1.11it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (12479 / 30000 Steps) (loss=1.47541): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (12500 / 30000 Steps) (loss=1.64012):  83%|████████▎ | 20/24 [00:20<00:03,  1.12it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (12480 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (12480 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.03s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.03s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (12480 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.69s/it]\u001b[A\n",
+      "Training (12500 / 30000 Steps) (loss=1.64012):  88%|████████▊ | 21/24 [00:42<00:22,  7.55s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.11252329498529434\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (12503 / 30000 Steps) (loss=1.59704): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
+      "Training (12527 / 30000 Steps) (loss=1.49524): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (12551 / 30000 Steps) (loss=1.86410): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (12575 / 30000 Steps) (loss=2.06310): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (12599 / 30000 Steps) (loss=1.46205): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (12623 / 30000 Steps) (loss=2.00819): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (12647 / 30000 Steps) (loss=1.56068): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (12671 / 30000 Steps) (loss=1.91482): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (12695 / 30000 Steps) (loss=1.53840): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (12719 / 30000 Steps) (loss=1.78944): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (12743 / 30000 Steps) (loss=1.75246): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (12767 / 30000 Steps) (loss=2.23596): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (12791 / 30000 Steps) (loss=1.56143): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (12815 / 30000 Steps) (loss=2.48896): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (12839 / 30000 Steps) (loss=1.58305): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (12863 / 30000 Steps) (loss=2.21133): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (12887 / 30000 Steps) (loss=2.02335): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s] \n",
+      "Training (12911 / 30000 Steps) (loss=1.85969): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (12935 / 30000 Steps) (loss=1.82117): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (12959 / 30000 Steps) (loss=1.55216): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (12983 / 30000 Steps) (loss=1.80917): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (13000 / 30000 Steps) (loss=1.71191):  67%|██████▋   | 16/24 [00:17<00:07,  1.11it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (12984 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (12984 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.03s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.03s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (12984 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (13000 / 30000 Steps) (loss=1.71191):  71%|███████   | 17/24 [00:39<00:52,  7.55s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.09339217096567154\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (13007 / 30000 Steps) (loss=1.61073): 100%|██████████| 24/24 [00:45<00:00,  1.91s/it]\n",
+      "Training (13031 / 30000 Steps) (loss=1.64728): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13055 / 30000 Steps) (loss=1.88418): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13079 / 30000 Steps) (loss=2.15990): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (13103 / 30000 Steps) (loss=2.05648): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13127 / 30000 Steps) (loss=2.08542): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (13151 / 30000 Steps) (loss=1.66662): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (13175 / 30000 Steps) (loss=1.69000): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (13199 / 30000 Steps) (loss=1.54646): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13223 / 30000 Steps) (loss=1.91987): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (13247 / 30000 Steps) (loss=1.87953): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (13271 / 30000 Steps) (loss=1.74172): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (13295 / 30000 Steps) (loss=1.73502): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (13319 / 30000 Steps) (loss=2.05865): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13343 / 30000 Steps) (loss=1.57452): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (13367 / 30000 Steps) (loss=2.06786): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (13391 / 30000 Steps) (loss=1.71718): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (13415 / 30000 Steps) (loss=1.97281): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (13439 / 30000 Steps) (loss=1.58712): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13463 / 30000 Steps) (loss=1.71977): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s] \n",
+      "Training (13487 / 30000 Steps) (loss=1.44692): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (13500 / 30000 Steps) (loss=1.85141):  50%|█████     | 12/24 [00:13<00:10,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (13488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (13488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.06s/it]\u001b[A\n",
+      "Validate (13488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.06s/it]\u001b[A\n",
+      "Validate (13488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.65s/it]\u001b[A\n",
+      "Validate (13488 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.65s/it]\u001b[A\n",
+      "Validate (13488 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.13s/it]\u001b[A\n",
+      "Validate (13488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.13s/it]\u001b[A\n",
+      "Validate (13488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (13488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (13488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (13488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (13488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (13500 / 30000 Steps) (loss=1.85141):  54%|█████▍    | 13/24 [00:35<01:23,  7.60s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.08462151885032654\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (13511 / 30000 Steps) (loss=1.71781): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
+      "Training (13535 / 30000 Steps) (loss=6.14073): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (13559 / 30000 Steps) (loss=1.55927): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13583 / 30000 Steps) (loss=1.83615): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (13607 / 30000 Steps) (loss=1.71865): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (13631 / 30000 Steps) (loss=1.63441): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (13655 / 30000 Steps) (loss=1.82784): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13679 / 30000 Steps) (loss=1.63667): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13703 / 30000 Steps) (loss=1.88138): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13727 / 30000 Steps) (loss=1.65120): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (13751 / 30000 Steps) (loss=1.85407): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13775 / 30000 Steps) (loss=1.69582): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (13799 / 30000 Steps) (loss=1.57240): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (13823 / 30000 Steps) (loss=1.62583): 100%|██████████| 24/24 [00:23<00:00,  1.11it/s]IOPub message rate exceeded.\n",
+      "The notebook server will temporarily stop sending output\n",
+      "to the client in order to avoid crashing it.\n",
+      "To change this limit, set the config variable\n",
+      "`--NotebookApp.iopub_msg_rate_limit`.\n",
+      "\n",
+      "Current values:\n",
+      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
+      "NotebookApp.rate_limit_window=3.0 (secs)\n",
+      "\n",
+      "Training (15143 / 30000 Steps) (loss=1.83970): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15167 / 30000 Steps) (loss=1.66505): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (15191 / 30000 Steps) (loss=1.57074): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15215 / 30000 Steps) (loss=1.72749): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15239 / 30000 Steps) (loss=1.56530): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (15263 / 30000 Steps) (loss=1.59868): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15287 / 30000 Steps) (loss=1.89288): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15311 / 30000 Steps) (loss=2.37840): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15335 / 30000 Steps) (loss=1.51758): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15359 / 30000 Steps) (loss=1.64895): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15383 / 30000 Steps) (loss=1.61334): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (15407 / 30000 Steps) (loss=2.79856): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15431 / 30000 Steps) (loss=1.62389): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (15455 / 30000 Steps) (loss=1.70353): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15479 / 30000 Steps) (loss=1.78808): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15500 / 30000 Steps) (loss=1.73282):  83%|████████▎ | 20/24 [00:20<00:03,  1.12it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (15480 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (15480 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.11s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.11s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.66s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.66s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.14s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.14s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (15480 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.69s/it]\u001b[A\n",
+      "Training (15500 / 30000 Steps) (loss=1.73282):  88%|████████▊ | 21/24 [00:42<00:22,  7.54s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.07921255379915237\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (15503 / 30000 Steps) (loss=2.00451): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
+      "Training (15527 / 30000 Steps) (loss=1.62787): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15551 / 30000 Steps) (loss=1.68343): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15575 / 30000 Steps) (loss=1.70755): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (15599 / 30000 Steps) (loss=1.93015): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15623 / 30000 Steps) (loss=5.34464): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15647 / 30000 Steps) (loss=1.73335): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15671 / 30000 Steps) (loss=1.84574): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (15695 / 30000 Steps) (loss=2.00053): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15719 / 30000 Steps) (loss=1.58993): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15743 / 30000 Steps) (loss=1.91924): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
+      "Training (15767 / 30000 Steps) (loss=1.78188): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15791 / 30000 Steps) (loss=2.37636): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (15815 / 30000 Steps) (loss=2.01652): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (15839 / 30000 Steps) (loss=1.62255): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15863 / 30000 Steps) (loss=2.15401): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (15887 / 30000 Steps) (loss=2.04368): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (15911 / 30000 Steps) (loss=1.54641): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (15935 / 30000 Steps) (loss=1.92847): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (15959 / 30000 Steps) (loss=5.25715): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (15983 / 30000 Steps) (loss=1.82406): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16000 / 30000 Steps) (loss=1.71104):  67%|██████▋   | 16/24 [00:16<00:07,  1.12it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (15984 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (15984 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.02s/it]\u001b[A\n",
+      "Validate (15984 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.02s/it]\u001b[A\n",
+      "Validate (15984 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (15984 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (15984 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (15984 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (15984 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (15984 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (15984 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (15984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (15984 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (16000 / 30000 Steps) (loss=1.71104):  71%|███████   | 17/24 [00:38<00:52,  7.53s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.09620887041091919\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (16007 / 30000 Steps) (loss=2.44960): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
+      "Training (16031 / 30000 Steps) (loss=2.16274): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (16055 / 30000 Steps) (loss=1.76895): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (16079 / 30000 Steps) (loss=1.91601): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16103 / 30000 Steps) (loss=1.95242): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (16127 / 30000 Steps) (loss=1.67782): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (16151 / 30000 Steps) (loss=2.04866): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16175 / 30000 Steps) (loss=1.67603): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16199 / 30000 Steps) (loss=1.59732): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16223 / 30000 Steps) (loss=1.73291): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (16247 / 30000 Steps) (loss=1.53592): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (16271 / 30000 Steps) (loss=1.87026): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16295 / 30000 Steps) (loss=1.68480): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (16319 / 30000 Steps) (loss=4.17143): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (16343 / 30000 Steps) (loss=1.85085): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16367 / 30000 Steps) (loss=1.99687): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16391 / 30000 Steps) (loss=1.83488): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16415 / 30000 Steps) (loss=1.67751): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (16439 / 30000 Steps) (loss=1.77609): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16463 / 30000 Steps) (loss=1.68322): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (16487 / 30000 Steps) (loss=2.76593): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16500 / 30000 Steps) (loss=1.76284):  50%|█████     | 12/24 [00:13<00:10,  1.10it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (16488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (16488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.14s/it]\u001b[A\n",
+      "Validate (16488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.14s/it]\u001b[A\n",
+      "Validate (16488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.67s/it]\u001b[A\n",
+      "Validate (16488 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.67s/it]\u001b[A\n",
+      "Validate (16488 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.15s/it]\u001b[A\n",
+      "Validate (16488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.15s/it]\u001b[A\n",
+      "Validate (16488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.10s/it]\u001b[A\n",
+      "Validate (16488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.10s/it]\u001b[A\n",
+      "Validate (16488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (16488 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (16488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.70s/it]\u001b[A\n",
+      "Training (16500 / 30000 Steps) (loss=1.76284):  54%|█████▍    | 13/24 [00:35<01:23,  7.63s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.098011814057827\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (16511 / 30000 Steps) (loss=1.59225): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
+      "Training (16535 / 30000 Steps) (loss=1.81074): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16559 / 30000 Steps) (loss=1.66221): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (16583 / 30000 Steps) (loss=1.65103): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16607 / 30000 Steps) (loss=1.78824): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (16631 / 30000 Steps) (loss=2.45114): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16655 / 30000 Steps) (loss=2.39868): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (16679 / 30000 Steps) (loss=1.75542): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (16703 / 30000 Steps) (loss=1.64762): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16727 / 30000 Steps) (loss=1.81028): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16751 / 30000 Steps) (loss=1.71452): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16775 / 30000 Steps) (loss=1.52099): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16799 / 30000 Steps) (loss=1.52047): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (16823 / 30000 Steps) (loss=1.86302): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (16847 / 30000 Steps) (loss=1.48955): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16871 / 30000 Steps) (loss=1.52795): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (16895 / 30000 Steps) (loss=1.51909): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16919 / 30000 Steps) (loss=1.65529): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (16943 / 30000 Steps) (loss=1.79362): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16967 / 30000 Steps) (loss=1.66853): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (16991 / 30000 Steps) (loss=1.53858): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (17000 / 30000 Steps) (loss=2.06464):  33%|███▎      | 8/24 [00:09<00:14,  1.07it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (16992 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (16992 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.05s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.05s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.64s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.64s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (16992 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (17000 / 30000 Steps) (loss=2.06464):  38%|███▊      | 9/24 [00:31<01:57,  7.83s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.0814722552895546\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (17015 / 30000 Steps) (loss=2.18220): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
+      "Training (17039 / 30000 Steps) (loss=2.03658): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (17063 / 30000 Steps) (loss=2.23238): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (17087 / 30000 Steps) (loss=1.74233): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (17111 / 30000 Steps) (loss=1.88046): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (17135 / 30000 Steps) (loss=1.68661): 100%|██████████| 24/24 [00:24<00:00,  1.01s/it]\n",
+      "Training (17159 / 30000 Steps) (loss=1.75913): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (17183 / 30000 Steps) (loss=1.90502): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (17207 / 30000 Steps) (loss=1.95610): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (17231 / 30000 Steps) (loss=8.81049): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (17255 / 30000 Steps) (loss=1.61983): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (17279 / 30000 Steps) (loss=2.39689): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (17303 / 30000 Steps) (loss=1.66409): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
+      "Training (17327 / 30000 Steps) (loss=2.19589): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (17351 / 30000 Steps) (loss=2.09781): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (17375 / 30000 Steps) (loss=1.80089): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (17399 / 30000 Steps) (loss=2.05174): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (17423 / 30000 Steps) (loss=2.08487): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (17447 / 30000 Steps) (loss=1.57698): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (17471 / 30000 Steps) (loss=1.92801): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (17495 / 30000 Steps) (loss=1.85291): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (17500 / 30000 Steps) (loss=1.74886):  17%|█▋        | 4/24 [00:06<00:22,  1.13s/it]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (17496 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (17496 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.02s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.02s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (17496 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (17500 / 30000 Steps) (loss=1.74886):  21%|██        | 5/24 [00:28<02:51,  9.01s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.08850126713514328\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (17519 / 30000 Steps) (loss=1.75699): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
+      "Training (17543 / 30000 Steps) (loss=1.79229): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (17567 / 30000 Steps) (loss=1.73489): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (17591 / 30000 Steps) (loss=1.83942): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (17615 / 30000 Steps) (loss=1.62040): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (17639 / 30000 Steps) (loss=1.86634): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (17783 / 30000 Steps) (loss=1.57847): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (17807 / 30000 Steps) (loss=1.63421): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (17831 / 30000 Steps) (loss=1.93467): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (17855 / 30000 Steps) (loss=1.90310): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (17879 / 30000 Steps) (loss=1.66063): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (17903 / 30000 Steps) (loss=2.17549): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s] \n",
+      "Training (17927 / 30000 Steps) (loss=2.14292): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (17951 / 30000 Steps) (loss=1.81027): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (17975 / 30000 Steps) (loss=1.52826): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (17999 / 30000 Steps) (loss=1.61435): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18000 / 30000 Steps) (loss=1.67718):   0%|          | 0/24 [00:02<?, ?it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (18000 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (18000 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.06s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.06s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.64s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.64s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.13s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.13s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
+      "Validate (18000 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.69s/it]\u001b[A\n",
+      "Training (18000 / 30000 Steps) (loss=1.67718):   4%|▍         | 1/24 [00:24<09:23, 24.48s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.08198585361242294\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (18023 / 30000 Steps) (loss=2.14477): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
+      "Training (18047 / 30000 Steps) (loss=1.62933): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (18071 / 30000 Steps) (loss=2.26788): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (18095 / 30000 Steps) (loss=2.25271): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18119 / 30000 Steps) (loss=1.76527): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18143 / 30000 Steps) (loss=1.63859): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (18167 / 30000 Steps) (loss=2.15393): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (18191 / 30000 Steps) (loss=1.63407): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (18215 / 30000 Steps) (loss=1.72186): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (18239 / 30000 Steps) (loss=1.76017): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (18263 / 30000 Steps) (loss=1.71748): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18287 / 30000 Steps) (loss=2.37135): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (18311 / 30000 Steps) (loss=2.01968): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (18335 / 30000 Steps) (loss=1.87879): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18359 / 30000 Steps) (loss=1.72722): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (18383 / 30000 Steps) (loss=1.54340): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (18407 / 30000 Steps) (loss=1.68834): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18431 / 30000 Steps) (loss=1.72274): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
+      "Training (18455 / 30000 Steps) (loss=1.71656): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
+      "Training (18479 / 30000 Steps) (loss=1.87595): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (18500 / 30000 Steps) (loss=1.63074):  83%|████████▎ | 20/24 [00:20<00:03,  1.11it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (18480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (18480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:19,  3.92s/it]\u001b[A\n",
+      "Validate (18480 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:19,  3.92s/it]\u001b[A\n",
+      "Validate (18480 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.59s/it]\u001b[A\n",
+      "Validate (18480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.59s/it]\u001b[A\n",
+      "Validate (18480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.10s/it]\u001b[A\n",
+      "Validate (18480 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.10s/it]\u001b[A\n",
+      "Validate (18480 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.07s/it]\u001b[A\n",
+      "Validate (18480 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.07s/it]\u001b[A\n",
+      "Validate (18480 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.74s/it]\u001b[A\n",
+      "Validate (18480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.74s/it]\u001b[A\n",
+      "Validate (18480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.66s/it]\u001b[A\n",
+      "Training (18500 / 30000 Steps) (loss=1.63074):  88%|████████▊ | 21/24 [00:42<00:22,  7.50s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.07472795993089676\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (18503 / 30000 Steps) (loss=1.72294): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
+      "Training (18527 / 30000 Steps) (loss=1.85699): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18551 / 30000 Steps) (loss=1.88507): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s] \n",
+      "Training (18575 / 30000 Steps) (loss=2.15083): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (18599 / 30000 Steps) (loss=1.70498): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (18623 / 30000 Steps) (loss=1.96934): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18647 / 30000 Steps) (loss=1.79790): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s] \n",
+      "Training (18671 / 30000 Steps) (loss=1.71376): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18695 / 30000 Steps) (loss=2.13562): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (18719 / 30000 Steps) (loss=1.73023): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18743 / 30000 Steps) (loss=1.96963): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s] \n",
+      "Training (18767 / 30000 Steps) (loss=2.29220): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (18791 / 30000 Steps) (loss=1.87343): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (18815 / 30000 Steps) (loss=1.95431): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (18839 / 30000 Steps) (loss=1.71830): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s] \n",
+      "Training (18863 / 30000 Steps) (loss=2.06890): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18887 / 30000 Steps) (loss=1.74112): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (18911 / 30000 Steps) (loss=1.82975): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (18935 / 30000 Steps) (loss=1.69898): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (18959 / 30000 Steps) (loss=1.77620): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (18983 / 30000 Steps) (loss=2.40519): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (19000 / 30000 Steps) (loss=1.78521):  67%|██████▋   | 16/24 [00:16<00:07,  1.12it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (18984 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (18984 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.03s/it]\u001b[A\n",
+      "Validate (18984 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.03s/it]\u001b[A\n",
+      "Validate (18984 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (18984 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
+      "Validate (18984 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (18984 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
+      "Validate (18984 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (18984 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
+      "Validate (18984 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (18984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
+      "Validate (18984 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
+      "Training (19000 / 30000 Steps) (loss=1.78521):  71%|███████   | 17/24 [00:38<00:52,  7.55s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.08991578966379166\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (19007 / 30000 Steps) (loss=1.66867): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
+      "Training (19031 / 30000 Steps) (loss=2.67283): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19055 / 30000 Steps) (loss=1.83468): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (19079 / 30000 Steps) (loss=2.33405): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19103 / 30000 Steps) (loss=1.68956): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (19127 / 30000 Steps) (loss=2.36081): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (19151 / 30000 Steps) (loss=2.13139): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (19175 / 30000 Steps) (loss=2.03213): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (19199 / 30000 Steps) (loss=1.69451): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (19223 / 30000 Steps) (loss=2.16839): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19247 / 30000 Steps) (loss=1.87738): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
+      "Training (19271 / 30000 Steps) (loss=1.85485): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (19295 / 30000 Steps) (loss=3.11746): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (19319 / 30000 Steps) (loss=1.84075): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
+      "Training (19343 / 30000 Steps) (loss=1.92438): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (19367 / 30000 Steps) (loss=2.05224): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (19391 / 30000 Steps) (loss=1.60599): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19415 / 30000 Steps) (loss=2.37670): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (19439 / 30000 Steps) (loss=1.68553): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19463 / 30000 Steps) (loss=1.90457): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19487 / 30000 Steps) (loss=1.74032): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19500 / 30000 Steps) (loss=1.67537):  50%|█████     | 12/24 [00:13<00:10,  1.11it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (19488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (19488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.14s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.14s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.67s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.67s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.15s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.15s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.11s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.11s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.77s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.77s/it]\u001b[A\n",
+      "Validate (19488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.70s/it]\u001b[A\n",
+      "Training (19500 / 30000 Steps) (loss=1.67537):  54%|█████▍    | 13/24 [00:35<01:23,  7.63s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.07180359959602356\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (19511 / 30000 Steps) (loss=1.92059): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it] \n",
+      "Training (19535 / 30000 Steps) (loss=2.07178): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (19559 / 30000 Steps) (loss=2.13432): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (19583 / 30000 Steps) (loss=2.24247): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (19607 / 30000 Steps) (loss=2.14878): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (19631 / 30000 Steps) (loss=1.80989): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (19655 / 30000 Steps) (loss=2.33858): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19679 / 30000 Steps) (loss=1.78372): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (19703 / 30000 Steps) (loss=1.82741): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19727 / 30000 Steps) (loss=2.23388): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (19751 / 30000 Steps) (loss=2.10488): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (19775 / 30000 Steps) (loss=1.85092): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19799 / 30000 Steps) (loss=2.06998): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19823 / 30000 Steps) (loss=2.44791): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (19847 / 30000 Steps) (loss=1.93086): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (19871 / 30000 Steps) (loss=1.82344): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19895 / 30000 Steps) (loss=2.08381): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (19919 / 30000 Steps) (loss=2.62876): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (19943 / 30000 Steps) (loss=1.78375): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (19967 / 30000 Steps) (loss=1.94173): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (19991 / 30000 Steps) (loss=1.64909): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (20000 / 30000 Steps) (loss=1.81226):  33%|███▎      | 8/24 [00:09<00:14,  1.07it/s]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (19992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
+      "Validate (19992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:19,  4.00s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:19,  4.00s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.61s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.61s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.11s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.11s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.07s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.07s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.74s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.74s/it]\u001b[A\n",
+      "Validate (19992 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.67s/it]\u001b[A\n",
+      "Training (20000 / 30000 Steps) (loss=1.81226):  38%|███▊      | 9/24 [00:31<01:57,  7.80s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.07197052240371704\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (20015 / 30000 Steps) (loss=1.89646): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
+      "Training (20039 / 30000 Steps) (loss=1.93255): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s] \n",
+      "Training (20063 / 30000 Steps) (loss=1.77097): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20087 / 30000 Steps) (loss=1.93809): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (20111 / 30000 Steps) (loss=1.66247): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (20135 / 30000 Steps) (loss=1.84815): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (20159 / 30000 Steps) (loss=1.83373): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20183 / 30000 Steps) (loss=1.77483): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20207 / 30000 Steps) (loss=1.73208): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (20231 / 30000 Steps) (loss=1.79430): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20255 / 30000 Steps) (loss=1.97952): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20279 / 30000 Steps) (loss=2.07711): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20303 / 30000 Steps) (loss=3.14760): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (20327 / 30000 Steps) (loss=1.96483): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20351 / 30000 Steps) (loss=1.94485): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (20375 / 30000 Steps) (loss=2.13720): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (20399 / 30000 Steps) (loss=1.92308): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (20423 / 30000 Steps) (loss=1.77186): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20447 / 30000 Steps) (loss=1.71648): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s] \n",
+      "Training (20471 / 30000 Steps) (loss=4.32078): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (20495 / 30000 Steps) (loss=1.72388): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20500 / 30000 Steps) (loss=1.91721):  17%|█▋        | 4/24 [00:05<00:21,  1.09s/it]\n",
+      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
+      "Validate (20496 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
+      "Validate (20496 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:22,  4.42s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:22,  4.42s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:15,  3.79s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:15,  3.79s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.21s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.21s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.14s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  67%|██████▋   | 4/6 [00:19<00:08,  4.14s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  83%|████████▎ | 5/6 [00:19<00:03,  3.79s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.79s/it]\u001b[A\n",
+      "Validate (20496 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.75s/it]\u001b[A\n",
+      "Training (20500 / 30000 Steps) (loss=1.91721):  21%|██        | 5/24 [00:28<02:53,  9.13s/it]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.08904116600751877\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training (20519 / 30000 Steps) (loss=3.33243): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
+      "Training (20543 / 30000 Steps) (loss=1.97818): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20567 / 30000 Steps) (loss=1.89210): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (20591 / 30000 Steps) (loss=1.64444): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20615 / 30000 Steps) (loss=2.15475): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (20639 / 30000 Steps) (loss=1.87721): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s] \n",
+      "Training (20663 / 30000 Steps) (loss=1.84025): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (20687 / 30000 Steps) (loss=1.59353): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (20711 / 30000 Steps) (loss=1.84514): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20735 / 30000 Steps) (loss=1.75540): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (20759 / 30000 Steps) (loss=1.79015): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (20783 / 30000 Steps) (loss=2.06733): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20807 / 30000 Steps) (loss=2.68113): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
+      "Training (20831 / 30000 Steps) (loss=1.85298): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
+      "Training (20855 / 30000 Steps) (loss=1.60886): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
+      "Training (20879 / 30000 Steps) (loss=1.65883): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (20903 / 30000 Steps) (loss=1.73903): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
+      "Training (20927 / 30000 Steps) (loss=2.25431): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
+      "Training (20951 / 30000 Steps) (loss=2.83517): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
+      "Training (20965 / 30000 Steps) (loss=1.50596):  58%|█████▊    | 14/24 [00:14<00:10,  1.06s/it]\n",
+      "\n",
+      "KeyboardInterrupt\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "def validation(epoch_iterator_val):\n",
+    "    model.eval()\n",
+    "    with torch.no_grad():\n",
+    "        for step, batch in enumerate(epoch_iterator_val):\n",
+    "            val_inputs, val_labels = (batch[\"image\"].cuda(), batch[\"label\"].cuda())\n",
+    "            with torch.cuda.amp.autocast():\n",
+    "                val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)\n",
+    "            val_labels_list = decollate_batch(val_labels)\n",
+    "            val_labels_convert = [\n",
+    "                post_label(val_label_tensor) for val_label_tensor in val_labels_list\n",
+    "            ]\n",
+    "            val_outputs_list = decollate_batch(val_outputs)\n",
+    "            val_output_convert = [\n",
+    "                post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list\n",
+    "            ]\n",
+    "            dice_metric(y_pred=val_output_convert, y=val_labels_convert)\n",
+    "            epoch_iterator_val.set_description(\n",
+    "                \"Validate (%d / %d Steps)\" % (global_step, 10.0)\n",
+    "            )\n",
+    "        mean_dice_val = dice_metric.aggregate().item()\n",
+    "        dice_metric.reset()\n",
+    "    return mean_dice_val\n",
+    "\n",
+    "\n",
+    "def train(global_step, train_loader, dice_val_best, global_step_best):\n",
+    "    model.train()\n",
+    "    epoch_loss = 0\n",
+    "    step = 0\n",
+    "    epoch_iterator = tqdm(\n",
+    "        train_loader, desc=\"Training (X / X Steps) (loss=X.X)\", dynamic_ncols=True\n",
+    "    )\n",
+    "    for step, batch in enumerate(epoch_iterator):\n",
+    "        step += 1\n",
+    "        x, y = (batch[\"image\"].cuda(), batch[\"label\"].cuda())\n",
+    "        with torch.cuda.amp.autocast():\n",
+    "            logit_map = model(x)\n",
+    "            loss = loss_function(logit_map, y)\n",
+    "        loss.backward()\n",
+    "        epoch_loss += loss.item()\n",
+    "        optimizer.step()\n",
+    "        optimizer.zero_grad()\n",
+    "        epoch_iterator.set_description(\n",
+    "            \"Training (%d / %d Steps) (loss=%2.5f)\"\n",
+    "            % (global_step, max_iterations, loss)\n",
+    "        )\n",
+    "        if (\n",
+    "            global_step % eval_num == 0 and global_step != 0\n",
+    "        ) or global_step == max_iterations:\n",
+    "            epoch_iterator_val = tqdm(\n",
+    "                val_loader, desc=\"Validate (X / X Steps) (dice=X.X)\", dynamic_ncols=True\n",
+    "            )\n",
+    "            dice_val = validation(epoch_iterator_val)\n",
+    "            epoch_loss /= step\n",
+    "            epoch_loss_values.append(epoch_loss)\n",
+    "            metric_values.append(dice_val)\n",
+    "            if dice_val > dice_val_best:\n",
+    "                dice_val_best = dice_val\n",
+    "                global_step_best = global_step\n",
+    "                torch.save(\n",
+    "                    model.state_dict(), os.path.join(root_dir, \"best_metric_model.pth\")\n",
+    "                )\n",
+    "                print(\n",
+    "                    \"Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}\".format(\n",
+    "                        dice_val_best, dice_val\n",
+    "                    )\n",
+    "                )\n",
+    "            else:\n",
+    "                print(\n",
+    "                    \"Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}\".format(\n",
+    "                        dice_val_best, dice_val\n",
+    "                    )\n",
+    "                )\n",
+    "        global_step += 1\n",
+    "    return global_step, dice_val_best, global_step_best\n",
+    "\n",
+    "\n",
+    "max_iterations = 30000\n",
+    "eval_num = 500\n",
+    "post_label = AsDiscrete(to_onehot=14)\n",
+    "post_pred = AsDiscrete(argmax=True, to_onehot=14)\n",
+    "dice_metric = DiceMetric(include_background=True, reduction=\"mean\", get_not_nans=False)\n",
+    "global_step = 0\n",
+    "dice_val_best = 0.0\n",
+    "global_step_best = 0\n",
+    "epoch_loss_values = []\n",
+    "metric_values = []\n",
+    "begin = time.time()\n",
+    "while global_step < max_iterations:\n",
+    "    global_step, dice_val_best, global_step_best = train(\n",
+    "        global_step, train_loader, dice_val_best, global_step_best\n",
+    "    )\n",
+    "end = time.time()\n",
+    "print(f\"Total train time: {end - begin:.2f} seconds\")\n",
+    "# model.load_state_dict(torch.load(os.path.join(root_dir, \"best_metric_model.pth\")))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "train completed, best_metric: 0.8452 at iteration: 28500\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(\n",
+    "    f\"train completed, best_metric: {dice_val_best:.4f} \"\n",
+    "    f\"at iteration: {global_step_best}\"\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 35,
+   "metadata": {
+    "scrolled": true
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Fri Jul 15 04:29:00 2022       \n",
+      "+-----------------------------------------------------------------------------+\n",
+      "| NVIDIA-SMI 450.119.04   Driver Version: 450.119.04   CUDA Version: 11.6     |\n",
+      "|-------------------------------+----------------------+----------------------+\n",
+      "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
+      "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
+      "|                               |                      |               MIG M. |\n",
+      "|===============================+======================+======================|\n",
+      "|   0  Tesla V100-SXM2...  On   | 00000000:89:00.0 Off |                    0 |\n",
+      "| N/A   38C    P0    48W / 163W |  32494MiB / 32510MiB |      0%      Default |\n",
+      "|                               |                      |                  N/A |\n",
+      "+-------------------------------+----------------------+----------------------+\n",
+      "                                                                               \n",
+      "+-----------------------------------------------------------------------------+\n",
+      "| Processes:                                                                  |\n",
+      "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
+      "|        ID   ID                                                   Usage      |\n",
+      "|=============================================================================|\n",
+      "+-----------------------------------------------------------------------------+\n"
+     ]
+    }
+   ],
+   "source": [
+    "torch.cuda.empty_cache()\n",
+    "\n",
+    "!nvidia-smi"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "'Tesla V100-SXM2-32GB-LS'"
+      ]
+     },
+     "execution_count": 11,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "torch.cuda.get_device_name(0)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Plot the loss and metric"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAskAAAGDCAYAAADZHo16AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAABw4ElEQVR4nO3dd5ycdbn//9c1M9tbstlNL5sGSWgBAqEqUpQiYBcUERWxHOyeI55i/Xp+HtuxcVBsiIKA2FBQikpvSSAEkpDey2Z7352dmc/vj/uezWzfTWZ3yr6fj8c+MnPf98z9mUly7zXXXJ/rY845RERERETksECqByAiIiIikm4UJIuIiIiI9KEgWURERESkDwXJIiIiIiJ9KEgWEREREelDQbKIiIiISB8KkiXpzKzVzBakehwiInJ0zMyZ2aJUjyOZzGy9mZ2X6nFI+lOQnGXMbKeZXejfvs7Mnhzj8z1qZtcnbnPOFTvnto/xORvMLG+szjFezOxLZvbrVI9DRLKTmf3NzL4ywPYrzeygmYWO4rkf9YPok/ps/4O//bwjfe4jHE+Vf95W/6fazP5iZhclHuecO8459+h4jk0yk4JkGdTRXDzHiplVAecCDrhiDJ4/7V6ziMhR+CVwjZlZn+3vAe5wzkWO8vk3A9fG75jZFOBMoOYon/doTHLOFQMnAQ8DfzCz61I4HslQCpKzlJktBX4EnOl/om70t+eZ2bfMbLf/KftHZlbg7zvPzPaa2efM7CDwCzOb7H8Sr/Gzt38xs9n+8V/DC1h/6J/jh/72nq/nzKzMzG73H7/LzP7TzAL+vuvM7El/PA1mtsPMLhnmpV0LPAvcBrw34TU1mtnxCa+/0sw6zGyqf/+NZrbWP+5pMzsx4did/mteB7SZWcjMbjKzbWbWYmYbzOzNCccHzezbZlbrj/lG/zWHEl7zz8zsgJntM7P/Z2bBI/g7vML/WrDRz9gsTdj3Of+5W8xsk5ld4G8/3cxWm1mz//f7ndGeV0Syyh+BKXjXagDMbDLwRuB2/5rxjH+dOWBmPzSz3FE8/x3AOxOucVcDfwDCCecLJFxT68zsHjMrT9j/Wz+r3WRmj5vZcQn7bjOzm83sfv9695yZLRzJwJxzB51z3wO+BPxPwu+exG9cg2b27wnX+zVmNsfft8TMHjazev86+45RvC+SBRQkZynn3Ebgw8AzfvnDJH/X14FjgOXAImAW8IWEh04HyoF5wA14/0Z+4d+fC3QAP/TP8R/AE8CN/jluHGAoPwDKgAXAa/GC3Pcl7F8JbAIqgG8APzPrl/FIdC3eRfkO4A1mNs051wX8Hu/iHPcO4DHn3CEzOxn4OfAhvF8WPwbus97lGlcDl+FlICLANrxfKmXAl4Ffm9kM/9gPApfgvYenAG/qM8bbgAje+3sy8HrgekbBzI4BfgN8EqgEHgD+bGa5ZnYscCNwmnOuBHgDsNN/6PeA7znnSoGFwD2jOa+IZBfnXAfedeDahM3vAF51zr0ERIFP4V2DzwQuAD46ilPsBzbgXefwz3N7n2M+hnedfC0wE2gAbk7Y/1dgMTAVeAHv+p7oKrzr8GRgK/C1UYwPvN8PU4FjB9j3abzr/6VAKfB+oN3MivCy0Hf6j70K+D8zWzbKc0smc87pJ4t+8IKlC/3b1wFPJuwzoA1YmLDtTGCHf/s8vE//+UM8/3KgIeH+o8D1fY5xeAFi0H++ZQn7PgQ8mjC+rQn7Cv3HTh/k3OcA3UCFf/9V4FP+7QuBbQnHPgVc69++Bfhqn+faBLw24T17/zDv61rgSv/2P4APJey70B93CJgGdAEFCfuvBv45yPN+Cfj1ANv/C7gn4X4A2Of/HS0CDvnnzenzuMfxfplUpPrfon70o5/0+PGvnY3xa7t/ffzUIMd+EvhDwn0HLBrk2EfxEgDX4H2oXwJs9vftBc7zb28ELkh43Az/Wh4a4Dkn+ecs8+/fBvw0Yf+leAH+QOOpil+L+2zP97ef7d/fyeHfk5vi1/Y+j3kn8ESfbT8Gvpjqv0/9jN+PMskTSyVeILrG/2qtEfibvz2uxjnXGb9jZoVm9mO/VKIZLwibNMLygQogB9iVsG0XXvY67mD8hnOu3b9ZPMjzvRd4yDlX69+/098G8E+g0MxWmle3vBzvKz/wsuCfib9m/3XPwctoxO1JPJGZXZtQntEIHO+/HvzH7RnksfP813wg4bE/xstEjMZMEt4351zMP88s59xWvF9kXwIOmdldZhZ/LR/A+6bgVTNbZWZvHOV5RSTLOOeeBGqBN/mlCqfjXT8xs2PMK6M76F/j/5vD17qR+j1wPt43XL8aYP88vLrg+DVxI14Ge5pf7vB1v9yhmcPfiiWO4WDC7XYG/x0xmPjvnPoB9s3B++ZwoDGv7PN7491437bKBKFJStnN9blfi1cucZxzbt8IH/MZvK+oVjrnDprZcuBFvKz0QMf3PV833sVmg79tLl5GdFTMq5t+BxA0r14aIA8vYD/JOfeSmd2Dl7WtBv7inGvxj9sDfM05N9RXdD2vw8zmAT/B+9rxGedc1MzWcvg1HwBmJzx2TsLtPXiZ5Ap3dBNi9gMnJIzJ/PPsA3DO3QncaWaleEH4/wDvcc5tAa72a+/eAtxrZlOcc21HMRYRyXy345VCHAs86Jyr9rffgndNv9o512JmnwTeNponds61m9lfgY/glXn1tQfv27qn+u4ws/cAV+J9M7YTr8StgcPX22R4M963b5sGGdtC4JUBtj/mnLuo/0NkolAmObtVA7PjkzD8bORPgP+1wxPaZpnZG4Z4jhK8wLrRn2jxxQHOMWBPZOdcFK8W7mtmVuIHn58GjqTl2ZvwMg/L8LLEy4GleDXR8Vq7O/G+Inu3fzvuJ8CH/SyzmVmRmV1mZiWDnKsIL2iuATCz9+FlkuPuAT7hv3eTgM8lvOYDwEPAt82s1J+wstDMXjvEawuYWX7CT55/jsvM7AIzy8H7sNIFPG1mx5rZ+f5xnXh/PzF/rNeYWaX/d93oP39siHOLyMRwO14g+kG8jhdxJUAz0GpmS/AC3SPx73glbDsH2PcjvN8D86BnYvWVCefvAurwvun87yM8fz9mNs3MbsT7vfV5/7rY10+Br5rZYv/3w4nmdej4C3CMmb3HzHL8n9MsYQK1ZD8FydntH8B64KCZxUsUPoc38eFZ/6utRxh4MkPcd4ECvKzws3jlGYm+B7zNvO4U3x/g8R/Dq4PeDjyJF7z+/Ahey3uBXzjndjtvxvJB59xBvEmE7zazkHPuOf9cM/EmggDgnFuN94vhh3gZiq149dADcs5tAL4NPIP3IeAEvBq+uJ/gBcLr8DIwD+BN1Iv6+68FcvGy5w3AvXg1eIO5Gi/Qjf9sc85twqvz+wHee385cLlzLoyXQf+6v/0gXinH5/3nuhhYb2ateH83Vzlv4o6ITGB+8Po0XhLgvoRdnwXeBbTgXdvuPsLn3++XdQzke/45HzKzFrzfJSv9fbfjlZbtw7tmPnsk5++j0czagJfxapjf7pwb7PfOd/CSEg/hfVj4Gd6ckha8yYhX4X2zdxDvG7uM788vI2fODfVtuYgMx7y2dT9yzs1L9VhEREQkOZRJFhklMysws0vN66c8C++rvD8M9zgRERHJHMoki4ySmRUCj+G1O+oA7gc+4ZxrTunAREREJGkUJIuIiIiI9KFyCxERERGRPhQki4iIiIj0kbLFRCoqKlxVVVWqTi8iclTWrFlT65yrHP7I7KBrtohksiO5ZqcsSK6qqmL16tWpOr2IyFExs13DH5U9dM0WkUx2JNdslVuIiIiIiPShIFlEREREpA8FySIiIiIifShIFhERERHpQ0GyiIiIiEgfCpJFRERERPpQkCwiIiIi0oeCZBERERGRPhQki4iIiIj0oSBZRERERKQPBckiIiIiIn1kVJD88t4mVu+sT/UwRERERCRJnHNsPdSKcy7VQ+klo4Lkbz60ia/evzHVwxAREREZFzUtXTy8oZrN1S10dkdTPZwx8bMnd3Dhdx7jtqd3pnoovYRSPYDRKM0Psbe+PdXDEBERkQmoszvKH1/cR3VzF00d3TR3dtPU0Y1zsKJqMmcumMLxs8oIBiwp54tEY1x/+2pe2tMIgBnMLCugqqKQy0+cyTtPm4NZcs6VKk0d3fzgH1vJDQb46l82ML+iiPOOnZrqYQEZFiSX5OfQ3BlJ9TBEREQkiZxzaR/sbdjfzCfvfpHN1a0AFOUGKSvIobQgh3A0xiMbqwEoyQ+xcn45V502lwuXTTuqc/70yR28tKeR/7xsKZUleeysbWdnXRsb9jdz0+9f5qW9jXz5iuPJDWVUYUAvP35sG00d3fz2w2fyhT+t52N3vsjvP3oWi6eV9Dt2/f4m2sNRTqsqH5exZVSQXJofoqWzO9XDEBERkSS576X9fOFPr1BelMtp88pZUTWZ06rKmTelcNSBczTmWLungeK8HBZUFpETHFnw6JyjpqWL3FCASYW5vfbFYo6fPrmdbz24mbLCHH5x3Wmcu7iCUJ/nPtTSybPb63lmWy2Pbarho3e8wD8++1pmTy4c1WuI23qohe88vJmLj5vOB86Z3+u9iMUc3354Ezf/cxvbDrVxyzWnMKU4b8Sv9Ug+kGyvaaWmpYuWzggtXd20dEZobO/mYHMnB5v8n+ZOivNCvP3U2bzjtDlMK80f8jmrmzv5+VM7uHL5TE6rKuen713BlT98ivf/chV/+pdzKC/y/i5qW7v49kObuGvVHk6ZO5nffeSsUY//SGRUkFySH6IrEiMciWX0pyYREZF09deXD7C3oYM3nzKLihEGXkciGnN888FN/OixbSyfM4mK4lz+tv4gd6/eA0DVlELuuuFMppcNHWgB7Kht4941e/j9C/s40NQJQE7QWFhZzLHTS1hYWUwwYMRijpiDqHO0dUXYVdfO7vo2dte309kdA2BBZRGnzp3MqfMms2hqMd9+aDPPbK/jDcdN4/97y4k9gVtfU0vyueKkmVxx0kz2N3Zw3rce5TsPbeY771x+RO/NZ3+7jsLcIF990/H9gtpAwPjXNyzhmGkl/Nu967jih0/xk2tXsGxm6aDPWd3cyX/98RVe3tfEHz569ojeV4CNB5r5+l9f5bHNNQPun1KUy7TSfGaU5bN87iR21rbx7Yc3892/b+H8JVN51+lzec0xlQOWoHz3kS1EY47PXHQsALMmFfCTa0/lnbc+y4d/tYZfvO807nhuFz/4+1Y6uqO8/+z5fPyCxSMadzJkWJCcA0BLZ/eIPzGJiIjIyOyqa+MTd68lHInxjQdf5fXHTefdp8/ljAVTCIyyzrY7GuN3a/ayv6mT846tZPnsST3P0dTRzSfuepFHN9Xw7pVz+eLlx5EbChCLObbVtPLsjnq++pcNfONvrw4ZZP5z0yFu+ec2nt9ZT8DgNcdUctMlS3AOXj3YwqaDzazaUc+f1u7v99j8nABzywuZW17EuYsrmTelkNauCC/sauCRjdX8ds1ewCur+MbbTuTtp84ecQZ25qQC3ndWFbc+sZ3rz10wZPA6kJ89uZ21exr53lXLqSwZPN65cvks5lcUccPta3jbj57mhtcs4K2nzGZO+eHstXOOu1ft4WsPbKQ76n0Q+NTda/n19SuHrJ3e19jBtx/axB9e3Edpfg6fu3gJJ80uozg/REl+DiX5IUryQ+SFgv0eu7O2jbtW7eHeNXt4eEM1J8+dxM3vOoWZkwp6jtlW08o9q/fwnjPmMXfK4fGePHcy33zbiXzirrWc/rVHaAtHed2xlfznG5exsLJ4VO/j0cqoILm0wBtuc2dEQbKIiEgSOef44n3ryQkYt31wJY9sOMTvXtjL/esOUDWlkGvOmMfbV8yhrCBnyOeJxRx/Xref7zy8mV113mT77/99CxXFeVywZCpnLCzn+3/fyp76dr725uN598p5PY8NBIzF00pYPK2EA40d/N+j27jmzHmcMndyv/Ns2N/MDbevZkZZAf928bG85eTZg2ZH410hggEjYEbAGDLgdc6xo7aNl/c1ccrcyb2CzpH66HmLuGvVHv7nb6/yy/ef3m//oeZO/uOPr7Cwspg3njiD42aWYmZsPdTKtx7azOuXTeOKk2YOe54TZ0/ivhvP5t9+t47vPrKF7z6yhTMXTOFtp87muFmlfPm+DTyzvY4zF0zh6289gee21/Nvv1vHrY9v5yPnLez3fJFojP99ZDM/eWIHADecu4CPnreIssKh/94TVVUUcdMlS/j0Rcfwp7X7+NJ963njD57ke1ct59zFlQB868FN5IcC3Hj+on6Pv3L5LPY3dvLXVw7w6YuOSdlEPktVT7oVK1a41atXj+oxj2yo5vrbV3PfjWdz4uxJYzMwEZERMLM1zrkVqR7HeDmSa7akl62HWnlhVwNXnjxzwOzf3145yId/vYb/vGwp15+7APCCywdePsAdz+1mza4GCnODvOWUWVx3VhWLpvaeWNUdjfHkllq+8eAmNh5oZsn0Ej538RJOnjuJxzbX8PCGah7bVENLV4QpRbnccs2pnD5/8AlYrV0Rzv/Wo8ycVMDvP3JWr0x2Z3eUN938FLWtYR785Llpmzi79fFt/PcDr3Ln9Ss5a1FFz/bG9jDv/PGz7KhrIxZzRGKOueWFXHbiDJ7eVseuujYe+tRrmFoyspKIuD317fzhxX3cu2Yvu/1uYCV5If7jsqU9nTCcc9x454s8uP4g937kLJbPmdTz+PZwhI/d+SJ/f/UQbzl5Fp95w7HMSsj+HqltNa185Ndr2HKolU9deAxnL6rgrbc8zScvXMwnLzzmqJ9/JI7kmp1RQfJz2+t4563Pcsf1Kzk74R+biMh4U5As4+HF3Q3c/M9tvO/sqiP6vdfQFubP6/bzuzV7eWlvEwDnL5nKLdec0itQbg9HuPDbj1GSn8NfPn7OgBPeXtnXxG1P7+S+tfsJR2OcXlVOKGjUtnZR09JFQ7s3sX5ueSGfef0xXH7izH4lGuFIjJf2NjK/omhE9c73rtnLZ3/7Et95x0m85ZTZPdv/+4GN3Pr4dn5x3Wm8bkl6tAsbSGd3lPO/9SgVJXn88aNnEwgYrV0R3v3T59h4oJnbrjuNpTNKeWjDQf6y7gBPb6sjGnN8953LedPJs474vM45Vu9qYPXOBt588qx+Gfam9m4u/f4ThILG/R8/l+K8EDUtXVz/y1W8vK+JL195PO85Y94gz35k2sMR/uMPr/CHF/eRnxOgOC/Eo//6OorzxqeoIeuD5A37m7n0+0/wo2tO4eLjZ4zRyEREhqcgWcZaU3s3l3zvcfb7E9EuWDKVz1+6lEVTD9dlRqIxnt1ez9/WH+Cgf1xcR3eU53fU0x11LJ1RyltP8YKu/3f/Rl53bCW3XHMq+TleoPw/f3uVWx7dxj0fOnPI7C54nQbuen439798kIKcAJUleVQU51FZksf8iiIuOX5G0ibXx2KON//fUxxs7uQfnzmPorwQz2yr410/fZZ3nT6Xr735hKScZyzFA/0fXH0yFy2bxvt+sYrnd9bzo2tO5aI+LeLq28LsqG3llLmTx7wl3vM76rnq1md488mz+ZfXLeS9v3iempYufnD1Kf3GlSzOOe54bjdf/csGvnzFcVx1+twxOc9Asj5I3lPfzrnf+CffeNuJvGPFnDEamYjI8BQky9EIR2L8z99e5cTZZVy5vH/G0DnHR+94gYc3VPObG85gza4Gbv7HVtq7o1yzci7nLZnKQ+ureXD9QerbwhTmBqmaUkRiXBUw4/T55bz1lNm9Jo795vndfP73L/PaYyr58XtOZW9DB5d873EuP2km33nH8nF49aOzZlcDb73laW583SJueO0CLvnuE+QEjQc+cS6Fuek/tSoac1z2/SdoD0c5Zloxf3/1EP/7jqPLFCfLdx7ezPf/voWi3CD5OUF+dt1pvcovxkp3NDbi9nzJciTX7PT/15Wg1J8s0NyhXskiIpKZItEYn7jrRf76ykHAK4m47uz5vY65a9Ue/vrKQW66ZAmnVZVzWlU5bz91Nt99ZAu/fm43v3xmF4W5QS5YOo3LTpjOecdO7ckKD+fq0+cSNONzv1/HB29fTSTqKMgJ8vlLlib9tSbDqfMmc+Xymdz6xHY2HGjmYHMn9374zIwIkMGbLPi5i5fwvttWsbu+na++6fi0CJABPn7+Ip7fUcehli5+cd1pzJtSNC7nHe8A+Uhlxr8wX7xupUWr7omISAaKxhyfuuelngD4xd0NfOnPG2gLR/mX13mz/LceauHLf17POYsquMGfQAcwpTiPr77peK47u4rdde2cuXDKiAPjvt5x2hww+Nzv1uEcfPXK44ZsNZZqn7t4CQ+uP8g/Xj3Exy9YzMkDdLtIZ+cdW8n7z57P/MqipNf6Ho1QMMAd158BkLSltLNJRgXJwYBRnBdSkCwiIhknFnP8273r+PNL+/ncxUv48GsXEonG+OxvX+KbD26itSvCJy5YzMd+s5bC3BDfecdJA/YmXlhZnJR+se9YMYeCnCDP7ajjXSvTJ3AbyMxJBXzp8uN4cmstHxugZVi6MzO+cPmyVA9jQAqOB5dRQTJ4q+5paWoRERlrq3bWEwrYkFlL5xyPbqrhjAVTKMgdPKsbizn+/Q8v87sX9vKpC4/p6U8bCgb4zjuWU5Ab4pZHt/Hg+oNsr2nj59etYOowS/omw+UnzeTyEfTiTQdXnT53XCd6iWRGUUiCkvwQzQqSRURkjH32ty/x5T9vGPKYZ7bV8b7bVnHFD59kc3XLgMfUt4X55N1ruWvVHv7ldQv5+AW9M6GBgPHfbz6e68+Zz/aaNq47q4rzl4xNdwERGbmMC5JL83NUbiEiMgQzu9jMNpnZVjO7aYD9c83sn2b2opmtM7NLUzHOdHaouZNdde1sPdTKUF2gNh70AuPa1i6u+OGT3PX87p7jYzHHPav2cP63H+WBlw/wmYuO4bOvP3bA1l5mxn9ctpS/fOwc/uuN6fm1vMhEk5HlFrWt4VQPQ0QkLZlZELgZuAjYC6wys/ucc4kp0f8E7nHO3WJmy4AHgKpxH2waW72rAfBWfTvQ1MnMQVYd23qohcmFOTz4qdfwqbvXctPvX+apbXW87+wqvv7Aqzy/s57TqibztTefwDHTSgZ8jjgz4/hZZUl/LSJyZDIwSM5hR21bqochIpKuTge2Oue2A5jZXcCVQGKQ7IB449wyYP+4jjADrNpZ33N7y6HWQYPkzdWtLJ5WwtSSfG5//0pueXQr33l4M39+aT+TCnP4xltP5G2nzh5wAp6IpLcMDJLV3UJEZAizgD0J9/cCK/sc8yXgITP7GFAEXDjQE5nZDcANAHPnTqwJU6t3NnDMtGI2V7eypbqF1x5T2e8Y5xxbqlt6Jr4FA8aN5y/mjAVT+OemQ3zgnAWUF+WO99BFJEkyria5JD+H5s7uIWvERERkSFcDtznnZgOXAr8ys36/D5xztzrnVjjnVlRW9g8Ss1VbV4QNB5p5w3HTKS/KZeuh1gGPq2nporkzwuKpvduxragq51/fsEQBskiGy7ggubQgRHfU0RWJpXooIiLpaB8wJ+H+bH9bog8A9wA4554B8oGKcRldBli7p5FozLGiqpxFU4vZMkiQvLna2z5crbGIZKaMC5JL8v2lqdUGTkRkIKuAxWY238xygauA+/ocsxu4AMDMluIFyTXjOso0tmpnPQGDU+ZOYvHUYjZXtwz47eWWQ15ni0XTjn5hDxFJPxkXJJfma2lqEZHBOOciwI3Ag8BGvC4W683sK2Z2hX/YZ4APmtlLwG+A65xq2Hqs3tnAkumllOTncMy0Elo6Ixxq6ep33ObqViYV5lBZnL7LOYvIkcvIiXugIFlEZDDOuQfw2rolbvtCwu0NwNnjPa5MEInGeGF3A28/dTZAT73xlupWpvVZAW/roRYWTy0esO+xiGS+jMsk95RbdKjcQkREkmvjgRbaw1FWVJUDh0sp4qUVcc45Nle3smiq6pFFslXGBcmlfpCsTLKIiCRbvD/yiqrJAFQW51FWkNNv8l5NaxdNHd0co3pkkayVcUHy4XILZZJFRCS5Vu+qZ9akAmaUeYuHmBmLpxazpbp3Jnmr39lisTLJIlkrg4NkZZJFRCR5nHOs2tnAaX4WOW7xtBI2V7f26nARzywvViZZJGtlXJBclBvCTJlkERFJrt317dS0dPXUI8ctnlpMU0c3ta3hnm2bq1sozQ8xtUSdLUSyVcYFyYGAUZwXolmZZBERSaJVOxsAOK1vkDzA5L0th1pZPK1EnS1EsljGBcngTd7TYiIiIpJMq3fWU5of6rfMdLzuOHF56q2HWjVpTyTLZWSQXJIfUk2yiIgk1epdDayoKicQ6J0dnlaaR0l+iM3+5L3a1i7q28Jq/yaS5TIySC7Nz1FNsoiIJE19W5ith1p7Wr8lOtzhwsskb+npbKFMskg2GzZINrOfm9khM3tlmONOM7OImb0tecMbmDLJIiKSTGt2efXIK+aVD7h/8dSSnnKLrX5t8jHTlEkWyWYjySTfBlw81AFmFgT+B3goCWMaVkl+SDXJIiKSNGt2NZATNE6cXTbg/sXTiqlrC1PX2sXm6lZK8kJMK1VnC5FsNmyQ7Jx7HKgf5rCPAb8DDiVjUMMpLchRJllERJJm08FmFlYWk58THHD/4mmHJ+9tOdTComnF6mwhkuWOuibZzGYBbwZuGcGxN5jZajNbXVNTc8TnjJdbJDZ2FxEROVKbq1t7AuGBxOuPNx9q9TpbaNKeSNZLxsS97wKfc87FhjvQOXerc26Fc25FZWXlEZ+wJD+HaMzR0R094ucQEREBaOuKsK+xg2OGmIg3oyyfotwgz++op7Y1rJX2RCaAUBKeYwVwl/+1UwVwqZlFnHN/TMJzDyhxaerC3GS8BBERmai29iwxPXh22MxYNK2Ef2ysBmCROluIZL2jziQ75+Y756qcc1XAvcBHxzJABi+TDNDcocl7IiJydOL9j4fLDi+eWkxb2PsGU50tRLLfsGlYM/sNcB5QYWZ7gS8COQDOuR+N6egGUepnkrU0tYiIHK2th1rJDQaYV1445HHxFfaK80LMKMsfj6GJSAoNGyQ7564e6ZM55647qtGMUDyTrAVFRETkaG2ubmFBZRGh4NBfrsaXp140VZ0tRCaCDF1x73BNsoiIyNHYXN06ovKJeB2yVtoTmRgyMkg+nElWkCwiIkcu3tliJIHvrEkFnLu4gouWTRuHkYlIqmVka4jSgnhNssotRETkyI2ks0VcIGD86gMrx3pIIpImMjKTXJATJBgw1SSLiMhRiXe2OEZ9j0Wkj4wMks2sZ9U9ERGRI7XlUCu5oQBzh+lsISITT0YGyYCCZBEROWqbq1tYUDF8ZwsRmXgy9qpQkpejcgsRETkqW0bY2UJEJp6MDZJLC0I0dyiTLCIiRybe2UL1yCIykIwNkkvyc9TdQkREjtgWv7PFoqnKJItIfxkcJKsmWUREjpw6W4jIUDI2SC7NV02yiIgcua1+Z4t5U4pSPRQRSUMZGySX5Ido7YoQi7lUD0VERDLQ5uoWFlYWEwxYqociImkoY4Pk0vwcYg7awiq5EBGR0dtS3Tqi5ahFZGLK2CC5JN9bmlp1ySIiMlqt6mwhIsPI4CA5B1CQLCIio7fV72yxWD2SRWQQGRwkxzPJmrwnIpLIzC42s01mttXMbhpg//+a2Vr/Z7OZNaZgmCl1uLOFgmQRGVgo1QM4Uiq3EBHpz8yCwM3ARcBeYJWZ3eec2xA/xjn3qYTjPwacPO4DTbEt1S3khgLMLS9M9VBEJE1lbCa5tMArt9CCIiIivZwObHXObXfOhYG7gCuHOP5q4DfjMrI0suVQqzpbiMiQMjZIjmeSm5VJFhFJNAvYk3B/r7+tHzObB8wH/jHI/hvMbLWZra6pqUn6QFNpS3WrJu2JyJAyNkgu7Zm4p0yyiMgRugq41zkXHWinc+5W59wK59yKysrKcR7a2Dnc2UL1yCIyuIwNkvNCAXKCpppkEZHe9gFzEu7P9rcN5ComYqmFP2lvkXoki8gQMjZINjNKtDS1iEhfq4DFZjbfzHLxAuH7+h5kZkuAycAz4zy+lHtlfzMAS6eXpngkIpLOMjZIBijND9HcoUyyiEiccy4C3Ag8CGwE7nHOrTezr5jZFQmHXgXc5ZxzqRhnKq3ZWU9lSR5zygtSPRQRSWMZ2wIOUCZZRGQAzrkHgAf6bPtCn/tfGs8xpZPVuxpYMW8yZupsISKDy+hMckl+SDXJIiIyYtXNnext6ODUeZNTPRQRSXMKkkVEZMJYvbMBgBVV5SkeiYikuwwPklVuISIiI7dqZz35OQGOm6lJeyIytIwOkkvzc7SYiIiIjNiaXQ0snzOJnGBG//oTkXGQ0VeJkvwQrV0RorEJNzlbRERGqa0rwoYDzayYp1ILERlexgfJ4K2eJCIiMpSX9jQSjTlOrdKkPREZXkYHyVqaWkRERmr1rgbM4JS5CpJFZHgZHSTHM8nqcCEiIsNZvauBY6eVUFaQk+qhiEgGyOggudS/0DV3KJMsIiKDi8YcL+5qUH9kERmxjA6SlUkWEZGR2FzdQktXhBWqRxaREcrwINmvSe5SJllERAa3emc9gDpbiMiIZXiQrEyyiIgMb/WuBqaW5DF7ckGqhyIiGUJBsoiIZL3VOxtYUTUZM0v1UEQkQ2R0kJwXCpIXCtCkiXsiIjKIg02d7GvsUKmFiIxKRgfJANPL8jnQ1JnqYYiISJpavcuvR9akPREZhYwPkueWF7K7vj3VwxARkTS1emcDBTlBls4oTfVQRCSDZHyQPKe8kD0KkkVEZBBrdjWwfM4kcoIZ/ytPRMZRxl8x5kwupL4tTGuXJu+JiEwkz++o5+dP7qA7Ghv0mFU769lwoFmlFiIyahkfJM8tLwRQNllEZAJpD0e48c4X+MpfNvC2Hz3Djtq2Xvudc/z0ie1cdeuzzJlcwDtPm5OikYpIpsqaIFl1ySIiE8etj2/nUEsXH79gMTtr27js+09w96rdOOdo6ezmX+58gf93/0YuWDKV+z52DrMnF6Z6yCKSYUKpHsDRUiZZRGRiqW7u5MePbefSE6bz6YuO4erT5/Dpu1/ic797mUc2HmJbTSu76tr590uX8MFzF6g3sogckYwPkssKcyjJDylIFhGZIL790CYisRifu3gJADPKCrjj+pX85IntfOuhTZQV5HLH9Ss5Y8GUFI9URDLZsEGymf0ceCNwyDl3/AD73w18DjCgBfiIc+6lZA90KGoDJyIyMWzY38xv1+zlA2fPZ96Uop7tgYDxodcu5NITZlCcF2JyUW4KRyki2WAkNcm3ARcPsX8H8Frn3AnAV4FbkzCuUVGQLCKS/Zxz/PcDGykryOFj5y8e8Jg55YUKkEUkKYYNkp1zjwP1Q+x/2jnX4N99FpidpLGN2NzyQvY0dBCLufE+tYiIjJNHN9fw5NZaPn7+YsoKc1I9HBHJcsnubvEB4K9Jfs5hzS4vJByJUdPaNd6nFhGRcRCJxvjv+zdSNaWQa86Yl+rhiMgEkLQg2cxehxckf26IY24ws9VmtrqmpiZZp1YbOBGRLPe39QfZcqiVmy5ZQm4o47uXikgGSMqVxsxOBH4KXOmcqxvsOOfcrc65Fc65FZWVlck4NZAQJNcpSBYRyUYbDzQTDBgXLJ2W6qGIyARx1EGymc0Ffg+8xzm3+eiHNHqzJhVgpkyyiEi22lXXzqxJBeQElUUWkfExkhZwvwHOAyrMbC/wRSAHwDn3I+ALwBTg//yG7RHn3IqxGvBAckMBZpTms6dBQbKISDbaU9/OvClaNU9Exs+wQbJz7uph9l8PXJ+0ER2hOeWFWlBERCRL7apv57ITZqR6GCIygWTN91bqlSwikp2aOrppbO/umX8iIjIesipIrm7uorM7muqhiIiklJldbGabzGyrmd00yDHvMLMNZrbezO4c7zGORnxStsotRGQ8ZU2QPMfPMOxVXXLaufmfW9mwvznVwxCZEMwsCNwMXAIsA642s2V9jlkMfB442zl3HPDJ8R7naOyqbwNgbnnRMEeKiCRP1gXJe+o7UjwSSRSNOb754Cb+sm5/qociMlGcDmx1zm13zoWBu4Ar+xzzQeDm+GqpzrlD4zzGUdnlZ5LnKpMsIuMoa4JkLSiSnsKRWK8/RWTMzQL2JNzf629LdAxwjJk9ZWbPmtnF4za6I7C7rp2K4lyK84aday4ikjRZc8WpKM6lICeoIDnNxGvEuxQki6STELAYr73nbOBxMzvBOdeYeJCZ3QDcADB37txxHuJhu+rbNGlPRMZd1mSSzYw55QUKktNMPDjuimhCpcg42QfMSbg/29+WaC9wn3Ou2zm3A9iMFzT3MlarpI7W7rp25k1RPbKIjK+sCZLBK7lQr+T0Eg+OlUkWGTergMVmNt/McoGrgPv6HPNHvCwyZlaBV36xfRzHOGJdkSgHmjuVSRaRcZdVQXJ8QRHnXKqHIr54cKzWfCLjwzkXAW4EHgQ2Avc459ab2VfM7Ar/sAeBOjPbAPwT+FfnXF1qRjy0vQ0dOKf2byIy/rKmJhm8THJbOEp9W5gpxXmpHo4AXd3xcgtlkkXGi3PuAeCBPtu+kHDbAZ/2f9KaeiSLSKpkVyZ5sjpcpJuecotuBckiMnq76rweyXNUbiEi4yyrguR4D809DeqVnC40cU9Ejsau+nYKc4NU6ttBERlnWRUkxzPJmryXPjRxT0SOxu66duaWF2JmqR6KiEwwWRUkF+QGqSzJ66lhk9RTTbKIHI1d9e3qbCEiKZFVQTLAnMnqlZxOVG4hIkcqFnPsrm/XpD0RSYmsC5LnlhcqSE4jmrgnIkfqUEsX4UiMuVpIRERSICuD5ANNHXRHFZSlg06VW4jIEYp3tpincgsRSYGsC5LnlBcSc7C/Mf07XHR2R/n//rqR9nAk1UMZM4cn7qncQkRGZ1e9eiSLSOpkXZAcn+CRCSUXa3Y18OPHtvP8jvpUD2XMJE7c00qIIjIau+vaCQaMmZMKUj0UEZmAsi5InjXZu5hmQia5rcvLIHeEszfLGi+zcA66owqSRWTkdtW3M2tSATnBrPtVJSIZIOuuPFNL8gE40NSZ4pEMr6PbC47bsjpIjg54W0RkOLvr2tT+TURSJuuC5NxQgIriPKqb0z9IbuvygsaOrK5Jjg14W0RkOLvq23tWUhURGW9ZFyQDzCjLz4hMcnzCXns2Z5K7FSSLyOg1dXTT2N6tzhYikjJZGSRPL8vnYAYEyfFM8oQpt+jO3tcpIskVXzlVnS1EJFWyMkjOmExyd3zinsotREQSxTsUzS3XQiIikhpZGSRPL8unqaM77btGtE+ITLKCZBEZvV313kIiqkkWkVTJziC51OtwcTDNJ++1hbO/BVxnQolFp8otRGSEdte1U1GcS3FeKNVDEZEJKjuD5LJ4G7j07pUczyRn94p7MQpygj23RURGYlddu9q/iUhKZWWQPKPMW1Ak3SfvtXfHg+TszbB2RaKUFniZIE3cE5GR2l3fzrwpqkcWkdTJyiA5Xm6R7pP32rsmRgu40vwc77YyySIyAl2RKPubOpRJFpGUysoguSA3yKTCnLTPJMcn7MWXp85GXZEYpQUKkkVk5PY3duIcCpJFJKWyMkgGL5uc9pnk+MS9LC5D6IpEKc0P9dwWERlOfVsXABUleSkeiYhMZNkbJJflp/3S1PEyi6wut0jMJHcrkywiw2ts7wZgkn/tEBFJhawNkjNhQZGemuRsLrdQTbKIjFJPkFyoIFlEUidrg+TppQXUtnYRTtPALBZzh7tbdEdxzqV4RMnnnKMrEqVE5RYiMgqNHfFMcm6KRyIiE1nWBskz/F7J6Vpy0RmJ4hyUFeTgXHZmWSMxR8xBYW6Q3GAgK1+jiCRfU0c3ZvR8wBYRSYWsDZKnlaX3qnvxOuRKf2JKNna4iAfFeaEgeaGAapJFZESa2sOU5ucQCFiqhyIiE1jWBsnxTHK6toGLr7ZXUex9nZiNk/fiy1Dn5QTIywmo3EJERqSxo1v1yCKSclkbJE9P8yC5zW//VlHsZZKzMUg+nEkOkBcKqtxCREaksb1bnS1EJOWyNkguyQtRlBtM2w4X7f2C5Cwst4hnkuPlFgqSRWQEGju6KSvUpD0RSa2sDZLNjOll+Rxs7kj1UAbUtya5I8szybmhQE/QLCIylOaObsqUSRaRFMvaIBm8kot0zSS39alJbsvmIDknQF6Oyi1EZGQa28MqtxCRlMvuILm0gOo0DZInZrlF9n0QEJHkisUcTZq4JyJpIKuD5Bll+VS3dBGNpd9CHW19yi2yf+KeapJFxouZXWxmm8xsq5ndNMD+68ysxszW+j/Xp2KcA2npihDze8iLiKRSVgfJ08vyicYcta1dqR5KPx0TqrtFkPycIJ3qkywy5swsCNwMXAIsA642s2UDHHq3c265//PTcR3kEJp6lqTWxD0RSa2sDpLjvZLTsS45XpNcXuT9IujIxnKLSEKfZJVbiIyX04GtzrntzrkwcBdwZYrHNGJN/pLUyiSLSKpldZA8rTTeKzn9Oly0hyMU5HgZ1pygZefEPT9znB8Ken2SlUkWGQ+zgD0J9/f62/p6q5mtM7N7zWzOQE9kZjeY2WozW11TUzMWY+2nsSMMoJpkEUm5YYNkM/u5mR0ys1cG2W9m9n2/9m2dmZ2S/GEemXReda8tHKUoLwhAQU4wK1vAdUb6rrinIFkkTfwZqHLOnQg8DPxyoIOcc7c651Y451ZUVlaOy8Aa4+UWyiSLSIqNJJN8G3DxEPsvARb7PzcAtxz9sJKjvCiX3GCAA83pFyR3hKMU5oYAKMwN0daVheUW3X0n7mXfBwGRNLQPSMwMz/a39XDO1Tnn4pM1fgqcOk5jG1ZjvNxCmWQRSbFhg2Tn3ONA/RCHXAnc7jzPApPMbEayBng0ehYUScdMcleEwlwvk1yYF6Q9CxfaSJy4p2WpRcbNKmCxmc03s1zgKuC+xAP6XKOvADaO4/iG1NTulVuoJllEUi2UhOcYrP7tQN8DzewGvGwzc+fOTcKph5euC4q0h6OHg+Tc7Cy3iGeOc/1McjgSwzmHmaV4ZCLZyzkXMbMbgQeBIPBz59x6M/sKsNo5dx/wcTO7AojgJUGuS9mA+2jq6KYgx/tgLSKSSskIkkfMOXcrcCvAihUrxqV58fTSfNbuaRyPU41KWzhCcV6Wl1tEYuQEjWDAyMsJ9GzLz9EvP5Gx5Jx7AHigz7YvJNz+PPD58R7XSDS2ayEREUkPyehuMWz9WyrNKMvnYHMnzqXXgiIdfTPJ2Vhu0R3ryQbF/1TJhYgMpbGjW6UWIpIWkhEk3wdc63e5OANocs71K7VIlell+YQjMRr8GdPpoi0cSZi4F8zSxUSi5IW8f2LxPzV5T0SG0qRMsoikiWHLLczsN8B5QIWZ7QW+COQAOOd+hPeV3qXAVqAdeN9YDfZIHF5QpKNn4Y500N6VmEkO0Z6l5Rb9gmT1ShaRITR2hJlfUZTqYYiIDB8kO+euHma/A/4laSNKsullBYDXK/m4mWUpHs1hbeEIRXkJmeRsLLeIxMjz64/jf6rcQkSG0tTRzaSC9EloiMjEldUr7kF6Lk0djTk6u2N9MslZGCR3q9xCREZHE/dEJF1kfZBcUZxHMGBUp9GCIvFJekUJNcnhaIxINLuyrL0yyaHD3S1ERAbS2R2lKxLTQiIikhayPkgOBoypJXlplUmO1x8XJHS3ALKu5KKzVybZL7dQTbKIDOLwktQqtxCR1Mv6IBlIu1X32vxOFkV5h8stgKwrueg1cS9H5RYiMrSm+JLUagEnImlgQgTJM8ryOdDUkeph9GgPe5nkxBZwiduzhRckq9xCREam0V+SWjXJIpIOJkSQPL20IK0yyfGeyEX9guTsyrJ2RaI9GWQtJiIiw2lUJllE0sjECJLL8mgLR2lKkwVF2vrVJPvlFtkWJHf375PcmWV11yKSPPFrtDLJIpIOJkSQvHzOZAAe3XwoxSPxtPepSS6YCOUWOSq3EJGhNXbEyy00cU9EUm9CBMkr5k1memk+f34pPVbL7ltuEQ+Wsy6THBmou0V2vUYRSZ6mjm6CAaPITxyIiKTShAiSAwHjjSfO4LHNh9Ki5OLwxD2/3CInS8stIrGeDHK+MskiMozG9m4mFeRgZqkeiojIxAiSAS4/aSbdUceD6w+meii0+a3eerpb5GVfuYVzjnAkRr6fQc4NKkgWkaE1dnRrIRERSRsTJkg+cXYZ86YU8ud1+1M9FNrDEcwOZ1ezsbtFPBiOZ5LNjLxQQH2SRWRQTX4mWUQkHUyYINnMuPzEmTy1tZba1q6UjqU9HKUoN9TzlWI825pVQbK/sl68Ftm7HdCKeyIyqMaOsCbtiUjamDBBMnglFzEHf305tRP42sORnuwxeDXThbnBnuWqs0E8YxyfuAeQlxNUuYWIDKqpo1s9kkUkbUyoIPnY6SUcM6045V0u2rqivYJk8Eou2rOo80NPuUVikKxyCxEZQmO7gmQRSR8TKkgGuPzEmTy/s579jalbpro9HO2ZtBdXkBukI5vKLeKZ5Jw+5RbKJIvIACLRGC2dES0kIiJpY+IFySfNBOD+danLJreHIz29keOKckM9K/Flg87ugTLJQdUki8iAmju9658m7olIuphwQXJVRREnzi5LaZeLtsEyydlebpGjcgsRGVhju1bbE5H0MuGCZPBKLtbtbWJnbVtKzt/eFelXk5xtmeTDE/dUbiEiw2vq8BZ6Uk2yiKSLCRkkX3biDAD+kqJs8mA1yVnVAq5Pn2Twyy0UJIvIABrjQbJqkkUkTUzIIHnmpAJOq5rMAy+nZvW9gWqSC7Ot3GLAmuQAXVn0GkUkeZravSBZNckiki4mZJAMcOq8crYcaiEac+N+7oFqkgtzQz3LVWeDeLlFfmJ3i5wgYWWSRWQAqkkWkXQzYYPk+RWFdEcd+xrGtxVcJBojHIkN2Ce5I5xFNcmDZZIVJIvIAOLlFqX5oWGOFBEZHxM4SC4GYEfd+E7eiy8Y0n/inreYiHPjn9keC4NP3MuebLmIJE9TRzcleSFCwQn7a0lE0syEvRpVVRQCsKOmdVzP2+6XVBTl9Z24F8K5w/2FM91gE/ey5fWJSHI1tXdr0p6IpJUJGyRXFudRnBdiZ137uJ63zS+pGKjcArxJfdlAfZJFZDQaO7q12p6IpJUJGySbGVUVhewY517J8Uxy/4l78SA5O4LIeBeL3GDvmuTuqEvJZEkRSW+N7WEmFWjSnoikjwkbJANUTSka/yDZzxQX9cskh/z9yQuSNx1s4YO3r05J9rYrEiMvFMDMerbF65PV4UJE+mrs6NZCIiKSViZ0kLygooi9De3jGrTFg+DCPjXJhX7f5LYklls8saWGhzdUj3sHDzgcJCeK31fJhYj01dyhmmQRSS8TOkiuqigi5mB3/fjVJbcNlkn2+wl3JDGTXNPaBUBL5/jXOXdFouTl9H6N8Ul8agMnMrbM7GIz22RmW83spiGOe6uZOTNbMZ7j68s5R2N7txYSEZG0MqGD5PkVRQDsHMeSi56a5L6Z5DEot6ht8ZrzpyRI7u6fSc73yy261OFCZMyYWRC4GbgEWAZcbWbLBjiuBPgE8Nz4jrC/tnCUSMxp4p6IpBUFyTCudcnxmuTCPlnWeLlFMrtb1PZkkruT9pwj1RWJ9VptDxIzySq3EBlDpwNbnXPbnXNh4C7gygGO+yrwP0DneA5uID2r7WninoikkQkdJE8qzGVSYc64LijS1lOTPFgLuCRmklNdbtGvJtnPJKvcQmQszQL2JNzf62/rYWanAHOcc/cP9URmdoOZrTaz1TU1Nckfqa+x3V9tT+UWIpJGJnSQDF42eUfN+GaSQwHr1RoNxqbcoq7Vy840pyCT3DlAuYUm7omknpkFgO8AnxnuWOfcrc65Fc65FZWVlWM2pmZ/SWqVW4hIOlGQPKWIneOZSe6KUpgb7NUaDRIyyV3Jyfo656hrS3UmuU+5RTxIVk2yyFjaB8xJuD/b3xZXAhwPPGpmO4EzgPtSOXmvUUGyiKQhBckVRRxo6hxxV4muSJRdRxFUd4Sj/RYSAcgJBsgJGu3dycmyNnV00x31Fu1ITZAc67UkNdDT7ULlFiJjahWw2Mzmm1kucBVwX3ync67JOVfhnKtyzlUBzwJXOOdWp2a4h8stVJMsIumkf7Q2wVTFO1zUtbF0RumAxzR1dPPopkM8tL6axzbX0NoV4U//cjYnzZk06vO1hSP96pHjCnNDScskx+uRIUUT91RuIZISzrmImd0IPAgEgZ8759ab2VeA1c65+4Z+hvHX2OFP3FMmWUTSyIQPkhPbwPUNksORGJ+8+0UeWl9NJOaoKM7j8pNm8IcX93HP6j1HFCS3h6MUDZBJBq/kIlk1yTV++zdIw3ILZZJFxpRz7gHggT7bvjDIseeNx5iG0tTeTW4o0K8jjohIKk34IDmeSd4+QBu4p7bV8sDLB7n69Lm8fcVsls+eRCBgdISj/Pml/fzXG5eN+qLe1hXpqT/uqzA3mLRyi3gmuSQvREtXalrA9csk56hPsoj019ShhUREJP1M+Jrk4rwQlSV5Ay4o8vCGagpzg3zx8mWcMncygYA32e6tp86muTPCIxurR32+ju7oEEFy8sst5lcWpU9NssotRGQAje3dKrUQkbQz4YNk8NvA9QmSYzHHwxuqOe/Yyn7Z4rMWVjC9NJ/frdk76nO1dUX6rbYXV5DEcova1i6CAWPO5MIUrbincgsRGZnGjrAm7YlI2lGQzMBt4F7a20hNSxcXLZvW7/hgwHjzKbN4fEsth1pGt1iVV5M8cCa5KJlBckuY8qJcSgtyUrjinhYTEZHhNbZ3ayEREUk7CpLxShJqW8O9Ft14aEM1wYBx/rH9g2SAt54ym2jM8acX94/qXF5N8mAT90JJW5a6trWLiuI8SvND455JjkRjRGKuXyY5J2iYeVlmEZG45g6VW4hI+lGQDFRNOdzhIu7hDdWsnF9O2SAX7kVTi1k+ZxL3rtmLc27E5xqqJrkgNzjifs3D8YLkXEryQ3RFYoTHMXsbzxT3nbhnZuSFAnQqkywiCRraNXFPRNKPgmRgQaUXJMfrkrfXtLL1UCuvH6DUItFbT53NpuoW1u9vHtF5wpEY3VFH0SA1yUW5QdqSFiSHqSzOoyTf+8UzniUXgwXJ3ragMski0qMjHKWjO0p5sWqSRSS9KEgG5pYXYnY4SH54g9e14qLjpg/5uMtPnEFuMMC9I5zAFy+lGDyTHEpKJtk552WSS/IoyfcC8vEsuYh3r8gboD1eXiigmmQR6VHf7vV0Ly9UkCwi6WVEQbKZXWxmm8xsq5ndNMD+uWb2TzN70czWmdmlyR/q2MnPCTKzrKCn3OKhDdUcN7OUWZMKhnzcpMJcLlw2lfte2j+icoZ4lniwxUSKcoOEozG6o0cXRLZ2ReiKxPxyi3gmeRyD5O4hMsk5CpJF5LD6Vj9ILlKQLCLpZdgg2cyCwM3AJcAy4GozW9bnsP8E7nHOnQxcBfxfsgc61uJt4GpaunhhdwOvXzZ0FjnubafOpr4tzKObDg17bIefSS4YoiYZOOoOF7X+L52K4sRMcirKLQbKJAfVJ1lEevRkkhUki0iaGUkm+XRgq3Nuu3MuDNwFXNnnGAfE13QuA0bX8iENVFUUsqO2jUc2VuMcA7Z+G8hrFldSUZzHbU/v5MXdDTS1Dx6MtnX5meS8wRcTAY665CK+kEhikNycinKLAWuSA1pxT0R61Ld51ysFySKSbkayLPUsYE/C/b3Ayj7HfAl4yMw+BhQBFw70RGZ2A3ADwNy5c0c71jE1v6KY5s4Id63aw+zJBSydUTKix4WCAa49cx7feXgzb/6/pwHvYj+/ooh3r5zLW06Z3XNsW09N8iDlFn7w3HaUbeBqWw4HycV5Kcwk5wwSJKvcQkR89W3etUlBsoikm5EEySNxNXCbc+7bZnYm8CszO9451ysacs7dCtwKsGLFipH3TRsH8ysKAXhpTyPvO7sKMxvxYz92/iIuPWEGO2rb2FHbyo7adh7fXMO3HtzEm0+e1fNc7V1D1yQX+BPdkpZJLsklJ+AFqqmpSVa5hYgMrb7NWx20NF8t4EQkvYwkSN4HzEm4P9vflugDwMUAzrlnzCwfqACGL9RNE/Mrintuj7QeOc7MWDS1mEVTiwGvTOPuVbv53O9eZuOBFpbN9CpR2v3WZ4PVJMdbw7V1HV1AW9MaxsybLR7/JJKK7hZ9V9yLb6trG/9lskUkPdW3dTO5MIdAYOSJCRGR8TCSmuRVwGIzm29muXgT8+7rc8xu4AIAM1sK5AM1yRzoWJs9uYBgwJhUmMNpVZOP+vlet2QqAP94tbpnW7sf/A5Wk9wzce8o+wjXtnYxuTCXUDBATjBAQU4wvSbuqSZZRHz1bV0qtRCRtDRskOyciwA3Ag8CG/G6WKw3s6+Y2RX+YZ8BPmhmLwG/Aa5zo1mGLg3kBAOcOLuMK06aSSh49O2jp5bkc9LsMh7ZeDiZHm8BN/iy1Ekqt2jxVtuLKxnnpamHnLiXE1C5hYj0aGjrVpAsImlpRDXJzrkHgAf6bPtCwu0NwNnJHdr4++2HzhxVLfJwLlg6jf99ZDM1LV1UluT1ZJIHW0wkXqt8tOUW3pLUeT33S/JDtHSNXya5s1sT90RkZOraujh2+sgmSouIjCetuJcgFAwQTGJd3AVLp+Ic/NPvodzeHSXXL4EYSLzcouOoyy3CfYLknHGeuBfPJA82cU9Bsoh4GtqVSRaR9KQgeQwtm1HKjLJ8/r7Rq0tu74pQOEg9MhzOJB/9YiL9M8nj2yd5iBX3QoGeIFpEJrZozNHQHtaS1CKSlhQkjyEz4/wlU3liSy2d3VHawtFB27+B1/nB7PAEvyPRHo7QHo5SUXL4l05pfk6KJu5pWWoRGVxTRzfOqUeyiKQnBclj7MKl02gPR3luRz3t4cig9cjgBdUFOcGjyiTXJSxJHVeSH6J1nCfuBQM24ATIvFCQSMwRiSpQFpno4qvtTVaQLCJpSEHyGDtz4RTycwL8fWM17eEohXlDz5UszA31dME4EjX+QiKVfSfujfNiIgNlkeFwdjmsIFlkwot/qJ9SlDfMkSIi409B8hjLzwlyzqJK/r7xEG1dEQpzBs8kg9f5ouMolqVOXJI6riQ/h47uKN3jFJh2RYYPktUrWUQa2r0geXKRVtsTkfSjIHkcXLh0KvsaO9iwv3nQhUTiCnOPrtyiNl5uUdK7TzIwbiUXXZHogJ0tAPL8Dwmd6pUsMuHVtSmTLCLpS0HyODjfX32vLRwddCGRuKMPkr1McuIvnZJ8L0szXiUXXZHYgEtSw/hkknfWtvHpe9Zq0RKRNNfQpkyyiKQvBcnjYGppPifOLgMGX5I6rjA3RPvRlFu0dlFWkENuQrlDPJPcPE4dLrya5EEyyf72sexw8bf1B/n9C/vYUt06ZucQkaNX1xamOC806PVCRCSVFCSPkwuWTAOgIGfsM8lTinvPFI8HyeOVSe6MRAdcbQ8SMsljmOXdXuMFx3sbOsbsHCJy9Brawsoii0jaUpA8Ti5Y6pVcjHlNckvv1fYASvLi5RbjmUkeJEjOiQfJY5dJ3l7TBsC+RgXJIumsri1MueqRRSRNKUgeJ8fNLOW6s6q4YOm0IY8rzAsddSa5sm+QPM6Z5CEn7sXLLcawJnl7rRck71eQLJLWGtrDTFGPZBFJUwqSx4mZ8aUrjmP5nElDHleYEzyqmuSa1i4qBi23GKdM8khawI1RuUVDW5h6fzLQPpVbiKS1+tYwk7UktYikKQXJaaYwN0hHd5RozI36sZ3dUVo6I/3LLVLQ3WLQmuQxLrfYXuvVI+cGA+xvUpAsks7q28P95lCIiKQLBclpZvG0EpyDf/3tS4Mu3fzqwWbe8L+P85PHt/faHs+gVpT0DpJzQwHyQgFautKo3GKMMsnb/HrkFVWTlUkWSWPt4Qid3TFlkkUkbSlITjNvPHEGn339Mfz+xX185I4X6OzuHUw+vKGat/7f02yqbuGWx7b12h/vkdw3kwxeNjktJu6NcZ/k7TVt5ASNlfOnUNcWpuMo6rtFMpWZXWxmm8xsq5ndNMD+D5vZy2a21syeNLNl4z3G+p6FRBQki0h6UpCcZsyMG89fzFeuPI6HN1Tz/ttW0doVwTnH/z26lRt+tZpFU4v59ttPor4tzF9fOdDz2MNBcv9fOqX5IZrHs9xi2JrksQmSt9W0UjWliLlTCgBUciETjpkFgZuBS4BlwNUDBMF3OudOcM4tB74BfGd8R3k4SJ6sIFlE0tTQTXslZa49s4qS/BCf/e063v3T56iaUsif1u7n8pNm8s23nUhuMMDN/9zK7c/s4s0nzwa89m8wWCY5NK7dLfJzBi63iG8fq3KL7TWtLJpazKxJhYA3eW9hZfGYnEskTZ0ObHXObQcws7uAK4EN8QOcc80JxxcBo58EcZTiQXK5gmQRSVPKJKexN588m1vefQobDzTzp7X7+cxFx/D9q5aTnxMkEDCuOWMeL+5u5JV9TYDX2QKgsiR15RbOuZFlkseg3CISjbG7vp0FlcXMnJQPqA1cMm2vaeVXz+xM9TBkeLOAPQn39/rbejGzfzGzbXiZ5I+P09h6KEgWkXSnIDnNvf646fzuw2dx5wdX8rELFmNmPfveeupsCnKC/OqZXYBXblGcFxowi1uSH6J1HDLJ3VGHc5A3SCY5FAwQDNiYlFvsaeigO+pYUFHE9NJ8AjZ+C4p0hKNZX/989+o9/Nef1tPUPj617TK2nHM3O+cWAp8D/nOgY8zsBjNbbWara2pqknp+Bckiku4UJGeAE2aXcdbCin7bywpyeNPJM/nTS/toau+mtjU8YD0yjF+5RadfRjFYJjm+byzKLbYd8tq/LZxaTCgYYHpp/rh1uPj4XS/ysd+8MC7nSpW6Vi+oibfZk7S1D5iTcH+2v20wdwFvGmiHc+5W59wK59yKysrK5I0QL0gOBYzSfFX9iUh6UpCc4d5zRhWd3TF+u2YPtS1dA9Yjw/iVW8TLKIYPkpOfSY4HbwsrvBrkWZMLxiWT7Jzj+R31rNvbNObnSqU6v5wnvuy3pK1VwGIzm29mucBVwH2JB5jZ4oS7lwFbxnF8gBckTy7K7fXtmIhIOtFH+Ay3bGYpK+ZN5tfP7iIUDLCwsmjA40ryQ7SFvUVKgoHk/FLq7I5S1xZm1qSCnm1dPZnkgcst4vvGoiZ5e00bU4pyKSv0Fk+ZNamA1bsakn6evg40ddLU4X0Aae2KUJyXnf+t6tqUSc4EzrmImd0IPAgEgZ8759ab2VeA1c65+4AbzexCoBtoAN473uOsbwtTrh7JIpLGlEnOAu85cx4769rZeqh1yEwykNS65K/+ZQMXf/fxXqUT8QzxYCvuxfeNRbnF9po2FiR8SJg5qYCDTZ1HtHrhaGw8cLhRwM7a7M2y1rYok5wpnHMPOOeOcc4tdM59zd/2BT9Axjn3Cefccc655c651znn1o/3GOvbwqpHFpG0piA5C1x8/PSeWuTBg2Qvu9mcpJKL+rYw967ZS0tnpFeZwUjLLTrHIpNc29qr3dusyQVEYo5DLZ1JP1eiVw+2JIwhOwNI5xy18UyygmRJgvr2MOVaklpE0piC5CyQFwpy1Wlzgf5LUsfFJ8cka/Leb57f3ZM1fm57Xc/2EZdbJDmTHJ+42DeTDIz55L0NB5qZVpqHGezI0gCytStCOBIjNxRgR13bmGfnJfup3EJE0p2C5Czx7jPmsmhqMctnTxpwf3GeV26RjMl73dEYv3pmF+csqmDJ9BKe21Hfs6+n3GKcJ+5t8+tkF1QcziTPjgfJYzx5b+OBZk6aPYmZZQXsyNJ63Xhni+WzJxGOxNR/Wo5KJBqjqaNb5RYiktYUJGeJGWUFPPLp13LC7LIB95ckMZP8t1cOcrC5k/edXcXK+eWs2dVAd9QLeg/XJA+RSc5JfpAcLwEYMJM8hgFdRzjKzto2ls4oZUFlETuytNyirs2rRz5t/mQge8tKZHw0dnTjnHoki0h6U5A8QfQEyV0jyySv3dNIe3jggPrnT+2gakohrzt2KisXTKE9HOVlf9W/ru6R9ElOfrnF9ppWcoLGnPLCnm1FeSEmFeaMabnFpuoWYg6WzihlfkUR22vbcC77ShFq/UzyiqpywHu/RY5UgxYSEZEMoCB5goh3txhJJvmfmw7xppuf4tqfPd9vFbm1exp5cXcj151VRSBgnD7fC5qe2+6VXMQzxPlDdbcIBZLeAm5bTStzywvJCfY+76xJBWNaGhDvbLHMD5JbOiM9rdKySbzcYsn0EkryQ5q8J0elTkGyiGQABckTxEjLLdq6IvznH15hWmkeL+xu4EO/XkM4oTTiF0/toCQvxNtWeAt6VRTnsbCyiOd2eJP3OrtHMnGvf7nFqweb+dr9G454QpjX/q243/aZk8Z2QZGNB5opyg0ye3IB8yu8Uo9sLLmo9RcSmVKUx4LKYvVKlqOiTLKIZAIFyRNEfk6Q3GBg2BZw335oM/saO7j5Xafw/73lBB7fXMMn736RaMxR3dzJ/esO8PYVc3otmLFywRRW72wgEo2NcOJe/3KL7z2yhZ88sYPnEyYBjlQ05thV196rHjlu1qQC9jV0jFkJxMYDzSyZUUogYD2TBrOxw0Vdaxel+SFyQwEWVhQpkyxHRZlkEckE2bk0mAyoJD80ZCZ57Z5GfvH0Dq45Yy4rqspZUVVOS2eE/3f/Rorz1jG1JJ+oc1x3VlWvx62cX86dz+1mw4HmhCB55BP36tvCPLKxGoD7XtrHmQunjOp17W1oJxyN9SxHnWj25ALawlGaOyI9K/Eli3OOVw+0cOXJMwGvL3NO0LJyUlttW7inB/eCyiJ+/+I+2sMRCnN1CZHRq/eD5MlqASciaUyZ5AlkqCC5Oxrjpt+tY1pJPv928ZKe7defu4CPn7+Ie1bv5f8e3cqFS6cxd0phr8eescALap/fUX+4T/IoapL/tHYf3VHHibPLuH/dgVFP6tvmTyJbOLV/Jjne4WJvY/uonnMk9jZ00NIVYemMUgCCAWPelKKsbANX19rFFH/hh3hZi7LJcqTq28KU5HnfTIiIpCtdoSaQkvycQfsk3/r4dl492MJXrjyO0vzeGddPXXQM151VhQOuP2d+v8dOK82nakohz26v7wl+c4PDl1vESyDuXbOX42eV8umLjqG5M8Jjm2pG9bp62r8NkEme5QfJ+xuTv+pefNJePEgGmF+RnW3g6lrDTCnyMsnZXHst46O+TavtiUj6U5A8gZTkh2gdIJO8vaaV7/19C5ccP53XHze9334z44uXL+OZmy5g5YKBSyFWzp/Cqp31dHZHyQ0GCARs0HHkhQLEHERijg37m1m/v5m3nzqHcxZVMKUolz+9tH9Ur2tbTRuTC3OYPEB94+FV944sk/zwhmquuvWZATtkbDzQgpnX8SFuQUURO+vas25Furq2cE8meX5FEWajyySHk9wXWzJbQ3tYpRYikvYUJE8gg5VbfPnPG8gLBfjyFccN+lgzY3pZ/qD7Vy4op6mjm5f2Ng45aQ8Ol2J0RWL8ds0ecoMBrjhpJqFggDeeOINHNlSPamXA7TWtA3a2AKgoziUvFGB/0+gyyc45bv7nVm741Wqe3V7Pz57c0e+YjQeaqZpS1Ksud35FUdatSBeJxmhoDzPFr0nOzwkys6xgxB0udtS2cfyXHuTRTYfGcpiSQbxvJhQki0h6U5A8gQxUbtHc2c0TW2q49sx5TC0dPAgeTjzDvGZXw5D1yHB4Ul9rZ4Q/rd3Phcum9mSBr1g+i65IjIfWV4/43Ntq2lg4QGcL8IL7eIeLkeoIR/n4XWv55oObuOKkmVx6wnTuXrWn33u38WAzS2eU9NqWjaUIDe3e6miVCV+PL6gceYeLhzccJByJ8aPHto3VECXDNLSH1dlCRNKeguQJZKBM8jPb6og5OHdx5VE996xJBcyeXEB31A3Z2QIOLzTy11cOUN8W5u2nzunZd8rcScwpLxi25KKzO8qaXQ389Int1LZ2DZpJBq/kYu8IM7sHmjp4x4+f4S/r9vO5i5fw3Xcu5yOvXURrV4S7V+3pOa61K8KuunaWTC/t9fj5ldkXJPf0SPYzyQALK4vZXtM6otZ6j2+uBeDZ7fWs3980NoOUjOGco65NQbKIpD8FyRNISX4OreEIsYR62ae21lKQE+SUuZOP+vlXzveyySPNJN/x3G6mluRx7uKKnn1mxpUnzeLJLTXUtHT1elw05vjfhzdz2fef4PgvPshbb3ma/3f/RmaU5XPOogoGM9JV91q7Irz55qfZUdvGT69dwUfOW4iZccLsMk6fX84vntpJJOrV1m462H/SHkBlcR7FeaGsCpLjq+0lfj2+oLKItnCUQ33+jvpqD0d4fkc971wxh8LcIL94audYDlUyQHs4SjgSU5AsImlPQfIEUpofwjloDR/OJj+5pZaVC8qT0opp5QJvierhMsnxmuWth1p58ymzCPXphHHl8pnEHNy/7nA2ORyJ8fHfvMj3/r6FkvwQH3rtAn78nlN59vMX8MznL+D4WWWDnm/mpAJqWrp6VgMczB9e2MvB5k5+9t4VXLB0Wq99158zn32NHTzol4FsONAC0K/cwsyYX1GUVb2S69r6Z5LjZSXx9nuDeXZ7HeFojMtPmslbT5nNfWv392SmZWLq6ZGsIFlE0pyC5Amk79LU+xo72F7bNmQWdjTOiGeSRzhxD+hVahG3eFoJS2eU9pRcdISjfPD21dz/8gH+49Kl3HXDmfzrG5bwhuOmDzmZMG7WZK/DxcEhJu855/jlM7s4aXbZgB08Llg6jXlTCvnZk9sBb9JeaX6op8VcIq8NXPb0Sq71M8kVvWqS/dUFh/kw8Phm75uKFVWTue7sKsLRGHc8u3vsBitpLx4ka+KeiKQ7BckTSHGe1/84PgHtqS1ereg5i5MTJM8pL2B6af7wQbKfaT557iQWTR24lvjK5TN5cXcjr+xr4tqfP8fjW2r4+ltO4IOvWTDqccUD2X1DlFw8s62OrYdaufbMqgH3BwPG+8+ezwu7G3lhdwOv+stRm/VvdTe/ooi9DR2jXhQlXdW1dhEKWK/+2TNK88nPCQw7ee+xzTWcsaCc/JwgCyuLOe/YSn793K6seW9k9JRJFpFMoSB5AumbSX5iay2VJXkcO61kqIeNmJnxhcuX8cFzhw5ki/K8cbzt1NmDHnPFSd5Sz2/70dOs3dPID64+matOn3tE4+oJkofocPHLZ3ZSXpTLZSfOGPSYt506m9L8ED99wlt4ZVmfeuS4BZVFOAe765K/yl8q1LV6k6wSe18HAsb8Cm/y3mB217Wzo7aN1x5zeFLo+8+eT01LF/evOzCmY5b0pUyyiGQKBckTyOEguZtYzPH01lrOWVQxYDb0SF16wgwuXDZtyGNOnFXGLe8+hXeu6F9qETdzUgFn+mUPt167gjeeOPOIxzS9LB+zwTPJ+xo7eHhDNe88bQ75OYPXUxflhbh65VweePkg7eFov3rkuGxrA1fX1kVFQj1y3ILKoWuvH9virZz4moQg+dzFFSyaWszPn9oxos4Ykn2USRaRTKEgeQIpyY+XW0TYeLCZurYwZyepHnk0AgHjkhNm9Juw19cP3nUyD3/qtbzu2KlHdb7cUICpJXmDBsl3PLsLgHevHD5Tfd1ZVYT8jGrfzhZxVVkWJNe0Hl5tL9HCiiL21LcPWjrx2KYa5pQX9HxoAO/bhvedXcUr+5pZvathzMYs6au+PUxO0CjJCw1/sIhICo0oSDazi81sk5ltNbObBjnmHWa2wczWm9mdyR2mJEOpn0lu7ozw1Fa/HjkFQfJIVRTnMae8MCnPNdiCIp3dUe5atYcLl05j9uThzzWjrIDLTpxBTtA4ZpAyldL8HCqK8zIqSB4qq1vXOnAmeX5lEbFBykrCkRjPbKvltcdU9vum4i0nz6asIIefD7CK4VhoD0d4aP1B/v0PL/OXdaNb8lySr77VW5I6md9giYiMhWGDZDMLAjcDlwDLgKvNbFmfYxYDnwfOds4dB3wy+UOVo3U4k9zNE1tqWTS1eETdIbLBrMmFbKpu6VdD+8DL3oImg03YG8hXrjieu244c8jSjAUZ1AZu1c56Tv/vvw9aJzzYEsILKrxJl9sGmLy3ZlcDbeEorxlgkZqC3CBXnz6XB9cf5M7nRtfp4kBTB9f89DnufG73kIF9Y3uYXz27i+t+8TzLv/IwN/xqDXc+t5vb1Kc55bSQiIhkipFkkk8HtjrntjvnwsBdwJV9jvkgcLNzrgHAOXcoucOUZMjPCRAKGHWtYVbtrE/rLHKyXbNyLtGY49LvP8FtT+3oWVDll8/sYkFlEWcv6t/2bTBlhTmcOm/oxVe8NnC9g8dHNx3iih8+yaaDLaN/AWPkyS21XPuz56lp6eLRTf3/27aHI3R0R3v1SI5bMMTqgo9triEUMM4a5N/Yjecv4tzFlfz7H17mq3/ZQDQ2fH1yc2c31/18FU9ureXf//AyN/7mRZr7LBXunOPeNXs5/9uP8V9/fIUdtW1cs3Ied16/kitOmjmiRWVkbB1s7pgwH85FJLONJEieBexJuL/X35boGOAYM3vKzJ41s4uTNUBJHjOjJD/Eo5sO0dkd67XSXbZbuWAKD33qNZyxYApf+vMGrvnZczzw8gFe2tPIe8+sSvpXv/Mri6hp6aKlsxvnHD9+bBvvv20V6/Y28ePHtiX1XEPZXN3Cn9buo7Ur0m/fwxuqef9tq5g3pZATZ5fx6gDBe89qewPUJJfk51BZkjdgh4vHNtewomoyxYPUnRbnhfjZe1dw3VlV/OzJHXzw9tUDjjGuKxLlQ7evYVtNK7/6wOn828XH8rdXDnLZ959g7Z5GALYeauGqW5/ls799ifkVRfz5xnN49LPn8YXLl3HWogrmTSnkYHNnz6qJkhr7GzuZOUB/cRGRdJOsmRMhYDFwHjAbeNzMTnDONSYeZGY3ADcAzJ17ZO285OiU5OewraaNUMAGXDQjm00rzecX153G3av28NW/bODpbXUU5QZ5yyl9P/MdvfhktVcPtnDHs7v449r9XHbiDIpyg/zxxf38+2VLB6zzTaY/vriPz/1uHV2RmP86Z3PNGfM4dnoJ9720n0/dvZbjZ5Xxy/edxg//sZVfPbuLaMwRTGj1Fl8dr2KAIBlgyfQS/rb+IGcvquDK5TMxMw41d7LxQDP/dvGxQ44vFAzwpSuOY+HUYr5033re+n9P89P3ruhXhx6LOf71t+t4Znsd33nHSZy7uJJzF1eycv4UPv6bF3nbLU9z8fHTeXD9QQpzQ3z9LSfwjhVzerWsA69jSszBoZYuBWkp0h6OUN8WHnARHhGRdDOSTPI+ILFX12x/W6K9wH3OuW7n3A5gM17Q3Itz7lbn3Arn3IrKyv61ijL24m3gTp47adAsXzYzM646fS5/++RruGjZND5+weKeWu1kigfJ1/9yNX96aT//+oZj+eHVJ3PDaxYQjsa4e9WeQR+7amc9D64/eMTnjkRjfO3+DXzy7rWcNGcSv/rA6bzh+OncvXoPb/ju41zxwyf5xF0vcuq8ydxx/UomFeayZEYpXZEYO+t6l070ZJKLBg7ov3Ll8SyaWswn717LB29fzcGmTh73F6lJ7I88lPecMY/b3nca+5s6OP/bj3qrK6470LOM+Dce3MR9/nv4llMO99Y+dd5kHvj4uVy4dBp/WXeAy0+cyd8/81quOn1uvwAZYIb/Fb9KLlJnf6O36qWCZBHJBCOJklYBi81sPl5wfBXwrj7H/BG4GviFmVXglV9sT+I4JUniQfI5iyb2h5Q55YX85NoVY/b8c8sLCQaMSDTGT96zoqd39KKpJZy1cAp3PLuLD71mQb82eE0d3Xz4V2voisR43X9NJXeY1Qv7amwP87HfvMgTW2q59sx5/Ncbl5ETDHDu4kr+87Jl/Hb1Hu5atYcLlkzlB1efQkGuN/lwyXSvU8erB1pYWHl4FcS6Nj+TXDJwkDy/ooh7P3wWv3hqB996aBMX/e9jzCjLp7Ikb9DFVgZy7uJK/vKxc7j9mV38+aX9PLyhmuK8EKfMm8zjm2t498q5fPS8hf0eV1aYwy3XnEJNaxdTS4auc01ceXHs/uZlKPEPKMrki0gmGDZIds5FzOxG4EEgCPzcObfezL4CrHbO3efve72ZbQCiwL865+rGcuByZOJZ03MWT6xSi/GWnxPkF9edxuzJBSyo7L309rVnVvHhX6/h768e4g3HTe+1738f3kydv9jCs9vrei3EMZzq5k7e/qNnONjUyf+89QTeeVrvkqbyolw+9NqFfOi1/YPNRVOLCQaMVw8291p1sLZ1+NXRggHj+nMXcOHSaXzud+t4bkc9bz1l9qjrvOdNKeK/3riMf790Kc9ur+OPL+7jb68c5A3HTePLVxw36POZ2bABMsAMPzCLZzNl/MV7lc+arCBZRNLfiL5vd849ADzQZ9sXEm474NP+j6Sx8sJcSvJDnDR7UqqHkvUGC3AvXDqVmWX5/OqZXb2C5I0Hmrn9mZ287dTZ3L/uAI9srB5VkPznl/azu76d3374TE6rKh/VWPNzgiyoKGLjgd6T92pbuyjOCw3Z7i6uqqKI33zwDP7+6iFOnF02qvMnCgaMsxdVcPaiCr7+1hMJGEmZWFmcF6KsIEflFim0v7GDYMCYNsg3EyIi6UQr7k0wH79wMb/6wMphV7uTsRMKBnjXyrk8ubWWrYe8zhDOOb5433rKCnL4z8uWcu7iCh7ZUD2qpZtf2dfE9NL8UQfIcUtmlPLqweZe2+oGWW1vMIGAcdGyaUwrTU6Lr2DAktp5ZOakAgXJKbSvoYPppfm6/ohIRtCVaoKZNamA5XMmpXoYE95Vp88lNxjg1/6S2Pe9tJ/nd9Tzr29YwqTCXC5cNo39TZ2s3988zDMd9sr+Zo6fdeQZ3CXTS9jb0NGr93BdW9eQpRaZZtak/EGXJ5ext6+xg5mT1CNZRDKDgmSRFKgozuPSE6Zz75q9HGru5L8f2MgJs8p452leI5kLlkzFzOtlPBLt4Qjbalo5ftbIJ8v1tXSGN3lvc0K/ZC+TnD1fjc8oK+BAU/bXJJvZxWa2ycy2mtlNA+z/tJltMLN1ZvZ3M5s3HuPa39ShSXsikjEUJIukyLVnVdHaFeGqW5+lurmLr1x5XE+P4inFeZw6d/KIg+QN+5txDo6feTSZZC/A3pgQJNe2hgftkZyJZk4qoKmje8iFSzKdmQWBm4FLgGXA1Wa2rM9hLwIrnHMnAvcC3xjrcUVjjgONnWr/JiIZQ0GySIqcPGcSx88qZXttG+9YMZuT5/Ze6vqiZdPYcKB5ROUBr+xrAuCEo5gwN6Msn5L8EK8e8Eo8YjFHfVvXmC96Mp7iX/UfyO6Si9OBrc657c65MHAXcGXiAc65fzrn2v27z+L1vx9TNS1dRGJOmWQRyRgKkkVSxMy48XWLWFhZxL9dvKTf/nhv5UdGkE1+eV8zFcV5TD2KrgFmxtLppT3LUze0h4m5odu/ZZrEXslZbBaQuFrNXn/bYD4A/HVMRwTsa/RicrV/E5FMoSBZJIUuPn4Gf//MeQNmaxdWFrOgsohHNg4fJK/f38QJs0qPuhPEkhklbDrYQizmevo1Z1NN8kz1Su7FzK4BVgDfHGT/DWa22sxW19TUHNW59mm1PRHJMAqSRdLYRUun8ez2ul4dJ/rq7I6y5VDrUXW2iFsyvZTWrgj7GjuobfVW2xtNC7h0N7Ukj2DAsr0N3D5gTsL92f62XszsQuA/gCucc10DPZFz7lbn3Arn3IrKyqNbpVOr7YlIplGQLJLGLlo2je6o47FNg2fxNh5oJhpzHHcUk/bilvgdLjYeaKbOX20vm2qSQ8EA00ry2N+U1UHyKmCxmc03s1zgKuC+xAPM7GTgx3gB8qHxGNS+hg7KCnIozhvRGlYiIimnIFkkjZ08dzJTinKH7HLxit9L+Wgm7cUdO80Lkl892EJdPJOcRTXJkP0LijjnIsCNwIPARuAe59x6M/uKmV3hH/ZNoBj4rZmtNbP7Bnm6pNnfqPZvIpJZ9JFeJI0FA8b5S6byt/UH6Y7GyBlgpbJX9jYxuTCHmWVHv0hDUV6IeVMKefVgMwsriwkYTCrMviB57Z7GVA9jTDnnHgAe6LPtCwm3LxzvMe1r7GD25MLxPq2IyBFTJlkkzV24bBotnRGe31E/4P5X9jdx/KyypC3fvGR6Ca8eaKG2NUx5UV5P7+ZsMXNSAQeaOojFRr7ktxy9fY0dzNJqeyKSQRQki6S5cxdXkBcK8MDLB/rt64pE2VzdkpRJe3FLppeys66NvQ3tWbWQSNysSfl0R13PxEQZe82d3bR0RtT+TUQyioJkkTRXmBvijSfO5Pcv7KOxPdxr3+aDrXRH3VGttNfX0hklxBys2dWQVZ0t4mZOjF7JaUWdLUQkEylIFskA1587n47uKHc8t7vX9lf2+yvtJTmTDNAejjKlKHs6W8TNKFOv5PGmIFlEMpGCZJEMsHRGKecuruCXT+8kHIn1bH9lXxMl+SHmlCcv+JhbXkhBThDIrh7JcfHFLA5kdxu4tLKvwXuvZytIFpEMoiBZJENcf+4CDrV0cd9L+3u2vbKvieNnJm/SHkAgYBwz3WsFl009kuNKC0IU5QZVbjGO9jV2khsMZOW/JxHJXgqSRTLEaxZXcOy0En76xHacc3RHY2w82JKU/sh9LfWD5GzrkQxgZlnfKznd7G/sYMakfAJZ1ilFRLKbgmSRDGFmfODc+bx6sIUnt9aypbqVcCTGcTNLk36uJVmcSYb4giKqSR4v+xo7mFmmUgsRySwKkkUyyJXLZ1JZksdPntgxJpP24s5aVEF5US7H+sFytlEmeXxptT0RyURacU8kg+SFgrz3zHl866HNdEdiFOUGqZpSlPTzHDOthBf+66KkP2+6mFmWT11bmM7uKPn+JEUZG93RGNXNneqRLCIZR5lkkQzz7pXzyM8J8Mz2Oo6bWaY6zyMws6fDhUouxtrBpk5iDq22JyIZR0GySIaZXJTL20+dA5DUlfYmkniQrJKLsRd/j2dNKkzxSERERkdBskgG+sA58ynICXLWwimpHkpGmqVV98bNvp6FRJRJFpHMoppkkQxUVVHE2i9eRG5Qn3OPxLSyPMyUSR4PWm1PRDKVgmSRDJUX0oSzI5UXClJZnKcgeRzsa+ykojhXEyRFJOMoDSUiE9IM9UoeF/vU/k1EMpSCZBGZkGZNylcmeRzsb+zoqQEXEckkCpJFZEKaWVbA/qYOnHOpHkrWcs6xr0GZZBHJTAqSRWRCmjmpgM7uGA3t3akeStZqbO+mozuqIFlEMpKCZBGZkNQreezt6+mRrCBZRDKPgmQRmZDUK3nsKUgWkUymIFlEJqQZ/uIWyiSPnZ7V9iYrSBaRzKMgWUQmpClFuZy9aArlRbmpHkrWKvff48mFOakeiojIqGkxERGZkMyMO64/I9XDyGpXLp/FlctnpXoYIiJHRJlkEREREZE+FCSLiIiIiPShIFlEREREpA8FySIiIiIifShIFhERERHpQ0GyiIiIiEgfCpJFRERERPpQkCwiIiIi0oeCZBERERGRPhQki4hkGTO72Mw2mdlWM7tpgP2vMbMXzCxiZm9LxRhFRNKdgmQRkSxiZkHgZuASYBlwtZkt63PYbuA64M7xHZ2ISOYIpXoAIiKSVKcDW51z2wHM7C7gSmBD/ADn3E5/XywVAxQRyQTKJIuIZJdZwJ6E+3v9baNmZjeY2WozW11TU5OUwYmIZIqUZZLXrFlTa2a7hjmsAqgdj/GMEY0/dTJ57KDxp9JIxz5vrAeSas65W4FbAcysZgTXbJgYf/fpSuNPnUweO0yM8Y/6mp2yINk5VzncMWa22jm3YjzGMxY0/tTJ5LGDxp9KmTx23z5gTsL92f62ozKSazZk9vuXyWMHjT+VMnnsoPEPRuUWIiLZZRWw2Mzmm1kucBVwX4rHJCKScRQki4hkEedcBLgReBDYCNzjnFtvZl8xsysAzOw0M9sLvB34sZmtT92IRUTSU7p3t7g11QM4Shp/6mTy2EHjT6VMHjsAzrkHgAf6bPtCwu1VeGUYYyGT379MHjto/KmUyWMHjX9A5pwbi+cVEREREclYKrcQEREREekjbYPk4ZZVTRUz22lmL5vZWjNb7W8rN7OHzWyL/+dkf7uZ2ff917DOzE5JeJ73+sdvMbP3juF4f25mh8zslYRtSRuvmZ3qvx9b/cfaOIz/S2a2z/87WGtmlybs+7w/lk1m9oaE7QP+e/InNz3nb7/bn+iUrLHPMbN/mtkGM1tvZp/wt2fE+z/E+DPl/c83s+fN7CV//F8e6pxmluff3+rvrzrS1zURpet7Ybpm65o98rHrmn14u67ZAM65tPsBgsA2YAGQC7wELEv1uPyx7QQq+mz7BnCTf/sm4H/825cCfwUMOAN4zt9eDmz3/5zs3548RuN9DXAK8MpYjBd43j/W/MdeMg7j/xLw2QGOXeb/W8kD5vv/hoJD/XsC7gGu8m//CPhIEsc+AzjFv10CbPbHmBHv/xDjz5T334Bi/3YO8Jz/Xg14TuCjwI/821cBdx/p65poP+n8XqBrtq7ZIx+7rtm6Zvf6SddMcs+yqs65MBBfVjVdXQn80r/9S+BNCdtvd55ngUlmNgN4A/Cwc67eOdcAPAxcPBYDc849DtSPxXj9faXOuWed9y/z9oTnGsvxD+ZK4C7nXJdzbgewFe/f0oD/nvxP8OcD9/qPT3wvkjH2A865F/zbLXidBmaRIe//EOMfTLq9/8451+rfzfF/3BDnTPx7uRe4wB/jqF5XssafYTLtvdA1W9fsgcaua7au2b2ka5CctGVVx4ADHjKzNWZ2g79tmnPugH/7IDDNvz3Y60j160vWeGf5t/tuHw83+l9v/Tz+1RejH/8UoNF5LbMStyed/zXQyXifjDPu/e8zfsiQ99/Mgma2FjiE94tq2xDn7Bmnv7/JH2O6/j9OJ+n8XuiarWv2qOmarWs2pG+QnM7Occ6dAlwC/IuZvSZxp//pMGNahmTaeH23AAuB5cAB4NspHc0wzKwY+B3wSedcc+K+THj/Bxh/xrz/zrmoc245Xruz04ElqR2RpICu2amXMdcM0DU7ldLtmp2uQfKYLKuaDM65ff6fh4A/4P0lVvtfo+D/ecg/fLDXkerXl6zx7qN3r9VxeR3OuWr/P1IM+Ane3wHDjHOg7XV4X4+F+mxPGjPLwbtY3eGc+72/OWPe/4HGn0nvf5xzrhH4J3DmEOfsGae/v8wfY7r+P04nafte6Jrdb7uu2UPQNVvX7L4DSbsfvEVOtuMVXMeLq49Lg3EVASUJt5/Gq0v7Jr2L+r/h376M3kX9z/vby4EdeAX9k/3b5WM47ip6T6JI2njpPwnh0nEY/4yE25/Cqz0COI7exfrb8Qr1B/33BPyW3hMCPprEcRtezdl3+2zPiPd/iPFnyvtfCUzybxcATwBvHOycwL/QexLIPUf6uibaT7q+F+iarWv26Mata7au2b3HlOz/HEl8sy7Fm5m5DfiPVI/HH9MC/019CVgfHxdeDczfgS3AIwn/GQy42X8NLwMrEp7r/XjF5FuB943hmH+D9/VKN179zQeSOV5gBfCK/5gf4i9QM8bj/5U/vnXAfX0uAP/hj2UTCbOGB/v35P+dPu+/rt8CeUkc+zl4X8utA9b6P5dmyvs/xPgz5f0/EXjRH+crwBeGOieQ79/f6u9fcKSvayL+pON7ga7ZumaPbuy6Zqf2/U+7a7ZW3BMRERER6SNda5JFRERERFJGQbKIiIiISB8KkkVERERE+lCQLCIiIiLSh4JkEREREZE+FCRLWjCzVv/PKjN7V5Kf+9/73H86mc8vIjLR6JotE4GCZEk3VcCoLrgJK/EMptcF1zl31ijHJCIiA6tC12zJUgqSJd18HTjXzNaa2afMLGhm3zSzVWa2zsw+BGBm55nZE2Z2H7DB3/ZHM1tjZuvN7AZ/29eBAv/57vC3xTMg5j/3K2b2spm9M+G5HzWze83sVTO7w8wsBe+FiEi60zVbstZwn+ZExttNwGedc28E8C+cTc6508wsD3jKzB7yjz0FON45t8O//37nXL2ZFQCrzOx3zrmbzOxG59zyAc71FmA5cBJQ4T/mcX/fyXhLW+4HngLOBp5M9osVEclwumZL1lImWdLd64FrzWwt8Bze8qCL/X3PJ1xsAT5uZi8BzwJzEo4bzDnAb5xzUedcNfAYcFrCc+91zsXwlvasSsJrERHJdrpmS9ZQJlnSnQEfc8492Guj2XlAW5/7FwJnOufazexRvHXdj1RXwu0o+r8iIjISumZL1lAmWdJNC1CScP9B4CNmlgNgZseYWdEAjysDGvyL7RLgjIR93fHH9/EE8E6/hq4SeA3wfFJehYjIxKBrtmQtfdKSdLMOiPpfwd0GfA/va7MX/IkYNcCbBnjc34APm9lGYBPe13dxtwLrzOwF59y7E7b/ATgTeAlwwL855w76F2wRERmertmStcw5l+oxiIiIiIikFZVbiIiIiIj0oSBZRERERKQPBckiIiIiIn0oSBYRERER6UNBsoiIiIhIHwqSRURERET6UJAsIiIiItKHgmQRERERkT7+fyGv3KjnodHkAAAAAElFTkSuQmCC\n",
+      "text/plain": [
+       "<Figure size 864x432 with 2 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.figure(\"train\", (12, 6))\n",
+    "plt.subplot(1, 2, 1)\n",
+    "plt.title(\"Iteration Average Loss\")\n",
+    "x = [eval_num * (i + 1) for i in range(len(epoch_loss_values))]\n",
+    "y = epoch_loss_values\n",
+    "plt.xlabel(\"Iteration\")\n",
+    "plt.plot(x, y)\n",
+    "plt.subplot(1, 2, 2)\n",
+    "plt.title(\"Val Mean Dice\")\n",
+    "x = [eval_num * (i + 1) for i in range(len(metric_values))]\n",
+    "y = metric_values\n",
+    "plt.xlabel(\"Iteration\")\n",
+    "plt.plot(x, y)\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Check best model output with the input image and label"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 1296x432 with 3 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "case_num = 4\n",
+    "model.load_state_dict(torch.load(os.path.join(root_dir, \"best_metric_model.pth\")))\n",
+    "model.eval()\n",
+    "with torch.no_grad():\n",
+    "    img_name = os.path.split(val_ds[case_num][\"image_meta_dict\"][\"filename_or_obj\"])[1]\n",
+    "    img = val_ds[case_num][\"image\"]\n",
+    "    label = val_ds[case_num][\"label\"]\n",
+    "    val_inputs = torch.unsqueeze(img, 1).cuda()\n",
+    "    val_labels = torch.unsqueeze(label, 1).cuda()\n",
+    "    val_outputs = sliding_window_inference(\n",
+    "        val_inputs, (96, 96, 96), 4, model, overlap=0.8\n",
+    "    )\n",
+    "    plt.figure(\"check\", (18, 6))\n",
+    "    plt.subplot(1, 3, 1)\n",
+    "    plt.title(\"image\")\n",
+    "    plt.imshow(val_inputs.cpu().numpy()[0, 0, :, :, slice_map[img_name]], cmap=\"gray\")\n",
+    "    plt.subplot(1, 3, 2)\n",
+    "    plt.title(\"label\")\n",
+    "    plt.imshow(val_labels.cpu().numpy()[0, 0, :, :, slice_map[img_name]])\n",
+    "    plt.subplot(1, 3, 3)\n",
+    "    plt.title(\"output\")\n",
+    "    plt.imshow(\n",
+    "        torch.argmax(val_outputs, dim=1).detach().cpu()[0, :, :, slice_map[img_name]]\n",
+    "    )\n",
+    "    plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Cleanup data directory\n",
+    "\n",
+    "Remove directory if a temporary was used."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "if directory is None:\n",
+    "    shutil.rmtree(root_dir)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.8.13"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}

From 67186713d8457d706711be4bc2289639261b512d Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Sat, 23 Jul 2022 23:18:34 -0700
Subject: [PATCH 06/17] ready, pending final checks

---
 .../swin_unetr_btcv_segmentation_3d.ipynb     | 2338 ++---------------
 1 file changed, 152 insertions(+), 2186 deletions(-)

diff --git a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
index 240f983ed2..7f91d3c1a0 100644
--- a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
+++ b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
@@ -15,20 +15,32 @@
     "1. Define a new transform according to MONAI transform API.\n",
     "1. Load Nifti image with metadata, load a list of images and stack them.\n",
     "1. Randomly adjust intensity for data augmentation.\n",
-    "1. Cache IO and transforms to accelerate training and validation.\n",
+    "1. Cache IO and transforms, ThreadDataLoader, and AMP to accelerate training and validation.\n",
     "1. Swin UNETR model, DiceCE loss function, Mean Dice metric for multi-organ segmentation task.\n",
     "\n",
-    "For this tutorial, the dataset needs to be downloaded from: https://www.synapse.org/#!Synapse:syn3193805/wiki/217752. \n",
+    "For this tutorial, the dataset needs to be downloaded from: https://www.synapse.org/#!Synapse:syn3193805/wiki/217752. More details are provided in the \"Download dataset\" section below.\n",
     "\n",
     "In addition, the json file for data splits needs to be downloaded from this [link](https://drive.google.com/file/d/1t4fIQQkONv7ArTSZe4Nucwkk1KfdUDvW/view?usp=sharing). Once downloaded, place the json file in the same folder as the dataset. \n",
     "\n",
     "For BTCV dataset, under Institutional Review Board (IRB) supervision, 50 abdomen CT scans of were randomly selected from a combination of an ongoing colorectal cancer chemotherapy trial, and a retrospective ventral hernia study. The 50 scans were captured during portal venous contrast phase with variable volume sizes (512 x 512 x 85 - 512 x 512 x 198) and field of views (approx. 280 x 280 x 280 mm3 - 500 x 500 x 650 mm3). The in-plane resolution varies from 0.54 x 0.54 mm2 to 0.98 x 0.98 mm2, while the slice thickness ranges from 2.5 mm to 5.0 mm. \n",
     "\n",
-    "Target: 13 abdominal organs including 1. Spleen 2. Right Kidney 3. Left Kideny 4.Gallbladder 5.Esophagus 6. Liver 7. Stomach 8.Aorta 9. IVC 10. Portal and Splenic Veins 11. Pancreas 12 Right adrenal gland 13 Left adrenal gland.\n",
-    "\n",
-    "Modality: CT\n",
-    "Size: 30 3D volumes (24 Training + 6 Testing)  \n",
-    "Challenge: BTCV MICCAI Challenge\n",
+    "- Target: 13 abdominal organs including \n",
+    "    1. Spleen \n",
+    "    2. Right Kidney \n",
+    "    3. Left Kideny \n",
+    "    4. Gallbladder \n",
+    "    5. Esophagus \n",
+    "    6. Liver \n",
+    "    7. Stomach \n",
+    "    8. Aorta \n",
+    "    9. IVC \n",
+    "    10. Portal and Splenic Veins \n",
+    "    11. Pancreas \n",
+    "    12. Right adrenal gland \n",
+    "    13. Left adrenal gland.\n",
+    "- Modality: CT\n",
+    "- Size: 30 3D volumes (24 Training + 6 Testing)\n",
+    "- Challenge: BTCV MICCAI Challenge\n",
     "\n",
     "The following figure shows image patches with the organ sub-regions that are annotated in the CT (top left) and the final labels for the whole dataset (right).\n",
     "\n",
@@ -79,54 +91,6 @@
     "If training from scratch is desired, please skip the step for initializing from pre-trained weights. "
    ]
   },
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "/opt/conda/lib/python3.8/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"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Tue Jul 12 02:08:22 2022       \n",
-      "+-----------------------------------------------------------------------------+\n",
-      "| NVIDIA-SMI 450.119.04   Driver Version: 450.119.04   CUDA Version: 11.6     |\n",
-      "|-------------------------------+----------------------+----------------------+\n",
-      "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
-      "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
-      "|                               |                      |               MIG M. |\n",
-      "|===============================+======================+======================|\n",
-      "|   0  Tesla V100-SXM2...  On   | 00000000:07:00.0 Off |                    0 |\n",
-      "| N/A   35C    P0    43W / 163W |      3MiB / 32510MiB |      0%      Default |\n",
-      "|                               |                      |                  N/A |\n",
-      "+-------------------------------+----------------------+----------------------+\n",
-      "                                                                               \n",
-      "+-----------------------------------------------------------------------------+\n",
-      "| Processes:                                                                  |\n",
-      "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
-      "|        ID   ID                                                   Usage      |\n",
-      "|=============================================================================|\n",
-      "|  No running processes found                                                 |\n",
-      "+-----------------------------------------------------------------------------+\n"
-     ]
-    }
-   ],
-   "source": [
-    "# TEMP CELL PLEASE DELETE\n",
-    "import torch\n",
-    "torch.cuda.is_available()\n",
-    "\n",
-    "!nvidia-smi"
-   ]
-  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -136,93 +100,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
-   "metadata": {
-    "scrolled": true
-   },
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
-      "Collecting nibabel==3.1.1\n",
-      "  Downloading nibabel-3.1.1-py3-none-any.whl (3.3 MB)\n",
-      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.3/3.3 MB\u001b[0m \u001b[31m102.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
-      "\u001b[?25hRequirement already satisfied: packaging>=14.3 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (21.3)\n",
-      "Requirement already satisfied: numpy>=1.13 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (1.22.3)\n",
-      "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.8/site-packages (from packaging>=14.3->nibabel==3.1.1) (3.0.8)\n",
-      "Installing collected packages: nibabel\n",
-      "  Attempting uninstall: nibabel\n",
-      "    Found existing installation: nibabel 3.2.2\n",
-      "    Uninstalling nibabel-3.2.2:\n",
-      "      Successfully uninstalled nibabel-3.2.2\n",
-      "Successfully installed nibabel-3.1.1\n",
-      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
-      "\u001b[0mLooking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
-      "Collecting tqdm==4.63.0\n",
-      "  Downloading tqdm-4.63.0-py2.py3-none-any.whl (76 kB)\n",
-      "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m76.6/76.6 kB\u001b[0m \u001b[31m28.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
-      "\u001b[?25hInstalling collected packages: tqdm\n",
-      "  Attempting uninstall: tqdm\n",
-      "    Found existing installation: tqdm 4.64.0\n",
-      "    Uninstalling tqdm-4.64.0:\n",
-      "      Successfully uninstalled tqdm-4.64.0\n",
-      "Successfully installed tqdm-4.63.0\n",
-      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
-      "\u001b[0m"
-     ]
-    }
-   ],
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
    "source": [
-    "# FIXME: compatible with 0.9.0 but not with 0.9.1rc\n",
-    "# !pip install 'git+https://github.com/Project-MONAI/MONAI#egg.gitmonai@0.8.1+271.g07de215c'\n",
-    "\n",
-    "!pip install nibabel==3.1.1\n",
-    "!pip install tqdm==4.63.0\n",
+    "!python -c \"import monai; import nibabel; import tqdm\" || pip install -q \"monai-weekly[nibabel, tqdm]\"\n",
     "!python -c \"import matplotlib\" || pip install -q matplotlib\n",
     "%matplotlib inline"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "MONAI version: 0.9.0\n",
-      "Numpy version: 1.22.3\n",
-      "Pytorch version: 1.12.0a0+bd13bc6\n",
-      "MONAI flags: HAS_EXT = True, USE_COMPILED = False\n",
-      "MONAI rev id: af0e0e9f757558d144b655c63afcea3a4e0a06f5\n",
-      "MONAI __file__: /opt/monai/monai/__init__.py\n",
-      "\n",
-      "Optional dependencies:\n",
-      "Pytorch Ignite version: 0.4.8\n",
-      "Nibabel version: 3.1.1\n",
-      "scikit-image version: 0.19.3\n",
-      "Pillow version: 9.0.1\n",
-      "Tensorboard version: 2.8.0\n",
-      "gdown version: 4.4.0\n",
-      "TorchVision version: 0.13.0a0\n",
-      "tqdm version: 4.64.0\n",
-      "lmdb version: 1.3.0\n",
-      "psutil version: 5.9.0\n",
-      "pandas version: 1.3.5\n",
-      "einops version: 0.4.1\n",
-      "transformers version: 4.19.4\n",
-      "mlflow version: 1.26.1\n",
-      "pynrrd version: 0.4.3\n",
-      "\n",
-      "For details about installing the optional dependencies, please visit:\n",
-      "    https://docs.monai.io/en/latest/installation.html#installing-the-recommended-dependencies\n",
-      "\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "import os\n",
     "import shutil\n",
@@ -231,13 +122,11 @@
     "import matplotlib.pyplot as plt\n",
     "import numpy as np\n",
     "from tqdm import tqdm\n",
-    "import time\n",
     "\n",
     "from monai.losses import DiceCELoss\n",
     "from monai.inferers import sliding_window_inference\n",
     "from monai.transforms import (\n",
     "    AsDiscrete,\n",
-    "    AddChanneld,\n",
     "    Compose,\n",
     "    CropForegroundd,\n",
     "    LoadImaged,\n",
@@ -248,11 +137,8 @@
     "    ScaleIntensityRanged,\n",
     "    Spacingd,\n",
     "    RandRotate90d,\n",
-    "    ToTensord,\n",
     "    EnsureTyped,\n",
-    "    ToDeviced\n",
     ")\n",
-    "from monai.utils import set_determinism\n",
     "\n",
     "from monai.config import print_config\n",
     "from monai.metrics import DiceMetric\n",
@@ -260,7 +146,6 @@
     "\n",
     "from monai.data import (\n",
     "    ThreadDataLoader,\n",
-    "    DataLoader,\n",
     "    CacheDataset,\n",
     "    load_decathlon_datalist,\n",
     "    decollate_batch,\n",
@@ -285,17 +170,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "/tmp/tmp8tsvk5fk\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "directory = os.environ.get(\"MONAI_DATA_DIRECTORY\")\n",
     "root_dir = tempfile.mkdtemp() if directory is None else directory\n",
@@ -307,23 +184,29 @@
    "metadata": {},
    "source": [
     "## Setup transforms for training and validation\n",
-    "To save on GPU memory utilization, the num_samples can be reduced to 2. "
+    "To save on GPU memory utilization, the num_samples can be reduced to 2. \n",
+    "\n",
+    "A note on design:\n",
+    "\n",
+    "- We are moving towards the use of MONAI's MetaTensor in place of numpy arrays or PyTorch tensors. MetaTensors have the benefit of carrying the metadata directly with the tensor, but in some use cases (like here with training, where training data are only used for computing loss and metadata is not useful), we can safely disregard the metadata to improve speed.\n",
+    "\n",
+    "- Hence, you will see `EnsureTyped` being used before the first random transform in the training transform chain, which caches the result of deterministic transforms on GPU, with `track_meta = False`. On the other hand, in the following demos we will display example validation images, which uses metadata, so we use `EnsureTyped` with `track_meta = True`. Since there are no random transforms during validation, tracking metadata for validation images causes virtually no slowdown (~0.5%)."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
     "num_samples = 4\n",
+    "\n",
     "os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\"\n",
     "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
     "\n",
     "train_transforms = Compose(\n",
     "    [\n",
-    "        LoadImaged(keys=[\"image\", \"label\"]),\n",
-    "        AddChanneld(keys=[\"image\", \"label\"]),\n",
+    "        LoadImaged(keys=[\"image\", \"label\"], ensure_channel_first=True),\n",
     "        Orientationd(keys=[\"image\", \"label\"], axcodes=\"RAS\"),\n",
     "        Spacingd(\n",
     "            keys=[\"image\", \"label\"],\n",
@@ -339,8 +222,7 @@
     "            clip=True,\n",
     "        ),\n",
     "        CropForegroundd(keys=[\"image\", \"label\"], source_key=\"image\"),\n",
-    "        EnsureTyped(keys=[\"image\", \"label\"]),\n",
-    "        ToDeviced(keys=[\"image\", \"label\"], device=device),\n",
+    "        EnsureTyped(keys=[\"image\", \"label\"], device=device, track_meta=False),\n",
     "        RandCropByPosNegLabeld(\n",
     "            keys=[\"image\", \"label\"],\n",
     "            label_key=\"label\",\n",
@@ -376,13 +258,11 @@
     "            offsets=0.10,\n",
     "            prob=0.50,\n",
     "        ),\n",
-    "#         ToTensord(keys=[\"image\", \"label\"]),\n",
     "    ]\n",
     ")\n",
     "val_transforms = Compose(\n",
     "    [\n",
-    "        LoadImaged(keys=[\"image\", \"label\"]),\n",
-    "        AddChanneld(keys=[\"image\", \"label\"]),\n",
+    "        LoadImaged(keys=[\"image\", \"label\"], ensure_channel_first=True),\n",
     "        Orientationd(keys=[\"image\", \"label\"], axcodes=\"RAS\"),\n",
     "        Spacingd(\n",
     "            keys=[\"image\", \"label\"],\n",
@@ -393,9 +273,7 @@
     "            keys=[\"image\"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True\n",
     "        ),\n",
     "        CropForegroundd(keys=[\"image\", \"label\"], source_key=\"image\"),\n",
-    "#        ToTensord(keys=[\"image\", \"label\"]),\n",
-    "        EnsureTyped(keys=[\"image\", \"label\"]),\n",
-    "        ToDeviced(keys=[\"image\", \"label\"], device=device),\n",
+    "        EnsureTyped(keys=[\"image\", \"label\"], device=device, track_meta=True),\n",
     "    ]\n",
     ")"
    ]
@@ -404,42 +282,32 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    " ## Download dataset and format in the folder.\n",
-    "1. Download dataset from here: https://www.synapse.org/#!Synapse:syn3193805/wiki/89480. After you open the link, navigate to the \"Files\" tab, then download Abdomen/RawData.zip. \n",
+    "## Download dataset and format in the folder\n",
+    "1. Download dataset from here: https://www.synapse.org/#!Synapse:syn3193805/wiki/89480. After you open the link, navigate to the \"Files\" tab, then download Abdomen/RawData.zip.\n",
     "\n",
     "    Note that you may need to register for an account on Synapse and consent to use agreements before being able to view/download this file. There are options to download directly from the browser or from the command line; please refer to Synapse API documentation for more info.\n",
     "\n",
+    "\n",
     "2. After downloading the zip file, unzip. Then put images from `RawData/Training/img` in `./data/imagesTr`, and put labels from `RawData/Training/label` in `./data/labelsTr`.\n",
     "\n",
+    "\n",
     "3. Make a JSON file to define train/val split and other relevant parameters. Place the JSON file at `./data/dataset_0.json`.\n",
     "\n",
-    "   You can download an example of the JSON file [here](https://drive.google.com/file/d/1EF2By3k1NWDIIoH4r_3Xj3Q9yUDxoF0U/view?usp=sharing). If you would like to use this directly, please move it into the `./data` folder.   "
+    "    You can download an example of the JSON file [here](https://drive.google.com/file/d/1t4fIQQkONv7ArTSZe4Nucwkk1KfdUDvW/view?usp=sharing). If you would like to use this directly, please move it into the `./data` folder."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Loading dataset: 100%|██████████| 24/24 [00:50<00:00,  2.11s/it]\n",
-      "Loading dataset: 100%|██████████| 6/6 [00:18<00:00,  3.10s/it]\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
-    "torch.cuda.empty_cache()\n",
     "data_dir = \"data/\"\n",
     "split_JSON = \"dataset_0.json\"\n",
     "\n",
     "datasets = data_dir + split_JSON\n",
     "datalist = load_decathlon_datalist(datasets, True, \"training\")\n",
     "val_files = load_decathlon_datalist(datasets, True, \"validation\")\n",
-    "\n",
-    "# TODO: try thread_workers\n",
     "train_ds = CacheDataset(\n",
     "    data=datalist,\n",
     "    transform=train_transforms,\n",
@@ -447,7 +315,7 @@
     "    cache_rate=1.0,\n",
     "    num_workers=8,\n",
     ")\n",
-    "train_loader = ThreadDataLoader(train_ds, batch_size=1, shuffle=True, use_thread_workers=False, num_workers=0)\n",
+    "train_loader = ThreadDataLoader(train_ds, num_workers=0, batch_size=1, shuffle=True)\n",
     "val_ds = CacheDataset(\n",
     "    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4\n",
     ")\n",
@@ -463,7 +331,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
@@ -475,7 +343,7 @@
     },
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "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",
       "text/plain": [
        "<Figure size 1296x432 with 2 Axes>"
       ]
@@ -496,7 +364,7 @@
     "    \"img0040.nii.gz\": 180,\n",
     "}\n",
     "case_num = 1\n",
-    "img_name = os.path.split(val_ds[case_num][\"image_meta_dict\"][\"filename_or_obj\"])[1]\n",
+    "img_name = os.path.split(val_ds[case_num]['image'].meta[\"filename_or_obj\"])[1]\n",
     "img = val_ds[case_num][\"image\"]\n",
     "label = val_ds[case_num][\"label\"]\n",
     "img_shape = img.shape\n",
@@ -518,15 +386,18 @@
    "source": [
     "### Create Swin UNETR model\n",
     "\n",
-    "In this section, we create Swin UNETR model for the 14-class multi-organ segmentation. We use a feature size of 48 which is compatible with self-supervised pre-trained weights. We also use gradient checkpointing (use_checkpoint) for more memory-efficient training. "
+    "In this section, we create a Swin UNETR model for the 14-class multi-organ segmentation. We use a feature size of 48, which is compatible with the self-supervised pre-trained weights. We also use gradient checkpointing (use_checkpoint) for more memory-efficient training. "
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
+    "os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\"\n",
+    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
+    "\n",
     "model = SwinUNETR(\n",
     "    img_size=(96, 96, 96),\n",
     "    in_channels=1,\n",
@@ -542,15 +413,13 @@
    "source": [
     "### Initialize Swin UNETR encoder from self-supervised pre-trained weights\n",
     "\n",
-    "In this section, we intialize the Swin UNETR encoder from pre-trained weights. The weights can be downloaded using the `wget` command below, or by clicking on [this link](https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/model_swinvit.pt). If training from scratch is desired, please skip this section."
+    "In this section, we intialize the Swin UNETR encoder from pre-trained weights. The weights can be downloaded using the wget command below, or by following [this link](https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/model_swinvit.pt) to GitHub. If training from scratch is desired, please skip this section."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
-   "metadata": {
-    "scrolled": false
-   },
+   "execution_count": null,
+   "metadata": {},
    "outputs": [],
    "source": [
     "# uncomment to download the pre-trained weights\n",
@@ -559,17 +428,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Using pretrained self-supervied Swin UNETR backbone weights !\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "weight = torch.load(\"./model_swinvit.pt\")\n",
     "model.load_from(weights=weight)\n",
@@ -585,7 +446,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -603,7 +464,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": null,
    "metadata": {
     "scrolled": true
    },
@@ -613,7 +474,7 @@
     "    model.eval()\n",
     "    with torch.no_grad():\n",
     "        for step, batch in enumerate(epoch_iterator_val):\n",
-    "            val_inputs, val_labels = (batch[\"image\"].cuda(device=device), batch[\"label\"].cuda(device=device))\n",
+    "            val_inputs, val_labels = (batch[\"image\"].cuda(), batch[\"label\"].cuda())\n",
     "            with torch.cuda.amp.autocast():\n",
     "                val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)\n",
     "            val_labels_list = decollate_batch(val_labels)\n",
@@ -640,16 +501,17 @@
     "    epoch_iterator = tqdm(\n",
     "        train_loader, desc=\"Training (X / X Steps) (loss=X.X)\", dynamic_ncols=True\n",
     "    )\n",
-    "\n",
     "    for step, batch in enumerate(epoch_iterator):\n",
     "        step += 1\n",
-    "        x, y = (batch[\"image\"].cuda(device=device), batch[\"label\"].cuda(device=device))\n",
+    "        x, y = (batch[\"image\"].cuda(), batch[\"label\"].cuda())\n",
     "        with torch.cuda.amp.autocast():\n",
     "            logit_map = model(x)\n",
     "            loss = loss_function(logit_map, y)\n",
-    "        loss.backward()\n",
+    "        scaler.scale(loss).backward()\n",
     "        epoch_loss += loss.item()\n",
-    "        optimizer.step()\n",
+    "        scaler.unscale_(optimizer)\n",
+    "        scaler.step(optimizer)\n",
+    "        scaler.update()\n",
     "        optimizer.zero_grad()\n",
     "        epoch_iterator.set_description(\n",
     "            \"Training (%d / %d Steps) (loss=%2.5f)\"\n",
@@ -662,7 +524,6 @@
     "                val_loader, desc=\"Validate (X / X Steps) (dice=X.X)\", dynamic_ncols=True\n",
     "            )\n",
     "            dice_val = validation(epoch_iterator_val)\n",
-    "            # FIXME: epoch_loss is a running average at time of validation??\n",
     "            epoch_loss /= step\n",
     "            epoch_loss_values.append(epoch_loss)\n",
     "            metric_values.append(dice_val)\n",
@@ -689,1996 +550,101 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
+    "max_iterations = 30000\n",
+    "eval_num = 500\n",
     "post_label = AsDiscrete(to_onehot=14)\n",
     "post_pred = AsDiscrete(argmax=True, to_onehot=14)\n",
-    "dice_metric = DiceMetric(include_background=True, reduction=\"mean\", get_not_nans=False)"
+    "dice_metric = DiceMetric(include_background=True, reduction=\"mean\", get_not_nans=False)\n",
+    "global_step = 0\n",
+    "dice_val_best = 0.0\n",
+    "global_step_best = 0\n",
+    "epoch_loss_values = []\n",
+    "metric_values = []\n",
+    "while global_step < max_iterations:\n",
+    "    global_step, dice_val_best, global_step_best = train(\n",
+    "        global_step, train_loader, dice_val_best, global_step_best\n",
+    "    )\n",
+    "model.load_state_dict(torch.load(os.path.join(root_dir, \"best_metric_model.pth\")))"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 13,
-   "metadata": {
-    "scrolled": true
-   },
+   "execution_count": 11,
+   "metadata": {},
    "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (23 / 30000 Steps) (loss=3.04756): 100%|██████████| 24/24 [00:24<00:00,  1.02s/it]\n",
-      "Training (47 / 30000 Steps) (loss=2.90349): 100%|██████████| 24/24 [00:22<00:00,  1.08it/s]\n",
-      "Training (71 / 30000 Steps) (loss=2.80009): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (95 / 30000 Steps) (loss=2.73891): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (119 / 30000 Steps) (loss=2.75600): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (143 / 30000 Steps) (loss=2.81679): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (167 / 30000 Steps) (loss=2.73382): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (191 / 30000 Steps) (loss=2.70652): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (215 / 30000 Steps) (loss=2.58161): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (239 / 30000 Steps) (loss=2.55389): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (263 / 30000 Steps) (loss=2.59991): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (287 / 30000 Steps) (loss=2.57285): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (311 / 30000 Steps) (loss=2.59925): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (335 / 30000 Steps) (loss=2.44729): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (359 / 30000 Steps) (loss=2.47251): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (383 / 30000 Steps) (loss=2.34368): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (407 / 30000 Steps) (loss=2.28056): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (431 / 30000 Steps) (loss=2.38724): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (455 / 30000 Steps) (loss=2.35536): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (479 / 30000 Steps) (loss=2.31131): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (500 / 30000 Steps) (loss=2.27539):  83%|████████▎ | 20/24 [00:19<00:03,  1.08it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[ANone of the inputs have requires_grad=True. Gradients will be None\n",
-      "\n",
-      "Validate (480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]        \u001b[A\n",
-      "Validate (480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.27s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:13,  3.27s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.97s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.97s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "positional arguments and argument \"destination\" are deprecated. nn.Module.state_dict will not accept them in the future. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.\n",
-      "Training (500 / 30000 Steps) (loss=2.27539):  88%|████████▊ | 21/24 [00:41<00:22,  7.52s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.09675732254981995\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (503 / 30000 Steps) (loss=2.31343): 100%|██████████| 24/24 [00:43<00:00,  1.83s/it]\n",
-      "Training (527 / 30000 Steps) (loss=2.29317): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (551 / 30000 Steps) (loss=2.26501): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (575 / 30000 Steps) (loss=2.28498): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (599 / 30000 Steps) (loss=2.19028): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (623 / 30000 Steps) (loss=2.50455): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (647 / 30000 Steps) (loss=2.18401): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (671 / 30000 Steps) (loss=2.24277): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (695 / 30000 Steps) (loss=2.21543): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (719 / 30000 Steps) (loss=2.26871): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (743 / 30000 Steps) (loss=2.23642): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (767 / 30000 Steps) (loss=2.19016): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (791 / 30000 Steps) (loss=2.15235): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (815 / 30000 Steps) (loss=2.14225): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (839 / 30000 Steps) (loss=2.09030): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (863 / 30000 Steps) (loss=2.10239): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (887 / 30000 Steps) (loss=2.16179): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (911 / 30000 Steps) (loss=2.03907): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (935 / 30000 Steps) (loss=2.04785): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (959 / 30000 Steps) (loss=2.10109): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (983 / 30000 Steps) (loss=2.12406): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1000 / 30000 Steps) (loss=2.07229):  67%|██████▋   | 16/24 [00:15<00:07,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]        \u001b[A\n",
-      "Validate (984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (1000 / 30000 Steps) (loss=2.07229):  71%|███████   | 17/24 [00:36<00:50,  7.24s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.09206199645996094\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (1007 / 30000 Steps) (loss=1.96543): 100%|██████████| 24/24 [00:43<00:00,  1.79s/it]\n",
-      "Training (1031 / 30000 Steps) (loss=1.98335): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1055 / 30000 Steps) (loss=1.97615): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1079 / 30000 Steps) (loss=1.95188): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1103 / 30000 Steps) (loss=2.07248): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1127 / 30000 Steps) (loss=2.11488): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1151 / 30000 Steps) (loss=1.84554): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1175 / 30000 Steps) (loss=2.08018): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1199 / 30000 Steps) (loss=1.95251): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1223 / 30000 Steps) (loss=2.01169): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1247 / 30000 Steps) (loss=2.16851): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (1271 / 30000 Steps) (loss=2.17117): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1295 / 30000 Steps) (loss=1.99893): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1319 / 30000 Steps) (loss=2.00061): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1343 / 30000 Steps) (loss=1.94732): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1367 / 30000 Steps) (loss=1.99271): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1391 / 30000 Steps) (loss=2.02488): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1415 / 30000 Steps) (loss=1.95091): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1439 / 30000 Steps) (loss=2.07077): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1463 / 30000 Steps) (loss=1.87043): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1487 / 30000 Steps) (loss=2.07426): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1500 / 30000 Steps) (loss=2.03805):  50%|█████     | 12/24 [00:11<00:10,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (1488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (1488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (1500 / 30000 Steps) (loss=2.03805):  54%|█████▍    | 13/24 [00:32<01:20,  7.29s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059749215841293\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (1511 / 30000 Steps) (loss=1.97013): 100%|██████████| 24/24 [00:43<00:00,  1.79s/it]\n",
-      "Training (1535 / 30000 Steps) (loss=2.08701): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1559 / 30000 Steps) (loss=1.91518): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1583 / 30000 Steps) (loss=1.86013): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (1607 / 30000 Steps) (loss=1.92820): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1631 / 30000 Steps) (loss=1.95914): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1655 / 30000 Steps) (loss=2.31439): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (1679 / 30000 Steps) (loss=1.93297): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1703 / 30000 Steps) (loss=1.82793): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1727 / 30000 Steps) (loss=1.77602): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1751 / 30000 Steps) (loss=1.87049): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1775 / 30000 Steps) (loss=1.81319): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1799 / 30000 Steps) (loss=1.86025): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1823 / 30000 Steps) (loss=1.81045): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1847 / 30000 Steps) (loss=1.76598): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (1871 / 30000 Steps) (loss=1.93609): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1895 / 30000 Steps) (loss=1.85808): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1919 / 30000 Steps) (loss=1.87683): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1943 / 30000 Steps) (loss=1.79905): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (1967 / 30000 Steps) (loss=1.93705): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (1991 / 30000 Steps) (loss=1.87536): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2000 / 30000 Steps) (loss=1.84261):  33%|███▎      | 8/24 [00:08<00:14,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (1992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (1992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (1992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (1992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (1992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (1992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (1992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (1992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (1992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (1992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (1992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (1992 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (2000 / 30000 Steps) (loss=1.84261):  38%|███▊      | 9/24 [00:29<01:52,  7.49s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.08243301510810852\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (2015 / 30000 Steps) (loss=1.85733): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (2039 / 30000 Steps) (loss=1.97582): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (2063 / 30000 Steps) (loss=1.80809): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2087 / 30000 Steps) (loss=1.76823): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2111 / 30000 Steps) (loss=1.85991): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2135 / 30000 Steps) (loss=2.01875): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2159 / 30000 Steps) (loss=1.75839): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2183 / 30000 Steps) (loss=1.74449): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2207 / 30000 Steps) (loss=1.76575): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2231 / 30000 Steps) (loss=1.82603): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2255 / 30000 Steps) (loss=1.89225): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2279 / 30000 Steps) (loss=1.76695): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2303 / 30000 Steps) (loss=1.88593): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2327 / 30000 Steps) (loss=1.90981): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2351 / 30000 Steps) (loss=1.91762): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2375 / 30000 Steps) (loss=2.09532): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2399 / 30000 Steps) (loss=2.09043): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2423 / 30000 Steps) (loss=1.89411): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2447 / 30000 Steps) (loss=1.64777): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (2471 / 30000 Steps) (loss=1.70874): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2495 / 30000 Steps) (loss=1.97091): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2500 / 30000 Steps) (loss=1.70573):  17%|█▋        | 4/24 [00:04<00:18,  1.08it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (2496 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (2496 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.17s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.17s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.27s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:13,  3.27s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.97s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.97s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (2500 / 30000 Steps) (loss=1.70573):  21%|██        | 5/24 [00:25<02:41,  8.52s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07071128487586975\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (2519 / 30000 Steps) (loss=2.08651): 100%|██████████| 24/24 [00:43<00:00,  1.79s/it]\n",
-      "Training (2543 / 30000 Steps) (loss=1.68501): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2567 / 30000 Steps) (loss=1.68676): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2591 / 30000 Steps) (loss=1.98606): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2615 / 30000 Steps) (loss=1.88070): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (2639 / 30000 Steps) (loss=1.85338): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2663 / 30000 Steps) (loss=1.75137): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2687 / 30000 Steps) (loss=1.68586): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2711 / 30000 Steps) (loss=1.85745): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2735 / 30000 Steps) (loss=1.83789): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2759 / 30000 Steps) (loss=2.14781): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2783 / 30000 Steps) (loss=1.65742): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2807 / 30000 Steps) (loss=1.61873): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2831 / 30000 Steps) (loss=1.69797): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2855 / 30000 Steps) (loss=1.73362): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2879 / 30000 Steps) (loss=1.64782): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2903 / 30000 Steps) (loss=1.96968): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2927 / 30000 Steps) (loss=1.81555): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2951 / 30000 Steps) (loss=1.90872): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (2975 / 30000 Steps) (loss=1.74729): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (2999 / 30000 Steps) (loss=1.71663): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3000 / 30000 Steps) (loss=1.77936):   0%|          | 0/24 [00:00<?, ?it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (3000 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (3000 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (3000 / 30000 Steps) (loss=1.77936):   4%|▍         | 1/24 [00:21<08:25, 21.97s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07060686498880386\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (3023 / 30000 Steps) (loss=1.69800): 100%|██████████| 24/24 [00:43<00:00,  1.79s/it]\n",
-      "Training (3047 / 30000 Steps) (loss=1.95848): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3071 / 30000 Steps) (loss=1.87359): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3095 / 30000 Steps) (loss=1.77414): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3119 / 30000 Steps) (loss=1.63642): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3143 / 30000 Steps) (loss=1.91925): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3167 / 30000 Steps) (loss=1.59526): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (3191 / 30000 Steps) (loss=1.67294): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3215 / 30000 Steps) (loss=1.83990): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3239 / 30000 Steps) (loss=1.98160): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3263 / 30000 Steps) (loss=1.66625): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3287 / 30000 Steps) (loss=1.59335): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3311 / 30000 Steps) (loss=1.71648): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3335 / 30000 Steps) (loss=1.88376): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3359 / 30000 Steps) (loss=1.71496): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3383 / 30000 Steps) (loss=1.86785): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3407 / 30000 Steps) (loss=1.98294): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (3431 / 30000 Steps) (loss=1.75702): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (3455 / 30000 Steps) (loss=1.79603): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3479 / 30000 Steps) (loss=1.63082): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3500 / 30000 Steps) (loss=1.85942):  83%|████████▎ | 20/24 [00:19<00:03,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (3480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (3480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (3480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (3480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (3480 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (3480 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (3480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (3480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (3480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (3480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (3480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (3480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (3500 / 30000 Steps) (loss=1.85942):  88%|████████▊ | 21/24 [00:40<00:21,  7.21s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.0705963596701622\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (3503 / 30000 Steps) (loss=2.01710): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (3527 / 30000 Steps) (loss=1.84506): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3551 / 30000 Steps) (loss=1.62561): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3575 / 30000 Steps) (loss=1.63863): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3599 / 30000 Steps) (loss=1.87775): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3623 / 30000 Steps) (loss=1.74231): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3647 / 30000 Steps) (loss=1.68101): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3671 / 30000 Steps) (loss=1.91874): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3695 / 30000 Steps) (loss=1.53870): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3719 / 30000 Steps) (loss=1.74763): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3743 / 30000 Steps) (loss=1.55972): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (3767 / 30000 Steps) (loss=1.74384): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3791 / 30000 Steps) (loss=1.85877): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3815 / 30000 Steps) (loss=1.76684): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3839 / 30000 Steps) (loss=1.61158): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3863 / 30000 Steps) (loss=1.95114): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3887 / 30000 Steps) (loss=1.91905): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (3911 / 30000 Steps) (loss=1.72093): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (3935 / 30000 Steps) (loss=1.68328): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (3959 / 30000 Steps) (loss=2.00625): 100%|██████████| 24/24 [00:22<00:00,  1.08it/s]\n",
-      "Training (4055 / 30000 Steps) (loss=3.08001): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4079 / 30000 Steps) (loss=1.88033): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4103 / 30000 Steps) (loss=1.88240): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4127 / 30000 Steps) (loss=2.10009): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4151 / 30000 Steps) (loss=1.95376): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4175 / 30000 Steps) (loss=1.64926): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4199 / 30000 Steps) (loss=1.67962): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4223 / 30000 Steps) (loss=1.56884): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4247 / 30000 Steps) (loss=1.79777): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4271 / 30000 Steps) (loss=1.81146): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4295 / 30000 Steps) (loss=1.69637): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4319 / 30000 Steps) (loss=1.96981): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4343 / 30000 Steps) (loss=1.98609): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4367 / 30000 Steps) (loss=1.74551): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4391 / 30000 Steps) (loss=1.86034): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4415 / 30000 Steps) (loss=1.67974): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (4439 / 30000 Steps) (loss=1.91140): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4463 / 30000 Steps) (loss=1.49962): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4487 / 30000 Steps) (loss=1.69141): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4500 / 30000 Steps) (loss=1.59464):  50%|█████     | 12/24 [00:11<00:11,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (4488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (4488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (4488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (4488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.25s/it]\u001b[A\n",
-      "Validate (4488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.25s/it]\u001b[A\n",
-      "Validate (4488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.90s/it]\u001b[A\n",
-      "Validate (4488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.90s/it]\u001b[A\n",
-      "Validate (4488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (4488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (4488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (4488 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (4488 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
-      "Training (4500 / 30000 Steps) (loss=1.59464):  54%|█████▍    | 13/24 [00:32<01:20,  7.27s/it]"
-     ]
-    },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07118804007768631\n"
+      "train completed, best_metric: 0.8430 at iteration: 29500\n"
      ]
-    },
+    }
+   ],
+   "source": [
+    "print(\n",
+    "    f\"train completed, best_metric: {dice_val_best:.4f} \"\n",
+    "    f\"at iteration: {global_step_best}\"\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Plot the loss and metric"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "metadata": {},
+   "outputs": [
     {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (4511 / 30000 Steps) (loss=1.71029): 100%|██████████| 24/24 [00:43<00:00,  1.79s/it]\n",
-      "Training (4535 / 30000 Steps) (loss=1.62508): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4559 / 30000 Steps) (loss=2.17477): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4583 / 30000 Steps) (loss=1.88658): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4607 / 30000 Steps) (loss=1.80879): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4631 / 30000 Steps) (loss=1.79553): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4655 / 30000 Steps) (loss=1.68824): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4679 / 30000 Steps) (loss=3.31505): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4703 / 30000 Steps) (loss=1.69691): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4727 / 30000 Steps) (loss=1.76048): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4751 / 30000 Steps) (loss=1.83267): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4775 / 30000 Steps) (loss=1.76414): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4799 / 30000 Steps) (loss=2.53666): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4823 / 30000 Steps) (loss=2.14007): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4847 / 30000 Steps) (loss=1.74145): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4871 / 30000 Steps) (loss=2.61676): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4895 / 30000 Steps) (loss=1.81788): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4919 / 30000 Steps) (loss=1.57590): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4943 / 30000 Steps) (loss=2.00811): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (4967 / 30000 Steps) (loss=1.80384): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (4991 / 30000 Steps) (loss=1.90285): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (5000 / 30000 Steps) (loss=1.88040):  33%|███▎      | 8/24 [00:08<00:14,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (4992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (4992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.24s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.24s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.90s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.90s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.94s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.94s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.65s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.65s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.49s/it]\u001b[A\n",
-      "Training (5000 / 30000 Steps) (loss=1.88040):  38%|███▊      | 9/24 [00:29<01:51,  7.46s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.0705961063504219\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (5015 / 30000 Steps) (loss=1.90351): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (5039 / 30000 Steps) (loss=1.82040): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (5063 / 30000 Steps) (loss=1.98175): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (5087 / 30000 Steps) (loss=1.83518): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (5111 / 30000 Steps) (loss=1.94957): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (5135 / 30000 Steps) (loss=2.21675): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (5159 / 30000 Steps) (loss=1.65691): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (5183 / 30000 Steps) (loss=2.06594): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (5207 / 30000 Steps) (loss=1.71019): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (5231 / 30000 Steps) (loss=1.69909): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (5255 / 30000 Steps) (loss=1.78258): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (5279 / 30000 Steps) (loss=1.83680): 100%|██████████| 24/24 [00:22<00:00,  1.08it/s]\n",
-      "Training (5303 / 30000 Steps) (loss=1.72917): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (5327 / 30000 Steps) (loss=1.78574): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (5351 / 30000 Steps) (loss=1.66677): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (5375 / 30000 Steps) (loss=1.91206): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (5399 / 30000 Steps) (loss=1.62288): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (5423 / 30000 Steps) (loss=1.64085): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (5425 / 30000 Steps) (loss=2.11989):   8%|▊         | 2/24 [00:01<00:20,  1.08it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (6719 / 30000 Steps) (loss=2.10010): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6743 / 30000 Steps) (loss=2.14659): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6767 / 30000 Steps) (loss=2.25052): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6791 / 30000 Steps) (loss=2.32982): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6815 / 30000 Steps) (loss=2.88283): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6839 / 30000 Steps) (loss=2.16736): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6863 / 30000 Steps) (loss=2.30831): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6887 / 30000 Steps) (loss=2.07512): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6911 / 30000 Steps) (loss=2.20173): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6935 / 30000 Steps) (loss=2.10239): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6959 / 30000 Steps) (loss=2.10825): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (6983 / 30000 Steps) (loss=2.14991): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7000 / 30000 Steps) (loss=2.20176):  67%|██████▋   | 16/24 [00:15<00:07,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (6984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (6984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (6984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (6984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (6984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (6984 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (6984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (6984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (6984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (6984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (6984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (6984 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (7000 / 30000 Steps) (loss=2.20176):  71%|███████   | 17/24 [00:36<00:50,  7.23s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (7007 / 30000 Steps) (loss=2.22435): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (7031 / 30000 Steps) (loss=1.94915): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (7055 / 30000 Steps) (loss=2.10985): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7079 / 30000 Steps) (loss=3.61836): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7103 / 30000 Steps) (loss=2.17272): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7127 / 30000 Steps) (loss=2.11024): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7151 / 30000 Steps) (loss=2.18739): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7175 / 30000 Steps) (loss=1.98527): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7199 / 30000 Steps) (loss=2.06582): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7223 / 30000 Steps) (loss=1.85681): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7247 / 30000 Steps) (loss=2.20366): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7271 / 30000 Steps) (loss=2.27249): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7295 / 30000 Steps) (loss=2.02298): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7319 / 30000 Steps) (loss=2.12844): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7343 / 30000 Steps) (loss=2.19865): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7367 / 30000 Steps) (loss=1.98583): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7391 / 30000 Steps) (loss=1.90731): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7415 / 30000 Steps) (loss=2.00232): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7439 / 30000 Steps) (loss=2.11610): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7463 / 30000 Steps) (loss=2.11672): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7487 / 30000 Steps) (loss=1.97471): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7500 / 30000 Steps) (loss=2.09074):  50%|█████     | 12/24 [00:11<00:11,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (7488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (7488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (7500 / 30000 Steps) (loss=2.09074):  54%|█████▍    | 13/24 [00:32<01:20,  7.29s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (7511 / 30000 Steps) (loss=1.93107): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (7535 / 30000 Steps) (loss=2.07361): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7559 / 30000 Steps) (loss=1.73974): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (7583 / 30000 Steps) (loss=1.86634): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7607 / 30000 Steps) (loss=1.71773): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7631 / 30000 Steps) (loss=2.12695): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7655 / 30000 Steps) (loss=2.08464): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7679 / 30000 Steps) (loss=1.89307): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7703 / 30000 Steps) (loss=2.23980): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7727 / 30000 Steps) (loss=2.06015): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7751 / 30000 Steps) (loss=1.92099): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7775 / 30000 Steps) (loss=1.92985): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7799 / 30000 Steps) (loss=1.91172): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7823 / 30000 Steps) (loss=1.72619): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7847 / 30000 Steps) (loss=1.83092): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (7871 / 30000 Steps) (loss=1.85543): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7895 / 30000 Steps) (loss=2.39690): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7919 / 30000 Steps) (loss=2.06412): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7943 / 30000 Steps) (loss=2.02304): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7967 / 30000 Steps) (loss=1.82633): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (7991 / 30000 Steps) (loss=1.97773): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (8000 / 30000 Steps) (loss=1.75278):  33%|███▎      | 8/24 [00:08<00:14,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (7992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (7992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (8000 / 30000 Steps) (loss=1.75278):  38%|███▊      | 9/24 [00:29<01:52,  7.48s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (8015 / 30000 Steps) (loss=1.83981): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (8039 / 30000 Steps) (loss=1.89129): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (8061 / 30000 Steps) (loss=1.97249):  92%|█████████▏| 22/24 [00:20<00:01,  1.10it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (9191 / 30000 Steps) (loss=1.82559): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9215 / 30000 Steps) (loss=2.22664): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9239 / 30000 Steps) (loss=3.52945): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9263 / 30000 Steps) (loss=1.69937): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9287 / 30000 Steps) (loss=1.96457): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9311 / 30000 Steps) (loss=2.37476): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9335 / 30000 Steps) (loss=3.65335): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9359 / 30000 Steps) (loss=2.01223): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9383 / 30000 Steps) (loss=1.98225): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9407 / 30000 Steps) (loss=1.96237): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9431 / 30000 Steps) (loss=2.09754): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9455 / 30000 Steps) (loss=2.20548): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9479 / 30000 Steps) (loss=1.84057): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9500 / 30000 Steps) (loss=1.92208):  83%|████████▎ | 20/24 [00:19<00:03,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (9480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (9480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (9480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (9480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.25s/it]\u001b[A\n",
-      "Validate (9480 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.25s/it]\u001b[A\n",
-      "Validate (9480 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (9480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (9480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (9480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (9480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (9480 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (9480 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
-      "Training (9500 / 30000 Steps) (loss=1.92208):  88%|████████▊ | 21/24 [00:40<00:21,  7.21s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (9503 / 30000 Steps) (loss=2.06457): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (9527 / 30000 Steps) (loss=1.92062): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9551 / 30000 Steps) (loss=1.92944): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (9575 / 30000 Steps) (loss=1.87242): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (9599 / 30000 Steps) (loss=2.29003): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9623 / 30000 Steps) (loss=2.05942): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9647 / 30000 Steps) (loss=2.01496): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9671 / 30000 Steps) (loss=2.16334): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9695 / 30000 Steps) (loss=2.43831): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9719 / 30000 Steps) (loss=1.83332): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9743 / 30000 Steps) (loss=1.79377): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9767 / 30000 Steps) (loss=1.83020): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9791 / 30000 Steps) (loss=1.81070): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (9815 / 30000 Steps) (loss=1.98005): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9839 / 30000 Steps) (loss=1.87129): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9863 / 30000 Steps) (loss=1.97455): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9887 / 30000 Steps) (loss=1.85374): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9911 / 30000 Steps) (loss=2.03599): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9935 / 30000 Steps) (loss=1.99218): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9959 / 30000 Steps) (loss=1.84849): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (9983 / 30000 Steps) (loss=4.62641): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10000 / 30000 Steps) (loss=2.10157):  67%|██████▋   | 16/24 [00:15<00:07,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (9984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]       \u001b[A\n",
-      "Validate (9984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (10000 / 30000 Steps) (loss=2.10157):  71%|███████   | 17/24 [00:36<00:50,  7.23s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059746980667114\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (10007 / 30000 Steps) (loss=2.04821): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (10031 / 30000 Steps) (loss=2.15589): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10055 / 30000 Steps) (loss=2.04477): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10079 / 30000 Steps) (loss=2.52525): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10103 / 30000 Steps) (loss=1.87341): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (10127 / 30000 Steps) (loss=2.04750): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10151 / 30000 Steps) (loss=1.67941): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10175 / 30000 Steps) (loss=2.09627): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10199 / 30000 Steps) (loss=2.10325): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10223 / 30000 Steps) (loss=1.72772): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10247 / 30000 Steps) (loss=1.73709): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10271 / 30000 Steps) (loss=2.13825): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10295 / 30000 Steps) (loss=2.38076): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (10319 / 30000 Steps) (loss=1.99312): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10343 / 30000 Steps) (loss=1.91718): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10367 / 30000 Steps) (loss=1.73751): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10391 / 30000 Steps) (loss=2.01234): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10415 / 30000 Steps) (loss=2.13904): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (10439 / 30000 Steps) (loss=1.90831): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10463 / 30000 Steps) (loss=1.90202): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10487 / 30000 Steps) (loss=1.94877): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (10500 / 30000 Steps) (loss=2.07265):  50%|█████     | 12/24 [00:11<00:10,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (10488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (10488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
-      "Training (10500 / 30000 Steps) (loss=2.07265):  54%|█████▍    | 13/24 [00:32<01:20,  7.29s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059737294912338\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (10511 / 30000 Steps) (loss=1.78768): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (10534 / 30000 Steps) (loss=1.97501):  96%|█████████▌| 23/24 [00:21<00:00,  1.10it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (11807 / 30000 Steps) (loss=1.89524): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (11831 / 30000 Steps) (loss=1.74966): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (11855 / 30000 Steps) (loss=2.43224): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (11879 / 30000 Steps) (loss=1.67055): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (11903 / 30000 Steps) (loss=1.77441): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (11927 / 30000 Steps) (loss=2.03059): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (11951 / 30000 Steps) (loss=1.83695): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (11975 / 30000 Steps) (loss=1.91604): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (11999 / 30000 Steps) (loss=2.13925): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12000 / 30000 Steps) (loss=3.23982):   0%|          | 0/24 [00:00<?, ?it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (12000 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (12000 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (12000 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (12000 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (12000 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (12000 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (12000 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (12000 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (12000 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (12000 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (12000 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (12000 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (12000 / 30000 Steps) (loss=3.23982):   4%|▍         | 1/24 [00:22<08:26, 22.00s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (12023 / 30000 Steps) (loss=1.82447): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (12047 / 30000 Steps) (loss=1.82887): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12071 / 30000 Steps) (loss=1.52439): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12095 / 30000 Steps) (loss=1.87138): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12119 / 30000 Steps) (loss=1.97862): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12143 / 30000 Steps) (loss=1.74948): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12167 / 30000 Steps) (loss=1.78102): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12191 / 30000 Steps) (loss=1.85116): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12215 / 30000 Steps) (loss=1.82716): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12239 / 30000 Steps) (loss=1.71390): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12263 / 30000 Steps) (loss=1.86355): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (12287 / 30000 Steps) (loss=1.94163): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12311 / 30000 Steps) (loss=1.95592): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12335 / 30000 Steps) (loss=1.90674): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12359 / 30000 Steps) (loss=1.51387): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12383 / 30000 Steps) (loss=1.79401): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12407 / 30000 Steps) (loss=2.12846): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12431 / 30000 Steps) (loss=1.88363): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12455 / 30000 Steps) (loss=1.55907): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12479 / 30000 Steps) (loss=1.92450): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12500 / 30000 Steps) (loss=1.80316):  83%|████████▎ | 20/24 [00:19<00:03,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (12480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (12480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (12500 / 30000 Steps) (loss=1.80316):  88%|████████▊ | 21/24 [00:40<00:21,  7.24s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (12503 / 30000 Steps) (loss=1.96900): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (12527 / 30000 Steps) (loss=1.71932): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12551 / 30000 Steps) (loss=2.44105): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12575 / 30000 Steps) (loss=1.82854): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12599 / 30000 Steps) (loss=1.87003): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12623 / 30000 Steps) (loss=1.98144): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12647 / 30000 Steps) (loss=1.74120): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (12671 / 30000 Steps) (loss=1.82985): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12695 / 30000 Steps) (loss=3.63615): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12719 / 30000 Steps) (loss=1.61349): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12743 / 30000 Steps) (loss=1.66895): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12767 / 30000 Steps) (loss=2.05730): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12791 / 30000 Steps) (loss=3.24769): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12815 / 30000 Steps) (loss=1.84073): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12839 / 30000 Steps) (loss=1.83791): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12863 / 30000 Steps) (loss=1.63371): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12887 / 30000 Steps) (loss=1.97341): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (12911 / 30000 Steps) (loss=1.69248): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12935 / 30000 Steps) (loss=1.78743): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (12959 / 30000 Steps) (loss=1.92369): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (12983 / 30000 Steps) (loss=1.73248): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (13000 / 30000 Steps) (loss=1.85713):  67%|██████▋   | 16/24 [00:15<00:07,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (12984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (12984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
-      "Training (13000 / 30000 Steps) (loss=1.85713):  71%|███████   | 17/24 [00:36<00:50,  7.22s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (13007 / 30000 Steps) (loss=1.91395): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (13031 / 30000 Steps) (loss=1.99260): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (13055 / 30000 Steps) (loss=1.97792): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (13079 / 30000 Steps) (loss=1.78542): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (13103 / 30000 Steps) (loss=1.85659): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (13127 / 30000 Steps) (loss=1.92632): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (13151 / 30000 Steps) (loss=1.81754): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (13152 / 30000 Steps) (loss=1.89086):   0%|          | 0/24 [00:00<?, ?it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (14279 / 30000 Steps) (loss=1.97302): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14303 / 30000 Steps) (loss=1.76956): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14327 / 30000 Steps) (loss=1.62204): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14351 / 30000 Steps) (loss=2.00075): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14375 / 30000 Steps) (loss=2.04387): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14399 / 30000 Steps) (loss=1.80620): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14423 / 30000 Steps) (loss=1.75732): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14447 / 30000 Steps) (loss=1.97194): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14471 / 30000 Steps) (loss=1.96930): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14495 / 30000 Steps) (loss=1.86121): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14500 / 30000 Steps) (loss=2.08860):  17%|█▋        | 4/24 [00:04<00:18,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (14496 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (14496 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (14496 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (14496 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (14496 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (14496 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (14496 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (14496 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (14496 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (14496 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (14496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (14496 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (14500 / 30000 Steps) (loss=2.08860):  21%|██        | 5/24 [00:25<02:41,  8.50s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07063417136669159\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (14519 / 30000 Steps) (loss=1.91196): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (14543 / 30000 Steps) (loss=2.08129): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14567 / 30000 Steps) (loss=2.49496): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14591 / 30000 Steps) (loss=2.31298): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14615 / 30000 Steps) (loss=1.64771): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14639 / 30000 Steps) (loss=2.52957): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (14663 / 30000 Steps) (loss=1.84937): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14687 / 30000 Steps) (loss=1.85760): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14711 / 30000 Steps) (loss=1.72808): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14735 / 30000 Steps) (loss=2.03447): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14759 / 30000 Steps) (loss=1.89378): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14783 / 30000 Steps) (loss=2.67441): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14807 / 30000 Steps) (loss=1.97396): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14831 / 30000 Steps) (loss=1.78090): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14855 / 30000 Steps) (loss=3.55123): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14879 / 30000 Steps) (loss=1.85744): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14903 / 30000 Steps) (loss=2.00595): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14927 / 30000 Steps) (loss=1.93430): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14951 / 30000 Steps) (loss=2.02877): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14975 / 30000 Steps) (loss=1.90165): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (14999 / 30000 Steps) (loss=2.20059): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15000 / 30000 Steps) (loss=1.80541):   0%|          | 0/24 [00:01<?, ?it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (15000 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (15000 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (15000 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (15000 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (15000 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (15000 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (15000 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (15000 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (15000 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (15000 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (15000 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (15000 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (15000 / 30000 Steps) (loss=1.80541):   4%|▍         | 1/24 [00:22<08:26, 22.04s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07060018181800842\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (15023 / 30000 Steps) (loss=1.99871): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (15047 / 30000 Steps) (loss=1.96666): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15071 / 30000 Steps) (loss=1.94317): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15095 / 30000 Steps) (loss=1.76293): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15119 / 30000 Steps) (loss=1.61266): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15143 / 30000 Steps) (loss=1.75006): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15167 / 30000 Steps) (loss=1.99979): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15191 / 30000 Steps) (loss=1.85684): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15215 / 30000 Steps) (loss=1.82103): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15239 / 30000 Steps) (loss=1.78026): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15263 / 30000 Steps) (loss=1.94322): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15287 / 30000 Steps) (loss=1.73688): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15311 / 30000 Steps) (loss=1.90172): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15335 / 30000 Steps) (loss=1.83768): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15359 / 30000 Steps) (loss=2.09235): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15383 / 30000 Steps) (loss=2.08885): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15407 / 30000 Steps) (loss=1.95160): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15431 / 30000 Steps) (loss=1.46536): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15455 / 30000 Steps) (loss=2.30151): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15479 / 30000 Steps) (loss=1.89452): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15500 / 30000 Steps) (loss=1.79847):  83%|████████▎ | 20/24 [00:19<00:03,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (15480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (15480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (15500 / 30000 Steps) (loss=1.79847):  88%|████████▊ | 21/24 [00:40<00:21,  7.23s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059748470783234\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (15503 / 30000 Steps) (loss=2.06165): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (15527 / 30000 Steps) (loss=1.82561): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15551 / 30000 Steps) (loss=1.95241): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15575 / 30000 Steps) (loss=1.67307): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15599 / 30000 Steps) (loss=2.27193): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (15614 / 30000 Steps) (loss=1.92388):  58%|█████▊    | 14/24 [00:13<00:09,  1.09it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (16871 / 30000 Steps) (loss=2.05397): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (16895 / 30000 Steps) (loss=1.79339): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (16919 / 30000 Steps) (loss=1.74023): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (16943 / 30000 Steps) (loss=1.65611): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (16967 / 30000 Steps) (loss=2.20335): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (16991 / 30000 Steps) (loss=1.91431): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (17000 / 30000 Steps) (loss=1.93411):  33%|███▎      | 8/24 [00:08<00:14,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (16992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (16992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (17000 / 30000 Steps) (loss=1.93411):  38%|███▊      | 9/24 [00:29<01:52,  7.48s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (17015 / 30000 Steps) (loss=1.51199): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (17039 / 30000 Steps) (loss=1.86803): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17063 / 30000 Steps) (loss=2.14775): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (17087 / 30000 Steps) (loss=1.87431): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17111 / 30000 Steps) (loss=2.15638): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17135 / 30000 Steps) (loss=1.84681): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17159 / 30000 Steps) (loss=1.77586): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17183 / 30000 Steps) (loss=2.58149): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17207 / 30000 Steps) (loss=2.62441): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17231 / 30000 Steps) (loss=2.00080): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (17255 / 30000 Steps) (loss=1.97522): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17279 / 30000 Steps) (loss=1.94671): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17303 / 30000 Steps) (loss=2.10150): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17327 / 30000 Steps) (loss=1.86351): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17351 / 30000 Steps) (loss=1.76649): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17375 / 30000 Steps) (loss=1.85833): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17399 / 30000 Steps) (loss=1.76396): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17423 / 30000 Steps) (loss=1.83632): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17447 / 30000 Steps) (loss=2.37074): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17471 / 30000 Steps) (loss=1.94557): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (17495 / 30000 Steps) (loss=1.82334): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17500 / 30000 Steps) (loss=2.04853):  17%|█▋        | 4/24 [00:04<00:18,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (17496 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (17496 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (17500 / 30000 Steps) (loss=2.04853):  21%|██        | 5/24 [00:25<02:41,  8.48s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (17519 / 30000 Steps) (loss=1.97825): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (17543 / 30000 Steps) (loss=1.77686): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17567 / 30000 Steps) (loss=2.08660): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17591 / 30000 Steps) (loss=1.69114): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17615 / 30000 Steps) (loss=3.15618): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17639 / 30000 Steps) (loss=2.46837): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17663 / 30000 Steps) (loss=1.90801): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17687 / 30000 Steps) (loss=1.86398): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17711 / 30000 Steps) (loss=2.52838): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17735 / 30000 Steps) (loss=1.85488): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17759 / 30000 Steps) (loss=1.79358): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17783 / 30000 Steps) (loss=1.85657): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17807 / 30000 Steps) (loss=1.57969): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (17831 / 30000 Steps) (loss=1.76049): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17855 / 30000 Steps) (loss=2.58910): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17879 / 30000 Steps) (loss=1.50078): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17903 / 30000 Steps) (loss=2.21724): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17927 / 30000 Steps) (loss=1.79426): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (17951 / 30000 Steps) (loss=1.57439): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17975 / 30000 Steps) (loss=3.44161): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (17999 / 30000 Steps) (loss=1.85103): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (18000 / 30000 Steps) (loss=2.06568):   0%|          | 0/24 [00:00<?, ?it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (18000 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (18000 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (18000 / 30000 Steps) (loss=2.06568):   4%|▍         | 1/24 [00:21<08:24, 21.94s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059751451015472\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (18023 / 30000 Steps) (loss=1.91046): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (18047 / 30000 Steps) (loss=1.77376): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (18071 / 30000 Steps) (loss=1.74509): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (18095 / 30000 Steps) (loss=1.92515): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (18119 / 30000 Steps) (loss=2.01573): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (18143 / 30000 Steps) (loss=1.78100): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (18167 / 30000 Steps) (loss=2.65431): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (18191 / 30000 Steps) (loss=3.30308): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (18215 / 30000 Steps) (loss=2.24336): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (18221 / 30000 Steps) (loss=2.80043):  25%|██▌       | 6/24 [00:05<00:16,  1.09it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (19487 / 30000 Steps) (loss=1.84788): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19500 / 30000 Steps) (loss=1.99503):  50%|█████     | 12/24 [00:11<00:10,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (19488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (19488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.25s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.25s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (19500 / 30000 Steps) (loss=1.99503):  54%|█████▍    | 13/24 [00:32<01:19,  7.27s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059739530086517\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (19511 / 30000 Steps) (loss=2.15546): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (19535 / 30000 Steps) (loss=2.16265): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19559 / 30000 Steps) (loss=2.11639): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19583 / 30000 Steps) (loss=2.04579): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19607 / 30000 Steps) (loss=1.67668): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19631 / 30000 Steps) (loss=2.25905): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19655 / 30000 Steps) (loss=1.94317): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19679 / 30000 Steps) (loss=1.69574): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19703 / 30000 Steps) (loss=1.79305): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19727 / 30000 Steps) (loss=1.96530): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (19751 / 30000 Steps) (loss=2.17987): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (19775 / 30000 Steps) (loss=2.07103): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (19799 / 30000 Steps) (loss=1.85445): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19823 / 30000 Steps) (loss=2.28405): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19847 / 30000 Steps) (loss=2.37703): 100%|██████████| 24/24 [00:22<00:00,  1.09it/s]\n",
-      "Training (19871 / 30000 Steps) (loss=1.48384): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19895 / 30000 Steps) (loss=1.98135): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19919 / 30000 Steps) (loss=1.64713): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (19943 / 30000 Steps) (loss=2.05469): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19967 / 30000 Steps) (loss=1.93943): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (19991 / 30000 Steps) (loss=1.64823): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20000 / 30000 Steps) (loss=1.65539):  33%|███▎      | 8/24 [00:08<00:14,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (19992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (19992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
-      "Training (20000 / 30000 Steps) (loss=1.65539):  38%|███▊      | 9/24 [00:29<01:52,  7.48s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (20015 / 30000 Steps) (loss=1.60423): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (20039 / 30000 Steps) (loss=1.54916): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20063 / 30000 Steps) (loss=2.57518): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20087 / 30000 Steps) (loss=1.74062): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20111 / 30000 Steps) (loss=1.84239): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (20135 / 30000 Steps) (loss=2.03068): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20159 / 30000 Steps) (loss=1.89472): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20183 / 30000 Steps) (loss=1.70734): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20207 / 30000 Steps) (loss=2.13889): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20231 / 30000 Steps) (loss=1.86293): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20255 / 30000 Steps) (loss=2.21442): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20279 / 30000 Steps) (loss=1.67967): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20303 / 30000 Steps) (loss=1.61414): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20327 / 30000 Steps) (loss=1.90312): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20351 / 30000 Steps) (loss=2.74694): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20375 / 30000 Steps) (loss=1.65601): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (20399 / 30000 Steps) (loss=1.87255): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20423 / 30000 Steps) (loss=1.80009): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20447 / 30000 Steps) (loss=2.14480): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20471 / 30000 Steps) (loss=1.98892): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20495 / 30000 Steps) (loss=1.86035): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20500 / 30000 Steps) (loss=1.84063):  17%|█▋        | 4/24 [00:04<00:18,  1.08it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (20496 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (20496 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
-      "Training (20500 / 30000 Steps) (loss=1.84063):  21%|██        | 5/24 [00:25<02:41,  8.49s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059755176305771\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (20519 / 30000 Steps) (loss=2.33358): 100%|██████████| 24/24 [00:42<00:00,  1.79s/it]\n",
-      "Training (20543 / 30000 Steps) (loss=2.00746): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20567 / 30000 Steps) (loss=1.80294): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20591 / 30000 Steps) (loss=1.91813): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20615 / 30000 Steps) (loss=1.82094): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20639 / 30000 Steps) (loss=2.03983): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20663 / 30000 Steps) (loss=1.80291): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (20687 / 30000 Steps) (loss=2.07999): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (20711 / 30000 Steps) (loss=1.92146): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20735 / 30000 Steps) (loss=1.62822): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (20759 / 30000 Steps) (loss=2.19601): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20783 / 30000 Steps) (loss=1.59426): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (20807 / 30000 Steps) (loss=1.82663): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (20831 / 30000 Steps) (loss=1.95577): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (22079 / 30000 Steps) (loss=1.67489): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22103 / 30000 Steps) (loss=1.73300): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22127 / 30000 Steps) (loss=1.97787): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22151 / 30000 Steps) (loss=1.66715): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22175 / 30000 Steps) (loss=1.68178): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22199 / 30000 Steps) (loss=1.76948): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22223 / 30000 Steps) (loss=1.75758): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22247 / 30000 Steps) (loss=1.89182): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22271 / 30000 Steps) (loss=1.90641): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (22295 / 30000 Steps) (loss=2.13553): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22319 / 30000 Steps) (loss=2.31815): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22343 / 30000 Steps) (loss=1.51507): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22367 / 30000 Steps) (loss=1.76590): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22391 / 30000 Steps) (loss=2.07999): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22415 / 30000 Steps) (loss=4.97741): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22439 / 30000 Steps) (loss=1.50233): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22463 / 30000 Steps) (loss=1.63693): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22487 / 30000 Steps) (loss=2.01772): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22500 / 30000 Steps) (loss=1.91705):  50%|█████     | 12/24 [00:11<00:10,  1.11it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (22488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (22488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.16s/it]\u001b[A\n",
-      "Validate (22488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.16s/it]\u001b[A\n",
-      "Validate (22488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.27s/it]\u001b[A\n",
-      "Validate (22488 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:13,  3.27s/it]\u001b[A\n",
-      "Validate (22488 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (22488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (22488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (22488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (22488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (22488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (22488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (22500 / 30000 Steps) (loss=1.91705):  54%|█████▍    | 13/24 [00:32<01:20,  7.29s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07064937800168991\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (22511 / 30000 Steps) (loss=1.68698): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (22535 / 30000 Steps) (loss=1.63871): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22559 / 30000 Steps) (loss=2.25367): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22583 / 30000 Steps) (loss=2.06416): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22607 / 30000 Steps) (loss=1.83227): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22631 / 30000 Steps) (loss=5.85255): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22655 / 30000 Steps) (loss=1.60104): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (22679 / 30000 Steps) (loss=1.69582): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22703 / 30000 Steps) (loss=1.99134): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22727 / 30000 Steps) (loss=2.05481): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22751 / 30000 Steps) (loss=2.27661): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22775 / 30000 Steps) (loss=3.47190): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (22799 / 30000 Steps) (loss=1.81042): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22823 / 30000 Steps) (loss=1.93108): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22847 / 30000 Steps) (loss=2.04849): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22871 / 30000 Steps) (loss=1.81162): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22895 / 30000 Steps) (loss=1.57702): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22919 / 30000 Steps) (loss=1.56488): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22943 / 30000 Steps) (loss=1.66431): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22967 / 30000 Steps) (loss=1.77493): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (22991 / 30000 Steps) (loss=1.85323): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23000 / 30000 Steps) (loss=2.16792):  33%|███▎      | 8/24 [00:08<00:14,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (22992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (22992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (22992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.15s/it]\u001b[A\n",
-      "Validate (22992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (22992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (22992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (22992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (22992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (22992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (22992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (22992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (22992 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (23000 / 30000 Steps) (loss=2.16792):  38%|███▊      | 9/24 [00:29<01:52,  7.48s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07060033828020096\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (23015 / 30000 Steps) (loss=2.04952): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (23039 / 30000 Steps) (loss=2.40743): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23063 / 30000 Steps) (loss=2.29741): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23087 / 30000 Steps) (loss=2.12759): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23111 / 30000 Steps) (loss=1.70525): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (23135 / 30000 Steps) (loss=1.76106): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23159 / 30000 Steps) (loss=1.72826): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23183 / 30000 Steps) (loss=1.72953): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23207 / 30000 Steps) (loss=1.72956): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23231 / 30000 Steps) (loss=2.05098): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23255 / 30000 Steps) (loss=1.78883): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23279 / 30000 Steps) (loss=1.72608): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23303 / 30000 Steps) (loss=1.86829): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23327 / 30000 Steps) (loss=1.71274): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23351 / 30000 Steps) (loss=1.66257): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23375 / 30000 Steps) (loss=1.82341): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23399 / 30000 Steps) (loss=2.08193): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23423 / 30000 Steps) (loss=1.92278): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (23445 / 30000 Steps) (loss=1.52153):  88%|████████▊ | 21/24 [00:19<00:02,  1.10it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (24695 / 30000 Steps) (loss=2.01804): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (24719 / 30000 Steps) (loss=2.03579): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (24743 / 30000 Steps) (loss=2.46138): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (24767 / 30000 Steps) (loss=1.75445): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (24791 / 30000 Steps) (loss=1.94900): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (24815 / 30000 Steps) (loss=1.49325): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (24839 / 30000 Steps) (loss=1.83015): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (24863 / 30000 Steps) (loss=2.09343): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (24887 / 30000 Steps) (loss=1.76656): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (24911 / 30000 Steps) (loss=1.70951): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (24935 / 30000 Steps) (loss=1.51334): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (24959 / 30000 Steps) (loss=1.63255): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (24983 / 30000 Steps) (loss=1.96455): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25000 / 30000 Steps) (loss=1.98859):  67%|██████▋   | 16/24 [00:15<00:07,  1.11it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (24984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (24984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (24984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (24984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (24984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (24984 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (24984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (24984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (24984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (24984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (24984 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (24984 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
-      "Training (25000 / 30000 Steps) (loss=1.98859):  71%|███████   | 17/24 [00:36<00:50,  7.21s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.0705978125333786\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (25007 / 30000 Steps) (loss=2.22888): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (25031 / 30000 Steps) (loss=2.12894): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25055 / 30000 Steps) (loss=1.67851): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25079 / 30000 Steps) (loss=4.27627): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25103 / 30000 Steps) (loss=2.11039): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25127 / 30000 Steps) (loss=1.83662): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25151 / 30000 Steps) (loss=1.89494): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25175 / 30000 Steps) (loss=2.07800): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (25199 / 30000 Steps) (loss=1.71037): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25223 / 30000 Steps) (loss=1.66549): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (25247 / 30000 Steps) (loss=1.62141): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25271 / 30000 Steps) (loss=1.73209): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25295 / 30000 Steps) (loss=2.35269): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (25319 / 30000 Steps) (loss=1.67006): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25343 / 30000 Steps) (loss=2.01230): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25367 / 30000 Steps) (loss=1.53763): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25391 / 30000 Steps) (loss=1.90676): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25415 / 30000 Steps) (loss=1.87406): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25439 / 30000 Steps) (loss=1.91836): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25463 / 30000 Steps) (loss=1.67455): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25487 / 30000 Steps) (loss=2.42163): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25500 / 30000 Steps) (loss=1.73544):  50%|█████     | 12/24 [00:11<00:10,  1.09it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (25488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (25488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (25488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (25488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (25488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.25s/it]\u001b[A\n",
-      "Validate (25488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (25488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (25488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (25488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (25488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (25488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (25488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (25500 / 30000 Steps) (loss=1.73544):  54%|█████▍    | 13/24 [00:32<01:20,  7.28s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07065335661172867\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (25511 / 30000 Steps) (loss=1.98163): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (25535 / 30000 Steps) (loss=1.90531): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25559 / 30000 Steps) (loss=1.66735): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (25583 / 30000 Steps) (loss=1.74367): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25607 / 30000 Steps) (loss=2.38333): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25631 / 30000 Steps) (loss=1.92686): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25655 / 30000 Steps) (loss=2.20560): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25679 / 30000 Steps) (loss=2.16599): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25703 / 30000 Steps) (loss=1.76526): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25727 / 30000 Steps) (loss=2.01786): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25751 / 30000 Steps) (loss=1.74469): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25775 / 30000 Steps) (loss=2.09772): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25799 / 30000 Steps) (loss=1.73560): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25823 / 30000 Steps) (loss=1.76547): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25847 / 30000 Steps) (loss=1.88350): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25871 / 30000 Steps) (loss=1.72797): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25895 / 30000 Steps) (loss=1.92903): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (25919 / 30000 Steps) (loss=1.98125): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25943 / 30000 Steps) (loss=2.05416): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25967 / 30000 Steps) (loss=1.89355): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (25991 / 30000 Steps) (loss=1.78587): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (26000 / 30000 Steps) (loss=2.70790):  33%|███▎      | 8/24 [00:08<00:14,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (25992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (25992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (25992 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (25992 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.25s/it]\u001b[A\n",
-      "Validate (25992 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.25s/it]\u001b[A\n",
-      "Validate (25992 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (25992 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (25992 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (25992 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (25992 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (25992 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (25992 / 10 Steps): 100%|██████████| 6/6 [00:20<00:00,  3.50s/it]\u001b[A\n",
-      "Training (26000 / 30000 Steps) (loss=2.70790):  38%|███▊      | 9/24 [00:29<01:52,  7.47s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07059632986783981\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (26015 / 30000 Steps) (loss=2.75751): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (26039 / 30000 Steps) (loss=2.22100): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (26042 / 30000 Steps) (loss=2.54282):  12%|█▎        | 3/24 [00:02<00:19,  1.09it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (27311 / 30000 Steps) (loss=2.06320): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27335 / 30000 Steps) (loss=1.96911): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27359 / 30000 Steps) (loss=2.08709): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27383 / 30000 Steps) (loss=2.45813): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27407 / 30000 Steps) (loss=1.77006): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (27431 / 30000 Steps) (loss=1.97804): 100%|██████████| 24/24 [00:21<00:00,  1.09it/s]\n",
-      "Training (27455 / 30000 Steps) (loss=1.70250): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27479 / 30000 Steps) (loss=1.85274): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27500 / 30000 Steps) (loss=2.18979):  83%|████████▎ | 20/24 [00:19<00:03,  1.11it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (27480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (27480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (27480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (27480 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (27480 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (27480 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (27480 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (27480 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (27480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.95s/it]\u001b[A\n",
-      "Validate (27480 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (27480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (27480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (27500 / 30000 Steps) (loss=2.18979):  88%|████████▊ | 21/24 [00:40<00:21,  7.21s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07224365323781967\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (27503 / 30000 Steps) (loss=1.50000): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (27527 / 30000 Steps) (loss=2.27667): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27551 / 30000 Steps) (loss=2.03874): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27575 / 30000 Steps) (loss=2.35553): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27599 / 30000 Steps) (loss=2.26227): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27623 / 30000 Steps) (loss=2.12316): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27647 / 30000 Steps) (loss=1.58768): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (27671 / 30000 Steps) (loss=2.10832): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27695 / 30000 Steps) (loss=1.50807): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27719 / 30000 Steps) (loss=1.97166): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27743 / 30000 Steps) (loss=2.02111): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27767 / 30000 Steps) (loss=2.17290): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27791 / 30000 Steps) (loss=1.62545): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27815 / 30000 Steps) (loss=1.69387): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27839 / 30000 Steps) (loss=2.31924): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27863 / 30000 Steps) (loss=2.24957): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27887 / 30000 Steps) (loss=1.89561): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27911 / 30000 Steps) (loss=2.24539): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27935 / 30000 Steps) (loss=1.56361): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27959 / 30000 Steps) (loss=2.01321): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (27983 / 30000 Steps) (loss=1.92082): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28000 / 30000 Steps) (loss=2.47622):  67%|██████▋   | 16/24 [00:15<00:07,  1.11it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (27984 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (27984 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (27984 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.14s/it]\u001b[A\n",
-      "Validate (27984 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (27984 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:13,  3.26s/it]\u001b[A\n",
-      "Validate (27984 / 10 Steps):  50%|█████     | 3/6 [00:09<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (27984 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.92s/it]\u001b[A\n",
-      "Validate (27984 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.97s/it]\u001b[A\n",
-      "Validate (27984 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.97s/it]\u001b[A\n",
-      "Validate (27984 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (27984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.67s/it]\u001b[A\n",
-      "Validate (27984 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.51s/it]\u001b[A\n",
-      "Training (28000 / 30000 Steps) (loss=2.47622):  71%|███████   | 17/24 [00:36<00:50,  7.25s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.0713159367442131\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (28007 / 30000 Steps) (loss=1.80550): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (28031 / 30000 Steps) (loss=1.86145): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28055 / 30000 Steps) (loss=1.85558): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28079 / 30000 Steps) (loss=1.91898): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28103 / 30000 Steps) (loss=1.92812): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28127 / 30000 Steps) (loss=1.91062): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28151 / 30000 Steps) (loss=2.35851): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28175 / 30000 Steps) (loss=1.96296): 100%|██████████| 24/24 [00:21<00:00,  1.11it/s]\n",
-      "Training (28199 / 30000 Steps) (loss=2.03080): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28223 / 30000 Steps) (loss=1.88645): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28247 / 30000 Steps) (loss=1.96491): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28271 / 30000 Steps) (loss=2.02651): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28295 / 30000 Steps) (loss=1.71030): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28319 / 30000 Steps) (loss=1.89605): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28343 / 30000 Steps) (loss=1.70242): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28367 / 30000 Steps) (loss=1.84909): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28391 / 30000 Steps) (loss=2.02234): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28415 / 30000 Steps) (loss=1.91935): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28439 / 30000 Steps) (loss=1.55340): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28463 / 30000 Steps) (loss=1.59120): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28487 / 30000 Steps) (loss=1.56324): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28500 / 30000 Steps) (loss=1.99294):  50%|█████     | 12/24 [00:11<00:10,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (28488 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (28488 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (28488 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:15,  3.13s/it]\u001b[A\n",
-      "Validate (28488 / 10 Steps):  33%|███▎      | 2/6 [00:06<00:12,  3.25s/it]\u001b[A\n",
-      "Validate (28488 / 10 Steps):  33%|███▎      | 2/6 [00:08<00:12,  3.25s/it]\u001b[A\n",
-      "Validate (28488 / 10 Steps):  50%|█████     | 3/6 [00:08<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (28488 / 10 Steps):  50%|█████     | 3/6 [00:14<00:08,  2.91s/it]\u001b[A\n",
-      "Validate (28488 / 10 Steps):  67%|██████▋   | 4/6 [00:14<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (28488 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:07,  3.96s/it]\u001b[A\n",
-      "Validate (28488 / 10 Steps):  83%|████████▎ | 5/6 [00:17<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (28488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.66s/it]\u001b[A\n",
-      "Validate (28488 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.50s/it]\u001b[A\n",
-      "Training (28500 / 30000 Steps) (loss=1.99294):  54%|█████▍    | 13/24 [00:32<01:19,  7.26s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.09675732254981995 Current Avg. Dice: 0.07071000337600708\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (28511 / 30000 Steps) (loss=1.62969): 100%|██████████| 24/24 [00:42<00:00,  1.78s/it]\n",
-      "Training (28535 / 30000 Steps) (loss=2.18971): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28559 / 30000 Steps) (loss=2.10305): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28583 / 30000 Steps) (loss=1.82222): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28607 / 30000 Steps) (loss=2.32003): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28631 / 30000 Steps) (loss=1.76598): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28655 / 30000 Steps) (loss=1.68438): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (28663 / 30000 Steps) (loss=2.01976):  29%|██▉       | 7/24 [00:07<00:15,  1.10it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (29927 / 30000 Steps) (loss=1.82648): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (29951 / 30000 Steps) (loss=2.04575): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (29975 / 30000 Steps) (loss=1.79833): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]\n",
-      "Training (29999 / 30000 Steps) (loss=1.80398): 100%|██████████| 24/24 [00:21<00:00,  1.10it/s]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Total train time: 28567.65 seconds\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n"
-     ]
-    }
-   ],
-   "source": [
-    "set_determinism(seed=0)\n",
-    "\n",
-    "# torch.cuda.empty_cache()\n",
-    "max_iterations = 30000\n",
-    "eval_num = 500\n",
-    "# max_iterations = 30\n",
-    "# eval_num = 29\n",
-    "global_step = 0\n",
-    "dice_val_best = 0.0\n",
-    "global_step_best = 0\n",
-    "epoch_loss_values = []\n",
-    "metric_values = []\n",
-    "\n",
-    "begin = time.time()\n",
-    "while global_step < max_iterations:\n",
-    "    global_step, dice_val_best, global_step_best = train(\n",
-    "        global_step, train_loader, dice_val_best, global_step_best\n",
-    "    )\n",
-    "print(f\"Total train time: {time.time() - begin:.2f} seconds\")\n",
-    "# model.load_state_dict(torch.load(os.path.join(root_dir, \"best_metric_model.pth\")))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 14,
-   "metadata": {
-    "scrolled": true
-   },
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "train completed, best_metric: 0.0968 at iteration: 500\n"
-     ]
-    }
-   ],
-   "source": [
-    "print(\n",
-    "    f\"train completed, best_metric: {dice_val_best:.4f} \"\n",
-    "    f\"at iteration: {global_step_best}\"\n",
-    ")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 15,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 864x432 with 2 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "plt.figure(\"train\", (12, 6))\n",
-    "plt.subplot(1, 2, 1)\n",
-    "plt.title(\"Iteration Average Loss\")\n",
-    "x = [eval_num * (i + 1) for i in range(len(epoch_loss_values))]\n",
-    "y = epoch_loss_values\n",
-    "plt.xlabel(\"Iteration\")\n",
-    "plt.plot(x, y)\n",
-    "plt.subplot(1, 2, 2)\n",
-    "plt.title(\"Val Mean Dice\")\n",
-    "x = [eval_num * (i + 1) for i in range(len(metric_values))]\n",
-    "y = metric_values\n",
-    "plt.xlabel(\"Iteration\")\n",
-    "plt.plot(x, y)\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Plot the loss and metric"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 12,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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",
-      "text/plain": [
-       "<Figure size 864x432 with 2 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "plt.figure(\"train\", (12, 6))\n",
-    "plt.subplot(1, 2, 1)\n",
-    "plt.title(\"Iteration Average Loss\")\n",
-    "x = [eval_num * (i + 1) for i in range(len(epoch_loss_values))]\n",
-    "y = epoch_loss_values\n",
-    "plt.xlabel(\"Iteration\")\n",
-    "plt.plot(x, y)\n",
-    "plt.subplot(1, 2, 2)\n",
-    "plt.title(\"Val Mean Dice\")\n",
-    "x = [eval_num * (i + 1) for i in range(len(metric_values))]\n",
-    "y = metric_values\n",
-    "plt.xlabel(\"Iteration\")\n",
-    "plt.plot(x, y)\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Check best model output with the input image and label"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 13,
-   "metadata": {},
-   "outputs": [
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 864x432 with 2 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plt.figure(\"train\", (12, 6))\n",
+    "plt.subplot(1, 2, 1)\n",
+    "plt.title(\"Iteration Average Loss\")\n",
+    "x = [eval_num * (i + 1) for i in range(len(epoch_loss_values))]\n",
+    "y = epoch_loss_values\n",
+    "plt.xlabel(\"Iteration\")\n",
+    "plt.plot(x, y)\n",
+    "plt.subplot(1, 2, 2)\n",
+    "plt.title(\"Val Mean Dice\")\n",
+    "x = [eval_num * (i + 1) for i in range(len(metric_values))]\n",
+    "y = metric_values\n",
+    "plt.xlabel(\"Iteration\")\n",
+    "plt.plot(x, y)\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Check best model output with the input image and label"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [
     {
      "data": {
       "image/png": "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",
@@ -2697,11 +663,11 @@
     "model.load_state_dict(torch.load(os.path.join(root_dir, \"best_metric_model.pth\")))\n",
     "model.eval()\n",
     "with torch.no_grad():\n",
-    "    img_name = os.path.split(val_ds[case_num][\"image_meta_dict\"][\"filename_or_obj\"])[1]\n",
+    "    img_name = os.path.split(val_ds[case_num]['image'].meta[\"filename_or_obj\"])[1]\n",
     "    img = val_ds[case_num][\"image\"]\n",
     "    label = val_ds[case_num][\"label\"]\n",
-    "    val_inputs = torch.unsqueeze(img, 1).cuda(device=device)\n",
-    "    val_labels = torch.unsqueeze(label, 1).cuda(device=device)\n",
+    "    val_inputs = torch.unsqueeze(img, 1).cuda()\n",
+    "    val_labels = torch.unsqueeze(label, 1).cuda()\n",
     "    val_outputs = sliding_window_inference(\n",
     "        val_inputs, (96, 96, 96), 4, model, overlap=0.8\n",
     "    )\n",
@@ -2731,7 +697,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [

From 0d417c56627521b60afe999fbe71e9b5985b4e8e Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Sun, 24 Jul 2022 08:01:54 -0700
Subject: [PATCH 07/17] applied autofixes

---
 3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb | 1 -
 1 file changed, 1 deletion(-)

diff --git a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
index 7f91d3c1a0..ba27d7b50a 100644
--- a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
+++ b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
@@ -120,7 +120,6 @@
     "import tempfile\n",
     "\n",
     "import matplotlib.pyplot as plt\n",
-    "import numpy as np\n",
     "from tqdm import tqdm\n",
     "\n",
     "from monai.losses import DiceCELoss\n",

From d581ba4a591c8248a00b4676d30b05e2d6f6b42d Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Sun, 24 Jul 2022 08:53:36 -0700
Subject: [PATCH 08/17] done

---
 3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
index ba27d7b50a..1c2a77f398 100644
--- a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
+++ b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
@@ -451,7 +451,8 @@
    "source": [
     "torch.backends.cudnn.benchmark = True\n",
     "loss_function = DiceCELoss(to_onehot_y=True, softmax=True)\n",
-    "optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)"
+    "optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)\n",
+    "scaler = torch.cuda.amp.GradScaler()"
    ]
   },
   {

From c06c46c68b2568a2d3e8867baaf81df7454caf9a Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Mon, 25 Jul 2022 12:55:20 -0700
Subject: [PATCH 09/17] removed unnecessary files

---
 3d_segmentation/swin_unetr_orig-091.ipynb | 2308 ---------------------
 3d_segmentation/swin_unetr_profiling.py   |  333 ---
 2 files changed, 2641 deletions(-)
 delete mode 100644 3d_segmentation/swin_unetr_orig-091.ipynb
 delete mode 100644 3d_segmentation/swin_unetr_profiling.py

diff --git a/3d_segmentation/swin_unetr_orig-091.ipynb b/3d_segmentation/swin_unetr_orig-091.ipynb
deleted file mode 100644
index 7e2e0e1928..0000000000
--- a/3d_segmentation/swin_unetr_orig-091.ipynb
+++ /dev/null
@@ -1,2308 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "# 3D Multi-organ Segmentation with Swin UNETR  (BTCV Challenge)\n",
-    "\n",
-    "\n",
-    "This tutorial uses a Swin UNETR [1] model for the task of multi-organ segmentation task using the BTCV challenge dataset. The architecture of Swin UNETR is demonstrated as below\n",
-    "![image](https://lh3.googleusercontent.com/pw/AM-JKLVx_J2DKYA7DCo3F_gGbK2e1sI_yzjYwQt-EWCirNGKsUv1hi7qLMofkY0r5xVXJNzhr8qenBkUJJYXtj49xsWJgOgbkBpcN7rz9axkeN3tgJbWldtZhYcBgYOlklzUS34eMCL-gRkxyFydJQ_Y1HAx=w1322-h518-no?authuser=2)\n",
-    "\n",
-    "The following features are included in this tutorial:\n",
-    "1. Transforms for dictionary format data.\n",
-    "1. Define a new transform according to MONAI transform API.\n",
-    "1. Load Nifti image with metadata, load a list of images and stack them.\n",
-    "1. Randomly adjust intensity for data augmentation.\n",
-    "1. Cache IO and transforms to accelerate training and validation.\n",
-    "1. Swin UNETR model, DiceCE loss function, Mean Dice metric for multi-organ segmentation task.\n",
-    "\n",
-    "For this tutorial, the dataset needs to be downloaded from: https://www.synapse.org/#!Synapse:syn3193805/wiki/217752. \n",
-    "\n",
-    "In addition, the json file for data splits needs to be downloaded from this [link](https://drive.google.com/file/d/1t4fIQQkONv7ArTSZe4Nucwkk1KfdUDvW/view?usp=sharing). Once downloaded, place the json file in the same folder as the dataset. \n",
-    "\n",
-    "For BTCV dataset, under Institutional Review Board (IRB) supervision, 50 abdomen CT scans of were randomly selected from a combination of an ongoing colorectal cancer chemotherapy trial, and a retrospective ventral hernia study. The 50 scans were captured during portal venous contrast phase with variable volume sizes (512 x 512 x 85 - 512 x 512 x 198) and field of views (approx. 280 x 280 x 280 mm3 - 500 x 500 x 650 mm3). The in-plane resolution varies from 0.54 x 0.54 mm2 to 0.98 x 0.98 mm2, while the slice thickness ranges from 2.5 mm to 5.0 mm. \n",
-    "\n",
-    "Target: 13 abdominal organs including 1. Spleen 2. Right Kidney 3. Left Kideny 4.Gallbladder 5.Esophagus 6. Liver 7. Stomach 8.Aorta 9. IVC 10. Portal and Splenic Veins 11. Pancreas 12 Right adrenal gland 13 Left adrenal gland.\n",
-    "\n",
-    "Modality: CT\n",
-    "Size: 30 3D volumes (24 Training + 6 Testing)  \n",
-    "Challenge: BTCV MICCAI Challenge\n",
-    "\n",
-    "The following figure shows image patches with the organ sub-regions that are annotated in the CT (top left) and the final labels for the whole dataset (right).\n",
-    "\n",
-    "Data, figures and resources are taken from: \n",
-    "\n",
-    "\n",
-    "1. [Self-Supervised Pre-Training of Swin Transformers\n",
-    "for 3D Medical Image Analysis](https://arxiv.org/abs/2111.14791)\n",
-    "\n",
-    "2. [Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images](https://arxiv.org/abs/2201.01266)\n",
-    "\n",
-    "3. [High-resolution 3D abdominal segmentation with random patch network fusion (MIA)](https://www.sciencedirect.com/science/article/abs/pii/S1361841520302589)\n",
-    "\n",
-    "4. [Efficient multi-atlas abdominal segmentation on clinically acquired CT with SIMPLE context learning (MIA)](https://www.sciencedirect.com/science/article/abs/pii/S1361841515000766?via%3Dihub)\n",
-    "\n",
-    "\n",
-    "![image](https://lh3.googleusercontent.com/pw/AM-JKLX0svvlMdcrchGAgiWWNkg40lgXYjSHsAAuRc5Frakmz2pWzSzf87JQCRgYpqFR0qAjJWPzMQLc_mmvzNjfF9QWl_1OHZ8j4c9qrbR6zQaDJWaCLArRFh0uPvk97qAa11HtYbD6HpJ-wwTCUsaPcYvM=w1724-h522-no?authuser=0)\n",
-    "\n",
-    "\n",
-    "\n",
-    "The image patches show anatomies of a subject, including: \n",
-    "1. large organs: spleen, liver, stomach. \n",
-    "2. Smaller organs: gallbladder, esophagus, kidneys, pancreas. \n",
-    "3. Vascular tissues: aorta, IVC, P&S Veins. \n",
-    "4. Glands: left and right adrenal gland\n",
-    "\n",
-    "If you find this tutorial helpful, please consider citing [1] and [2]:\n",
-    "\n",
-    "[1]: Tang, Y., Yang, D., Li, W., Roth, H.R., Landman, B., Xu, D., Nath, V. and Hatamizadeh, A., 2022. Self-supervised pre-training of swin transformers for 3d medical image analysis. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 20730-20740).\n",
-    "\n",
-    "[2]: Hatamizadeh, A., Nath, V., Tang, Y., Yang, D., Roth, H. and Xu, D., 2022. Swin UNETR: Swin Transformers for Semantic Segmentation of Brain Tumors in MRI Images. arXiv preprint arXiv:2201.01266.\n",
-    "\n",
-    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/Project-MONAI/tutorials/blob/main/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb)\n"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "# Pre-trained Swin UNETR Encoder\n",
-    "\n",
-    "We use weights from self-supervised pre-training of Swin UNETR encoder (3D Swin Tranformer) on a cohort of 5050 CT scans from publicly available datasets. The encoder is pre-trained using reconstructin, rotation prediction and contrastive learning pre-text tasks as shown below. For more details, please refer to [1] (CVPR paper) and see this [repository](https://github.com/Project-MONAI/research-contributions/tree/main/SwinUNETR/Pretrain). \n",
-    "\n",
-    "![image](https://lh3.googleusercontent.com/pw/AM-JKLVLgduGZ9naCSasWg09U665NBdd3UD4eLTy15wJiwbmKLS_p5WSZ2MBcRePEJO2tv9X3TkC52MsbnomuPy5JT3vSVeCji1MOEuAzcsxily88TdbHuAt6PzccefwKupbXyOCumK5hzz5Ul38kZnlEQ84=w397-h410-no?authuser=2)\n",
-    "\n",
-    "Please download the pre-trained weights from this [link](https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/model_swinvit.pt) and place it in the root directory of this tutorial. \n",
-    "\n",
-    "If training from scratch is desired, please skip the step for initializing from pre-trained weights. "
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Setup environment"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "/opt/conda/lib/python3.8/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"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Fri Jul 15 20:35:49 2022       \n",
-      "+-----------------------------------------------------------------------------+\n",
-      "| NVIDIA-SMI 450.119.04   Driver Version: 450.119.04   CUDA Version: 11.6     |\n",
-      "|-------------------------------+----------------------+----------------------+\n",
-      "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
-      "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
-      "|                               |                      |               MIG M. |\n",
-      "|===============================+======================+======================|\n",
-      "|   0  Tesla V100-SXM2...  On   | 00000000:89:00.0 Off |                    0 |\n",
-      "| N/A   36C    P0    43W / 163W |      0MiB / 32510MiB |      0%      Default |\n",
-      "|                               |                      |                  N/A |\n",
-      "+-------------------------------+----------------------+----------------------+\n",
-      "                                                                               \n",
-      "+-----------------------------------------------------------------------------+\n",
-      "| Processes:                                                                  |\n",
-      "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
-      "|        ID   ID                                                   Usage      |\n",
-      "|=============================================================================|\n",
-      "|  No running processes found                                                 |\n",
-      "+-----------------------------------------------------------------------------+\n"
-     ]
-    }
-   ],
-   "source": [
-    "import torch\n",
-    "torch.cuda.empty_cache()\n",
-    "\n",
-    "!nvidia-smi"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 2,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
-      "Requirement already satisfied: monai==0.9.1rc3 in /opt/conda/lib/python3.8/site-packages (0.9.1rc3)\n",
-      "Requirement already satisfied: numpy>=1.17 in /opt/conda/lib/python3.8/site-packages (from monai==0.9.1rc3) (1.22.3)\n",
-      "Requirement already satisfied: torch>=1.7 in /opt/conda/lib/python3.8/site-packages (from monai==0.9.1rc3) (1.12.0a0+bd13bc6)\n",
-      "Requirement already satisfied: typing_extensions in /opt/conda/lib/python3.8/site-packages (from torch>=1.7->monai==0.9.1rc3) (4.1.1)\n",
-      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
-      "\u001b[0mLooking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
-      "Requirement already satisfied: nibabel==3.1.1 in /opt/conda/lib/python3.8/site-packages (3.1.1)\n",
-      "Requirement already satisfied: packaging>=14.3 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (21.3)\n",
-      "Requirement already satisfied: numpy>=1.13 in /opt/conda/lib/python3.8/site-packages (from nibabel==3.1.1) (1.22.3)\n",
-      "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.8/site-packages (from packaging>=14.3->nibabel==3.1.1) (3.0.8)\n",
-      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
-      "\u001b[0mLooking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\n",
-      "Requirement already satisfied: tqdm==4.63.0 in /opt/conda/lib/python3.8/site-packages (4.63.0)\n",
-      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\u001b[33m\n",
-      "\u001b[0m"
-     ]
-    }
-   ],
-   "source": [
-    "# !pip install 'git+https://github.com/Project-MONAI/MONAI#egg.gitmonai@0.8.1+271.g07de215c'\n",
-    "!pip install monai==0.9.1rc3\n",
-    "!pip install nibabel==3.1.1\n",
-    "!pip install tqdm==4.63.0\n",
-    "!python -c \"import matplotlib\" || pip install -q matplotlib\n",
-    "%matplotlib inline"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "/opt/conda/lib/python3.8/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"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "MONAI version: 0.9.1rc3\n",
-      "Numpy version: 1.22.3\n",
-      "Pytorch version: 1.12.0a0+bd13bc6\n",
-      "MONAI flags: HAS_EXT = False, USE_COMPILED = False, USE_META_DICT = False\n",
-      "MONAI rev id: 7a5de8b7b9db101a431e70ae2aa8ea7ebb8dfffe\n",
-      "MONAI __file__: /opt/conda/lib/python3.8/site-packages/monai/__init__.py\n",
-      "\n",
-      "Optional dependencies:\n",
-      "Pytorch Ignite version: 0.4.8\n",
-      "Nibabel version: 3.1.1\n",
-      "scikit-image version: 0.19.3\n",
-      "Pillow version: 9.0.1\n",
-      "Tensorboard version: 2.8.0\n",
-      "gdown version: 4.4.0\n",
-      "TorchVision version: 0.13.0a0\n",
-      "tqdm version: 4.63.0\n",
-      "lmdb version: 1.3.0\n",
-      "psutil version: 5.9.0\n",
-      "pandas version: 1.3.5\n",
-      "einops version: 0.4.1\n",
-      "transformers version: 4.19.4\n",
-      "mlflow version: 1.26.1\n",
-      "pynrrd version: 0.4.3\n",
-      "\n",
-      "For details about installing the optional dependencies, please visit:\n",
-      "    https://docs.monai.io/en/latest/installation.html#installing-the-recommended-dependencies\n",
-      "\n"
-     ]
-    }
-   ],
-   "source": [
-    "import os\n",
-    "import shutil\n",
-    "import tempfile\n",
-    "\n",
-    "import matplotlib.pyplot as plt\n",
-    "import numpy as np\n",
-    "from tqdm import tqdm\n",
-    "import time\n",
-    "\n",
-    "from monai.losses import DiceCELoss\n",
-    "from monai.inferers import sliding_window_inference\n",
-    "from monai.transforms import (\n",
-    "    AsDiscrete,\n",
-    "    AddChanneld,\n",
-    "    Compose,\n",
-    "    CropForegroundd,\n",
-    "    LoadImaged,\n",
-    "    Orientationd,\n",
-    "    RandFlipd,\n",
-    "    RandCropByPosNegLabeld,\n",
-    "    RandShiftIntensityd,\n",
-    "    ScaleIntensityRanged,\n",
-    "    Spacingd,\n",
-    "    RandRotate90d,\n",
-    "    ToTensord,\n",
-    "    ToDeviced,\n",
-    ")\n",
-    "\n",
-    "from monai.config import print_config\n",
-    "from monai.metrics import DiceMetric\n",
-    "from monai.networks.nets import SwinUNETR\n",
-    "\n",
-    "from monai.data import (\n",
-    "    DataLoader,\n",
-    "    ThreadDataLoader,\n",
-    "    CacheDataset,\n",
-    "    load_decathlon_datalist,\n",
-    "    decollate_batch,\n",
-    ")\n",
-    "\n",
-    "\n",
-    "import torch\n",
-    "\n",
-    "print_config()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Setup data directory\n",
-    "\n",
-    "You can specify a directory with the `MONAI_DATA_DIRECTORY` environment variable.  \n",
-    "This allows you to save results and reuse downloads.  \n",
-    "If not specified a temporary directory will be used."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 2,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "/tmp/tmpt26y6vq7\n"
-     ]
-    }
-   ],
-   "source": [
-    "directory = os.environ.get(\"MONAI_DATA_DIRECTORY\")\n",
-    "root_dir = tempfile.mkdtemp() if directory is None else directory\n",
-    "print(root_dir)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Setup transforms for training and validation\n",
-    "To save on GPU memory utilization, the num_samples can be reduced to 2. "
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 3,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "num_samples = 4\n",
-    "\n",
-    "os.environ[\"CUDA_DEVICE_ORDER\"] = \"PCI_BUS_ID\"\n",
-    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
-    "\n",
-    "train_transforms = Compose(\n",
-    "    [\n",
-    "        LoadImaged(keys=[\"image\", \"label\"]),\n",
-    "        AddChanneld(keys=[\"image\", \"label\"]),\n",
-    "        Orientationd(keys=[\"image\", \"label\"], axcodes=\"RAS\"),\n",
-    "        Spacingd(\n",
-    "            keys=[\"image\", \"label\"],\n",
-    "            pixdim=(1.5, 1.5, 2.0),\n",
-    "            mode=(\"bilinear\", \"nearest\"),\n",
-    "        ),\n",
-    "        ScaleIntensityRanged(\n",
-    "            keys=[\"image\"],\n",
-    "            a_min=-175,\n",
-    "            a_max=250,\n",
-    "            b_min=0.0,\n",
-    "            b_max=1.0,\n",
-    "            clip=True,\n",
-    "        ),\n",
-    "        CropForegroundd(keys=[\"image\", \"label\"], source_key=\"image\"),\n",
-    "        RandCropByPosNegLabeld(\n",
-    "            keys=[\"image\", \"label\"],\n",
-    "            label_key=\"label\",\n",
-    "            spatial_size=(96, 96, 96),\n",
-    "            pos=1,\n",
-    "            neg=1,\n",
-    "            num_samples=num_samples,\n",
-    "            image_key=\"image\",\n",
-    "            image_threshold=0,\n",
-    "        ),\n",
-    "        RandFlipd(\n",
-    "            keys=[\"image\", \"label\"],\n",
-    "            spatial_axis=[0],\n",
-    "            prob=0.10,\n",
-    "        ),\n",
-    "        RandFlipd(\n",
-    "            keys=[\"image\", \"label\"],\n",
-    "            spatial_axis=[1],\n",
-    "            prob=0.10,\n",
-    "        ),\n",
-    "        RandFlipd(\n",
-    "            keys=[\"image\", \"label\"],\n",
-    "            spatial_axis=[2],\n",
-    "            prob=0.10,\n",
-    "        ),\n",
-    "        RandRotate90d(\n",
-    "            keys=[\"image\", \"label\"],\n",
-    "            prob=0.10,\n",
-    "            max_k=3,\n",
-    "        ),\n",
-    "        RandShiftIntensityd(\n",
-    "            keys=[\"image\"],\n",
-    "            offsets=0.10,\n",
-    "            prob=0.50,\n",
-    "        ),\n",
-    "        ToTensord(keys=[\"image\", \"label\"]),\n",
-    "#        ToDeviced(keys=[\"image\", \"label\"], device=device),\n",
-    "    ]\n",
-    ")\n",
-    "val_transforms = Compose(\n",
-    "    [\n",
-    "        LoadImaged(keys=[\"image\", \"label\"]),\n",
-    "        AddChanneld(keys=[\"image\", \"label\"]),\n",
-    "        Orientationd(keys=[\"image\", \"label\"], axcodes=\"RAS\"),\n",
-    "        Spacingd(\n",
-    "            keys=[\"image\", \"label\"],\n",
-    "            pixdim=(1.5, 1.5, 2.0),\n",
-    "            mode=(\"bilinear\", \"nearest\"),\n",
-    "        ),\n",
-    "        ScaleIntensityRanged(\n",
-    "            keys=[\"image\"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True\n",
-    "        ),\n",
-    "        CropForegroundd(keys=[\"image\", \"label\"], source_key=\"image\"),\n",
-    "       # ToDeviced(keys=[\"image\", \"label\"], device=device),\n",
-    "        ToTensord(keys=[\"image\", \"label\"]),\n",
-    "    ]\n",
-    ")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 4,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Loading dataset: 100%|██████████| 24/24 [01:33<00:00,  3.91s/it]\n",
-      "Loading dataset: 100%|██████████| 6/6 [00:28<00:00,  4.78s/it]\n"
-     ]
-    }
-   ],
-   "source": [
-    "data_dir = \"data/\"\n",
-    "split_JSON = \"dataset_0.json\"\n",
-    "\n",
-    "datasets = data_dir + split_JSON\n",
-    "datalist = load_decathlon_datalist(datasets, True, \"training\")\n",
-    "val_files = load_decathlon_datalist(datasets, True, \"validation\")\n",
-    "train_ds = CacheDataset(\n",
-    "    data=datalist,\n",
-    "    transform=train_transforms,\n",
-    "    cache_num=24,\n",
-    "    cache_rate=1.0,\n",
-    "    num_workers=8,\n",
-    ")\n",
-    "train_loader = DataLoader(train_ds, batch_size=1, shuffle=True, num_workers=8, pin_memory=True)\n",
-    "# train_loader = ThreadDataLoader(train_ds, num_workers=0, batch_size=1, shuffle=True)\n",
-    "val_ds = CacheDataset(\n",
-    "    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4\n",
-    ")\n",
-    "val_loader = DataLoader(val_ds, batch_size=1, shuffle=False, num_workers=4, pin_memory=True)\n",
-    "# val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Check data shape and visualize"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 5,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "image shape: torch.Size([1, 255, 223, 276]), label shape: torch.Size([1, 255, 223, 276])\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 1296x432 with 2 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "slice_map = {\n",
-    "    \"img0035.nii.gz\": 170,\n",
-    "    \"img0036.nii.gz\": 230,\n",
-    "    \"img0037.nii.gz\": 204,\n",
-    "    \"img0038.nii.gz\": 204,\n",
-    "    \"img0039.nii.gz\": 204,\n",
-    "    \"img0040.nii.gz\": 180,\n",
-    "}\n",
-    "case_num = 1\n",
-    "img_name = os.path.split(val_ds[case_num][\"image_meta_dict\"][\"filename_or_obj\"])[1]\n",
-    "img = val_ds[case_num][\"image\"]\n",
-    "label = val_ds[case_num][\"label\"]\n",
-    "img_shape = img.shape\n",
-    "label_shape = label.shape\n",
-    "print(f\"image shape: {img_shape}, label shape: {label_shape}\")\n",
-    "plt.figure(\"image\", (18, 6))\n",
-    "plt.subplot(1, 2, 1)\n",
-    "plt.title(\"image\")\n",
-    "plt.imshow(img[0, :, :, slice_map[img_name]].detach().cpu(), cmap=\"gray\")\n",
-    "plt.subplot(1, 2, 2)\n",
-    "plt.title(\"label\")\n",
-    "plt.imshow(label[0, :, :, slice_map[img_name]].detach().cpu())\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Create Swin UNETR model\n",
-    "\n",
-    "In this scetion, we create Swin UNETR model for the 14-class multi-organ segmentation. We use a feature size of 48 which is compatible with self-supervised pre-trained weights. We also use gradient checkpointing (use_checkpoint) for more memory-efficient training. "
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 6,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "model = SwinUNETR(\n",
-    "    img_size=(96, 96, 96),\n",
-    "    in_channels=1,\n",
-    "    out_channels=14,\n",
-    "    feature_size=48,\n",
-    "    use_checkpoint=True,\n",
-    ").to(device)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Initialize Swin UNETR encoder from self-supervised pre-trained weights\n",
-    "\n",
-    "In this section, we intialize the Swin UNETR encoder from weights downloaded from this [link](https://github.com/Project-MONAI/MONAI-extra-test-data/releases/download/0.8.1/model_swinvit.pt). If training from scratch is desired, please skip this section."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 7,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Using pretrained self-supervied Swin UNETR backbone weights !\n"
-     ]
-    }
-   ],
-   "source": [
-    "weight = torch.load(\"./model_swinvit.pt\")\n",
-    "model.load_from(weights=weight)\n",
-    "print(\"Using pretrained self-supervied Swin UNETR backbone weights !\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Optimizer and loss function"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 8,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "torch.backends.cudnn.benchmark = True\n",
-    "loss_function = DiceCELoss(to_onehot_y=True, softmax=True)\n",
-    "optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Execute a typical PyTorch training process"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 9,
-   "metadata": {
-    "scrolled": true
-   },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (23 / 30000 Steps) (loss=3.18401): 100%|██████████| 24/24 [00:38<00:00,  1.62s/it]\n",
-      "Training (47 / 30000 Steps) (loss=3.07850): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (71 / 30000 Steps) (loss=3.11205): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (95 / 30000 Steps) (loss=3.10495): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (119 / 30000 Steps) (loss=3.04469): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (143 / 30000 Steps) (loss=2.98636): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (167 / 30000 Steps) (loss=2.85411): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (191 / 30000 Steps) (loss=2.71353): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (215 / 30000 Steps) (loss=2.65432): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (239 / 30000 Steps) (loss=2.71864): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (263 / 30000 Steps) (loss=2.65412): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (287 / 30000 Steps) (loss=2.70267): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (311 / 30000 Steps) (loss=2.63068): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (335 / 30000 Steps) (loss=2.54479): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (359 / 30000 Steps) (loss=2.56929): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (383 / 30000 Steps) (loss=2.54598): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (407 / 30000 Steps) (loss=2.54486): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (431 / 30000 Steps) (loss=2.67341): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (455 / 30000 Steps) (loss=2.57942): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (479 / 30000 Steps) (loss=2.47799): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (500 / 30000 Steps) (loss=2.68251):  83%|████████▎ | 20/24 [00:20<00:03,  1.12it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[ANone of the inputs have requires_grad=True. Gradients will be None\n",
-      "\n",
-      "Validate (480 / 10 Steps):   0%|          | 0/6 [00:06<?, ?it/s]        \u001b[A\n",
-      "Validate (480 / 10 Steps):  17%|█▋        | 1/6 [00:06<00:31,  6.29s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  17%|█▋        | 1/6 [00:09<00:31,  6.29s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:18,  4.56s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  33%|███▎      | 2/6 [00:12<00:18,  4.56s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  50%|█████     | 3/6 [00:12<00:10,  3.63s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  50%|█████     | 3/6 [00:17<00:10,  3.63s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  67%|██████▋   | 4/6 [00:17<00:08,  4.40s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  67%|██████▋   | 4/6 [00:20<00:08,  4.40s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  83%|████████▎ | 5/6 [00:20<00:03,  3.96s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps):  83%|████████▎ | 5/6 [00:24<00:03,  3.96s/it]\u001b[A\n",
-      "Validate (480 / 10 Steps): 100%|██████████| 6/6 [00:24<00:00,  4.07s/it]\u001b[A\n",
-      "positional arguments and argument \"destination\" are deprecated. nn.Module.state_dict will not accept them in the future. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.\n",
-      "Training (500 / 30000 Steps) (loss=2.68251):  88%|████████▊ | 21/24 [00:45<00:25,  8.48s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.10523679107427597\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (503 / 30000 Steps) (loss=2.53570): 100%|██████████| 24/24 [00:48<00:00,  2.01s/it]\n",
-      "Training (527 / 30000 Steps) (loss=2.46318): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (551 / 30000 Steps) (loss=2.37891): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (575 / 30000 Steps) (loss=2.55742): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (599 / 30000 Steps) (loss=2.47565): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (623 / 30000 Steps) (loss=2.52184): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (647 / 30000 Steps) (loss=2.34970): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (671 / 30000 Steps) (loss=2.38541): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (695 / 30000 Steps) (loss=2.32277): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (719 / 30000 Steps) (loss=2.29923): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (743 / 30000 Steps) (loss=2.22664): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (767 / 30000 Steps) (loss=2.17650): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (791 / 30000 Steps) (loss=2.33309): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (815 / 30000 Steps) (loss=2.35450): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (839 / 30000 Steps) (loss=2.33414): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (863 / 30000 Steps) (loss=2.23909): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (887 / 30000 Steps) (loss=2.25688): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (911 / 30000 Steps) (loss=2.09836): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (935 / 30000 Steps) (loss=2.14540): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (959 / 30000 Steps) (loss=2.18719): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (983 / 30000 Steps) (loss=2.17081): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (1000 / 30000 Steps) (loss=2.13638):  67%|██████▋   | 16/24 [00:16<00:07,  1.12it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (984 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]        \u001b[A\n",
-      "Validate (984 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.02s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.02s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (984 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (1000 / 30000 Steps) (loss=2.13638):  71%|███████   | 17/24 [00:38<00:52,  7.54s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.07569018751382828\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (1007 / 30000 Steps) (loss=2.15620): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
-      "Training (1031 / 30000 Steps) (loss=2.23309): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1055 / 30000 Steps) (loss=2.18655): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (1079 / 30000 Steps) (loss=2.23391): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1103 / 30000 Steps) (loss=2.04205): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1127 / 30000 Steps) (loss=2.06677): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1151 / 30000 Steps) (loss=2.01106): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (1175 / 30000 Steps) (loss=2.01657): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1199 / 30000 Steps) (loss=2.04089): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1223 / 30000 Steps) (loss=1.95371): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (1247 / 30000 Steps) (loss=1.93151): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (1271 / 30000 Steps) (loss=2.18730): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1295 / 30000 Steps) (loss=1.89861): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1319 / 30000 Steps) (loss=1.89109): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1343 / 30000 Steps) (loss=2.08866): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1367 / 30000 Steps) (loss=1.99767): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1391 / 30000 Steps) (loss=1.92629): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1415 / 30000 Steps) (loss=1.89440): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1439 / 30000 Steps) (loss=1.99090): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (1463 / 30000 Steps) (loss=2.22657): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1487 / 30000 Steps) (loss=1.92651): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1500 / 30000 Steps) (loss=1.90171):  50%|█████     | 12/24 [00:13<00:10,  1.11it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (1488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
-      "Validate (1488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.05s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.05s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.64s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.64s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.13s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.13s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (1488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.69s/it]\u001b[A\n",
-      "Training (1500 / 30000 Steps) (loss=1.90171):  54%|█████▍    | 13/24 [00:35<01:23,  7.60s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.08025763183832169\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (1511 / 30000 Steps) (loss=2.25799): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
-      "Training (1535 / 30000 Steps) (loss=2.16516): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (1559 / 30000 Steps) (loss=1.88683): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1583 / 30000 Steps) (loss=1.93661): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (1607 / 30000 Steps) (loss=2.15409): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1631 / 30000 Steps) (loss=2.03945): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1655 / 30000 Steps) (loss=2.16960): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (1679 / 30000 Steps) (loss=1.93416): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1703 / 30000 Steps) (loss=1.89512): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1727 / 30000 Steps) (loss=1.89541): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1751 / 30000 Steps) (loss=2.10734): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (1775 / 30000 Steps) (loss=1.77242): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1799 / 30000 Steps) (loss=2.02548): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1823 / 30000 Steps) (loss=1.97165): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (1847 / 30000 Steps) (loss=1.98719): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (1871 / 30000 Steps) (loss=1.89960): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (1895 / 30000 Steps) (loss=1.82791): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1919 / 30000 Steps) (loss=2.12694): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (1943 / 30000 Steps) (loss=1.93836): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (1967 / 30000 Steps) (loss=1.99752): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (1991 / 30000 Steps) (loss=2.00321): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (2111 / 30000 Steps) (loss=1.87203): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2135 / 30000 Steps) (loss=1.83229): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (2159 / 30000 Steps) (loss=1.97508): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2183 / 30000 Steps) (loss=1.93266): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2207 / 30000 Steps) (loss=1.96755): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2231 / 30000 Steps) (loss=2.01195): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (2255 / 30000 Steps) (loss=1.87510): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2279 / 30000 Steps) (loss=1.74624): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (2303 / 30000 Steps) (loss=1.99921): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2327 / 30000 Steps) (loss=1.96728): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (2351 / 30000 Steps) (loss=1.77553): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2375 / 30000 Steps) (loss=1.96980): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (2399 / 30000 Steps) (loss=1.96445): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (2423 / 30000 Steps) (loss=1.95579): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2447 / 30000 Steps) (loss=1.82860): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (2471 / 30000 Steps) (loss=1.88289): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (2495 / 30000 Steps) (loss=1.86248): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2500 / 30000 Steps) (loss=1.96714):  17%|█▋        | 4/24 [00:05<00:21,  1.08s/it]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (2496 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
-      "Validate (2496 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.00s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.00s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.62s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.62s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.11s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.11s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (2496 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (2500 / 30000 Steps) (loss=1.96714):  21%|██        | 5/24 [00:27<02:50,  8.97s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.07246538996696472\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (2519 / 30000 Steps) (loss=1.76536): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
-      "Training (2543 / 30000 Steps) (loss=1.87272): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (2567 / 30000 Steps) (loss=2.24276): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (2591 / 30000 Steps) (loss=1.79909): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (2615 / 30000 Steps) (loss=1.85001): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (2639 / 30000 Steps) (loss=1.79896): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (2663 / 30000 Steps) (loss=1.71249): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2687 / 30000 Steps) (loss=1.89300): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (2711 / 30000 Steps) (loss=1.88981): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (2735 / 30000 Steps) (loss=1.81299): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (2759 / 30000 Steps) (loss=1.86044): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2783 / 30000 Steps) (loss=1.80493): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (2807 / 30000 Steps) (loss=1.88155): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (2831 / 30000 Steps) (loss=2.06799): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (2855 / 30000 Steps) (loss=1.82452): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (2879 / 30000 Steps) (loss=1.90656): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2903 / 30000 Steps) (loss=1.62591): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2927 / 30000 Steps) (loss=1.92314): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (2951 / 30000 Steps) (loss=1.80026): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (2975 / 30000 Steps) (loss=1.80520): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (2999 / 30000 Steps) (loss=1.79177): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (3000 / 30000 Steps) (loss=1.91799):   0%|          | 0/24 [00:02<?, ?it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (3000 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
-      "Validate (3000 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.02s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.02s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.62s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.62s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (3000 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (3000 / 30000 Steps) (loss=1.91799):   4%|▍         | 1/24 [00:24<09:30, 24.79s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.09305248409509659\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (3023 / 30000 Steps) (loss=1.88577): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
-      "Training (3047 / 30000 Steps) (loss=1.74826): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (3071 / 30000 Steps) (loss=1.76287): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (3095 / 30000 Steps) (loss=1.66413): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (3119 / 30000 Steps) (loss=1.70303): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (3143 / 30000 Steps) (loss=1.70093): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (3167 / 30000 Steps) (loss=1.65261): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (3191 / 30000 Steps) (loss=1.73050): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (3215 / 30000 Steps) (loss=2.02838): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (3239 / 30000 Steps) (loss=1.80735): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (3263 / 30000 Steps) (loss=1.60768): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (3287 / 30000 Steps) (loss=1.59386): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (3311 / 30000 Steps) (loss=1.88477): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (3335 / 30000 Steps) (loss=1.74821): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (3359 / 30000 Steps) (loss=1.58187): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (3383 / 30000 Steps) (loss=1.72284): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (3407 / 30000 Steps) (loss=1.76967): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (3431 / 30000 Steps) (loss=1.58086): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (3455 / 30000 Steps) (loss=1.64616): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (3464 / 30000 Steps) (loss=1.96805):  33%|███▎      | 8/24 [00:09<00:15,  1.06it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (4511 / 30000 Steps) (loss=1.71779): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
-      "Training (4535 / 30000 Steps) (loss=1.59975): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (4559 / 30000 Steps) (loss=1.95123): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (4583 / 30000 Steps) (loss=1.68883): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (4607 / 30000 Steps) (loss=2.08812): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (4631 / 30000 Steps) (loss=1.75501): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (4655 / 30000 Steps) (loss=1.70849): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (4679 / 30000 Steps) (loss=1.89590): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (4703 / 30000 Steps) (loss=1.58810): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (4727 / 30000 Steps) (loss=1.58057): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (4751 / 30000 Steps) (loss=1.63159): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (4775 / 30000 Steps) (loss=1.61231): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (4799 / 30000 Steps) (loss=1.57463): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (4823 / 30000 Steps) (loss=1.65194): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (4847 / 30000 Steps) (loss=1.70500): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (4871 / 30000 Steps) (loss=1.84604): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (4895 / 30000 Steps) (loss=1.57213): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (4919 / 30000 Steps) (loss=1.68219): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (4943 / 30000 Steps) (loss=2.06159): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (4967 / 30000 Steps) (loss=1.72085): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (4991 / 30000 Steps) (loss=1.89126): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5000 / 30000 Steps) (loss=1.59737):  33%|███▎      | 8/24 [00:09<00:14,  1.07it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (4992 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
-      "Validate (4992 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.09s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.09s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.66s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.66s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.14s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.14s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.10s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.10s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.77s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.77s/it]\u001b[A\n",
-      "Validate (4992 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.70s/it]\u001b[A\n",
-      "Training (5000 / 30000 Steps) (loss=1.59737):  38%|███▊      | 9/24 [00:31<01:58,  7.87s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.07061345130205154\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (5015 / 30000 Steps) (loss=1.81268): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
-      "Training (5039 / 30000 Steps) (loss=1.55078): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5063 / 30000 Steps) (loss=1.44579): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5087 / 30000 Steps) (loss=1.62221): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (5111 / 30000 Steps) (loss=2.66676): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5135 / 30000 Steps) (loss=2.57928): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5159 / 30000 Steps) (loss=1.67426): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5183 / 30000 Steps) (loss=1.69841): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (5207 / 30000 Steps) (loss=1.53911): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5231 / 30000 Steps) (loss=1.62084): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5255 / 30000 Steps) (loss=1.49868): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5279 / 30000 Steps) (loss=1.70572): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5303 / 30000 Steps) (loss=1.66650): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5327 / 30000 Steps) (loss=2.02091): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (5351 / 30000 Steps) (loss=1.63479): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5375 / 30000 Steps) (loss=1.77253): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5399 / 30000 Steps) (loss=1.70641): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (5423 / 30000 Steps) (loss=3.40272): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5447 / 30000 Steps) (loss=1.70207): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (5471 / 30000 Steps) (loss=1.50105): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5495 / 30000 Steps) (loss=1.67170): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5500 / 30000 Steps) (loss=1.56272):  17%|█▋        | 4/24 [00:06<00:22,  1.12s/it]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (5496 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
-      "Validate (5496 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.03s/it]\u001b[A\n",
-      "Validate (5496 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.03s/it]\u001b[A\n",
-      "Validate (5496 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (5496 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (5496 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (5496 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (5496 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (5496 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (5496 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (5496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (5496 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (5500 / 30000 Steps) (loss=1.56272):  21%|██        | 5/24 [00:28<02:51,  9.02s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.07060054689645767\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (5519 / 30000 Steps) (loss=1.59257): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
-      "Training (5543 / 30000 Steps) (loss=1.86619): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5567 / 30000 Steps) (loss=1.56463): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (5591 / 30000 Steps) (loss=1.95694): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5615 / 30000 Steps) (loss=1.88110): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (5639 / 30000 Steps) (loss=1.79884): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5663 / 30000 Steps) (loss=1.66443): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5687 / 30000 Steps) (loss=1.47059): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5711 / 30000 Steps) (loss=1.85159): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5735 / 30000 Steps) (loss=1.79990): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (5759 / 30000 Steps) (loss=1.96519): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (5783 / 30000 Steps) (loss=1.55542): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5807 / 30000 Steps) (loss=1.43398): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (5831 / 30000 Steps) (loss=1.57149): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (5855 / 30000 Steps) (loss=1.49405): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (5879 / 30000 Steps) (loss=1.72335): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (5883 / 30000 Steps) (loss=1.76143):  17%|█▋        | 4/24 [00:05<00:22,  1.10s/it]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (7175 / 30000 Steps) (loss=1.97261): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (7199 / 30000 Steps) (loss=1.58691): 100%|██████████| 24/24 [00:23<00:00,  1.00it/s]\n",
-      "Training (7223 / 30000 Steps) (loss=1.79031): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (7247 / 30000 Steps) (loss=1.60217): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7271 / 30000 Steps) (loss=2.03104): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (7295 / 30000 Steps) (loss=1.54362): 100%|██████████| 24/24 [00:23<00:00,  1.00it/s]\n",
-      "Training (7319 / 30000 Steps) (loss=1.54586): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7343 / 30000 Steps) (loss=1.68927): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7367 / 30000 Steps) (loss=2.33010): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7391 / 30000 Steps) (loss=1.55624): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (7415 / 30000 Steps) (loss=1.61111): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (7439 / 30000 Steps) (loss=1.79105): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (7463 / 30000 Steps) (loss=1.40654): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (7487 / 30000 Steps) (loss=1.52006): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (7500 / 30000 Steps) (loss=1.72726):  50%|█████     | 12/24 [00:13<00:10,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (7488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
-      "Validate (7488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.18s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.18s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.69s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.69s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.16s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.16s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.11s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.11s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.77s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.77s/it]\u001b[A\n",
-      "Validate (7488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.71s/it]\u001b[A\n",
-      "Training (7500 / 30000 Steps) (loss=1.72726):  54%|█████▍    | 13/24 [00:35<01:24,  7.65s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.08777529001235962\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (7511 / 30000 Steps) (loss=1.54710): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
-      "Training (7535 / 30000 Steps) (loss=1.55861): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7559 / 30000 Steps) (loss=1.42543): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (7583 / 30000 Steps) (loss=1.59021): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7607 / 30000 Steps) (loss=1.79987): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7631 / 30000 Steps) (loss=1.86244): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (7655 / 30000 Steps) (loss=1.61107): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (7679 / 30000 Steps) (loss=1.57211): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7703 / 30000 Steps) (loss=1.50199): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7727 / 30000 Steps) (loss=2.25318): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (7751 / 30000 Steps) (loss=1.59236): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7775 / 30000 Steps) (loss=1.68014): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7799 / 30000 Steps) (loss=1.36067): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7823 / 30000 Steps) (loss=1.66846): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7847 / 30000 Steps) (loss=1.79751): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (7871 / 30000 Steps) (loss=1.57973): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (7895 / 30000 Steps) (loss=1.90628): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (7919 / 30000 Steps) (loss=3.27129): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (7943 / 30000 Steps) (loss=1.73865): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (7967 / 30000 Steps) (loss=2.42319): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (7991 / 30000 Steps) (loss=1.71041): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (8000 / 30000 Steps) (loss=1.91273):  33%|███▎      | 8/24 [00:10<00:15,  1.05it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (7992 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
-      "Validate (7992 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.15s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.15s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.68s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.68s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.15s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.15s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.10s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.10s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (7992 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.71s/it]\u001b[A\n",
-      "Training (8000 / 30000 Steps) (loss=1.91273):  38%|███▊      | 9/24 [00:32<01:58,  7.91s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.09312240034341812\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (8015 / 30000 Steps) (loss=1.65287): 100%|██████████| 24/24 [00:46<00:00,  1.92s/it]\n",
-      "Training (8039 / 30000 Steps) (loss=1.75467): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (8063 / 30000 Steps) (loss=1.52946): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (8087 / 30000 Steps) (loss=1.99682): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (8111 / 30000 Steps) (loss=1.69648): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (8135 / 30000 Steps) (loss=1.58250): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (8159 / 30000 Steps) (loss=1.71409): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (8183 / 30000 Steps) (loss=1.61727): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (8207 / 30000 Steps) (loss=1.50567): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (8231 / 30000 Steps) (loss=1.67341): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (8255 / 30000 Steps) (loss=2.14218): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (8279 / 30000 Steps) (loss=1.36899): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (8303 / 30000 Steps) (loss=1.73868): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (8327 / 30000 Steps) (loss=1.77028): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (8351 / 30000 Steps) (loss=1.72490): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (8375 / 30000 Steps) (loss=5.67763): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (8399 / 30000 Steps) (loss=1.56018): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (8423 / 30000 Steps) (loss=1.92986): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (8447 / 30000 Steps) (loss=1.67452): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (8471 / 30000 Steps) (loss=1.36784): 100%|██████████| 24/24 [00:23<00:00,  1.00it/s]\n",
-      "Training (8495 / 30000 Steps) (loss=1.71360): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (8500 / 30000 Steps) (loss=1.65908):  17%|█▋        | 4/24 [00:06<00:23,  1.18s/it]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (8496 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
-      "Validate (8496 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:21,  4.20s/it]\u001b[A\n",
-      "Validate (8496 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:21,  4.20s/it]\u001b[A\n",
-      "Validate (8496 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.70s/it]\u001b[A\n",
-      "Validate (8496 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.70s/it]\u001b[A\n",
-      "Validate (8496 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.16s/it]\u001b[A\n",
-      "Validate (8496 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.16s/it]\u001b[A\n",
-      "Validate (8496 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.10s/it]\u001b[A\n",
-      "Validate (8496 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.10s/it]\u001b[A\n",
-      "Validate (8496 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (8496 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (8496 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.71s/it]\u001b[A\n",
-      "Training (8500 / 30000 Steps) (loss=1.65908):  21%|██        | 5/24 [00:28<02:52,  9.10s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10523679107427597 Current Avg. Dice: 0.08893509954214096\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (8517 / 30000 Steps) (loss=1.72599):  88%|████████▊ | 21/24 [00:44<00:02,  1.08it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (9839 / 30000 Steps) (loss=1.51064): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (9863 / 30000 Steps) (loss=1.97902): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (9887 / 30000 Steps) (loss=1.91444): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (9911 / 30000 Steps) (loss=1.44662): 100%|██████████| 24/24 [00:23<00:00,  1.00it/s]\n",
-      "Training (9935 / 30000 Steps) (loss=1.47294): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (9959 / 30000 Steps) (loss=1.47041): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (9983 / 30000 Steps) (loss=1.36296): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (10000 / 30000 Steps) (loss=2.28292):  67%|██████▋   | 16/24 [00:17<00:07,  1.11it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (9984 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]       \u001b[A\n",
-      "Validate (9984 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.13s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.13s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.67s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.67s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.15s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.15s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.10s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.10s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (9984 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.70s/it]\u001b[A\n",
-      "Training (10000 / 30000 Steps) (loss=2.28292):  71%|███████   | 17/24 [00:40<00:54,  7.83s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Saved ! Current Best Avg. Dice: 0.10929516702890396 Current Avg. Dice: 0.10929516702890396\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (10007 / 30000 Steps) (loss=1.96494): 100%|██████████| 24/24 [00:46<00:00,  1.95s/it]\n",
-      "Training (10031 / 30000 Steps) (loss=1.57745): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10055 / 30000 Steps) (loss=1.45904): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (10079 / 30000 Steps) (loss=1.63818): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (10103 / 30000 Steps) (loss=2.50689): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (10127 / 30000 Steps) (loss=1.73940): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (10151 / 30000 Steps) (loss=1.46289): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10175 / 30000 Steps) (loss=1.45769): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (10199 / 30000 Steps) (loss=1.78349): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (10223 / 30000 Steps) (loss=1.69970): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10247 / 30000 Steps) (loss=1.50493): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10271 / 30000 Steps) (loss=1.51015): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10295 / 30000 Steps) (loss=1.54784): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (10319 / 30000 Steps) (loss=1.37233): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (10343 / 30000 Steps) (loss=1.48155): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (10367 / 30000 Steps) (loss=1.54143): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (10391 / 30000 Steps) (loss=1.70770): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (10415 / 30000 Steps) (loss=2.30014): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (10439 / 30000 Steps) (loss=1.79295): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (10463 / 30000 Steps) (loss=1.55343): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10487 / 30000 Steps) (loss=2.00099): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (10500 / 30000 Steps) (loss=1.65427):  50%|█████     | 12/24 [00:13<00:10,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (10488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (10488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.13s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.13s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.67s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.67s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.14s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.14s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (10488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.69s/it]\u001b[A\n",
-      "Training (10500 / 30000 Steps) (loss=1.65427):  54%|█████▍    | 13/24 [00:35<01:23,  7.62s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10929516702890396 Current Avg. Dice: 0.10640474408864975\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (10511 / 30000 Steps) (loss=1.52513): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
-      "Training (10535 / 30000 Steps) (loss=1.67685): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (10559 / 30000 Steps) (loss=1.44793): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (10583 / 30000 Steps) (loss=3.96370): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (10607 / 30000 Steps) (loss=1.67806): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10631 / 30000 Steps) (loss=1.59580): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10655 / 30000 Steps) (loss=1.57280): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10679 / 30000 Steps) (loss=1.75091): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (10703 / 30000 Steps) (loss=1.66555): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (10727 / 30000 Steps) (loss=1.67437): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (10751 / 30000 Steps) (loss=1.44212): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (10775 / 30000 Steps) (loss=1.55636): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10799 / 30000 Steps) (loss=1.54452): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (10823 / 30000 Steps) (loss=1.62608): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (10847 / 30000 Steps) (loss=1.61273): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (10871 / 30000 Steps) (loss=1.66387): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (10895 / 30000 Steps) (loss=1.68224): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (10919 / 30000 Steps) (loss=1.69885): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (10943 / 30000 Steps) (loss=2.09697): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (10967 / 30000 Steps) (loss=5.33873): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (10991 / 30000 Steps) (loss=1.53754): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (11000 / 30000 Steps) (loss=1.61410):  33%|███▎      | 8/24 [00:10<00:15,  1.05it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (10992 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (10992 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.00s/it]\u001b[A\n",
-      "Validate (10992 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.00s/it]\u001b[A\n",
-      "Validate (10992 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.62s/it]\u001b[A\n",
-      "Validate (10992 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.62s/it]\u001b[A\n",
-      "Validate (10992 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.11s/it]\u001b[A\n",
-      "Validate (10992 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.11s/it]\u001b[A\n",
-      "Validate (10992 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (10992 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (10992 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (10992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (10992 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (11000 / 30000 Steps) (loss=1.61410):  38%|███▊      | 9/24 [00:32<01:57,  7.84s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.10929516702890396 Current Avg. Dice: 0.10596194118261337\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (11015 / 30000 Steps) (loss=2.03147): 100%|██████████| 24/24 [00:45<00:00,  1.91s/it]\n",
-      "Training (11039 / 30000 Steps) (loss=1.42184): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (11063 / 30000 Steps) (loss=1.60865): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (11087 / 30000 Steps) (loss=1.88724): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (11111 / 30000 Steps) (loss=1.37930): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (11135 / 30000 Steps) (loss=2.38266): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (11159 / 30000 Steps) (loss=1.86418): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (11173 / 30000 Steps) (loss=1.72598):  58%|█████▊    | 14/24 [00:14<00:08,  1.11it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (12479 / 30000 Steps) (loss=1.47541): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (12500 / 30000 Steps) (loss=1.64012):  83%|████████▎ | 20/24 [00:20<00:03,  1.12it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (12480 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (12480 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.03s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.03s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (12480 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.69s/it]\u001b[A\n",
-      "Training (12500 / 30000 Steps) (loss=1.64012):  88%|████████▊ | 21/24 [00:42<00:22,  7.55s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.11252329498529434\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (12503 / 30000 Steps) (loss=1.59704): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
-      "Training (12527 / 30000 Steps) (loss=1.49524): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (12551 / 30000 Steps) (loss=1.86410): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (12575 / 30000 Steps) (loss=2.06310): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (12599 / 30000 Steps) (loss=1.46205): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (12623 / 30000 Steps) (loss=2.00819): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (12647 / 30000 Steps) (loss=1.56068): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (12671 / 30000 Steps) (loss=1.91482): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (12695 / 30000 Steps) (loss=1.53840): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (12719 / 30000 Steps) (loss=1.78944): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (12743 / 30000 Steps) (loss=1.75246): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (12767 / 30000 Steps) (loss=2.23596): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (12791 / 30000 Steps) (loss=1.56143): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (12815 / 30000 Steps) (loss=2.48896): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (12839 / 30000 Steps) (loss=1.58305): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (12863 / 30000 Steps) (loss=2.21133): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (12887 / 30000 Steps) (loss=2.02335): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s] \n",
-      "Training (12911 / 30000 Steps) (loss=1.85969): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (12935 / 30000 Steps) (loss=1.82117): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (12959 / 30000 Steps) (loss=1.55216): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (12983 / 30000 Steps) (loss=1.80917): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (13000 / 30000 Steps) (loss=1.71191):  67%|██████▋   | 16/24 [00:17<00:07,  1.11it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (12984 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (12984 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.03s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.03s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (12984 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (13000 / 30000 Steps) (loss=1.71191):  71%|███████   | 17/24 [00:39<00:52,  7.55s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.09339217096567154\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (13007 / 30000 Steps) (loss=1.61073): 100%|██████████| 24/24 [00:45<00:00,  1.91s/it]\n",
-      "Training (13031 / 30000 Steps) (loss=1.64728): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13055 / 30000 Steps) (loss=1.88418): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13079 / 30000 Steps) (loss=2.15990): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (13103 / 30000 Steps) (loss=2.05648): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13127 / 30000 Steps) (loss=2.08542): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (13151 / 30000 Steps) (loss=1.66662): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (13175 / 30000 Steps) (loss=1.69000): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (13199 / 30000 Steps) (loss=1.54646): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13223 / 30000 Steps) (loss=1.91987): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (13247 / 30000 Steps) (loss=1.87953): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (13271 / 30000 Steps) (loss=1.74172): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (13295 / 30000 Steps) (loss=1.73502): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (13319 / 30000 Steps) (loss=2.05865): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13343 / 30000 Steps) (loss=1.57452): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (13367 / 30000 Steps) (loss=2.06786): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (13391 / 30000 Steps) (loss=1.71718): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (13415 / 30000 Steps) (loss=1.97281): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (13439 / 30000 Steps) (loss=1.58712): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13463 / 30000 Steps) (loss=1.71977): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s] \n",
-      "Training (13487 / 30000 Steps) (loss=1.44692): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (13500 / 30000 Steps) (loss=1.85141):  50%|█████     | 12/24 [00:13<00:10,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (13488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (13488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.06s/it]\u001b[A\n",
-      "Validate (13488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.06s/it]\u001b[A\n",
-      "Validate (13488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.65s/it]\u001b[A\n",
-      "Validate (13488 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.65s/it]\u001b[A\n",
-      "Validate (13488 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.13s/it]\u001b[A\n",
-      "Validate (13488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.13s/it]\u001b[A\n",
-      "Validate (13488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (13488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (13488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (13488 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (13488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (13500 / 30000 Steps) (loss=1.85141):  54%|█████▍    | 13/24 [00:35<01:23,  7.60s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.08462151885032654\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (13511 / 30000 Steps) (loss=1.71781): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
-      "Training (13535 / 30000 Steps) (loss=6.14073): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (13559 / 30000 Steps) (loss=1.55927): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13583 / 30000 Steps) (loss=1.83615): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (13607 / 30000 Steps) (loss=1.71865): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (13631 / 30000 Steps) (loss=1.63441): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (13655 / 30000 Steps) (loss=1.82784): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13679 / 30000 Steps) (loss=1.63667): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13703 / 30000 Steps) (loss=1.88138): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13727 / 30000 Steps) (loss=1.65120): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (13751 / 30000 Steps) (loss=1.85407): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13775 / 30000 Steps) (loss=1.69582): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (13799 / 30000 Steps) (loss=1.57240): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (13823 / 30000 Steps) (loss=1.62583): 100%|██████████| 24/24 [00:23<00:00,  1.11it/s]IOPub message rate exceeded.\n",
-      "The notebook server will temporarily stop sending output\n",
-      "to the client in order to avoid crashing it.\n",
-      "To change this limit, set the config variable\n",
-      "`--NotebookApp.iopub_msg_rate_limit`.\n",
-      "\n",
-      "Current values:\n",
-      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
-      "NotebookApp.rate_limit_window=3.0 (secs)\n",
-      "\n",
-      "Training (15143 / 30000 Steps) (loss=1.83970): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15167 / 30000 Steps) (loss=1.66505): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (15191 / 30000 Steps) (loss=1.57074): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15215 / 30000 Steps) (loss=1.72749): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15239 / 30000 Steps) (loss=1.56530): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (15263 / 30000 Steps) (loss=1.59868): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15287 / 30000 Steps) (loss=1.89288): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15311 / 30000 Steps) (loss=2.37840): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15335 / 30000 Steps) (loss=1.51758): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15359 / 30000 Steps) (loss=1.64895): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15383 / 30000 Steps) (loss=1.61334): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (15407 / 30000 Steps) (loss=2.79856): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15431 / 30000 Steps) (loss=1.62389): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (15455 / 30000 Steps) (loss=1.70353): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15479 / 30000 Steps) (loss=1.78808): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15500 / 30000 Steps) (loss=1.73282):  83%|████████▎ | 20/24 [00:20<00:03,  1.12it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (15480 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (15480 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.11s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.11s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.66s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.66s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.14s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.14s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (15480 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.69s/it]\u001b[A\n",
-      "Training (15500 / 30000 Steps) (loss=1.73282):  88%|████████▊ | 21/24 [00:42<00:22,  7.54s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.07921255379915237\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (15503 / 30000 Steps) (loss=2.00451): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
-      "Training (15527 / 30000 Steps) (loss=1.62787): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15551 / 30000 Steps) (loss=1.68343): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15575 / 30000 Steps) (loss=1.70755): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (15599 / 30000 Steps) (loss=1.93015): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15623 / 30000 Steps) (loss=5.34464): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15647 / 30000 Steps) (loss=1.73335): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15671 / 30000 Steps) (loss=1.84574): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (15695 / 30000 Steps) (loss=2.00053): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15719 / 30000 Steps) (loss=1.58993): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15743 / 30000 Steps) (loss=1.91924): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
-      "Training (15767 / 30000 Steps) (loss=1.78188): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15791 / 30000 Steps) (loss=2.37636): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (15815 / 30000 Steps) (loss=2.01652): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (15839 / 30000 Steps) (loss=1.62255): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15863 / 30000 Steps) (loss=2.15401): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (15887 / 30000 Steps) (loss=2.04368): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (15911 / 30000 Steps) (loss=1.54641): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (15935 / 30000 Steps) (loss=1.92847): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (15959 / 30000 Steps) (loss=5.25715): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (15983 / 30000 Steps) (loss=1.82406): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16000 / 30000 Steps) (loss=1.71104):  67%|██████▋   | 16/24 [00:16<00:07,  1.12it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (15984 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (15984 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.02s/it]\u001b[A\n",
-      "Validate (15984 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.02s/it]\u001b[A\n",
-      "Validate (15984 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (15984 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (15984 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (15984 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (15984 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (15984 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (15984 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (15984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (15984 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (16000 / 30000 Steps) (loss=1.71104):  71%|███████   | 17/24 [00:38<00:52,  7.53s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.09620887041091919\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (16007 / 30000 Steps) (loss=2.44960): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
-      "Training (16031 / 30000 Steps) (loss=2.16274): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (16055 / 30000 Steps) (loss=1.76895): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (16079 / 30000 Steps) (loss=1.91601): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16103 / 30000 Steps) (loss=1.95242): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (16127 / 30000 Steps) (loss=1.67782): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (16151 / 30000 Steps) (loss=2.04866): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16175 / 30000 Steps) (loss=1.67603): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16199 / 30000 Steps) (loss=1.59732): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16223 / 30000 Steps) (loss=1.73291): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (16247 / 30000 Steps) (loss=1.53592): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (16271 / 30000 Steps) (loss=1.87026): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16295 / 30000 Steps) (loss=1.68480): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (16319 / 30000 Steps) (loss=4.17143): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (16343 / 30000 Steps) (loss=1.85085): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16367 / 30000 Steps) (loss=1.99687): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16391 / 30000 Steps) (loss=1.83488): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16415 / 30000 Steps) (loss=1.67751): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (16439 / 30000 Steps) (loss=1.77609): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16463 / 30000 Steps) (loss=1.68322): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (16487 / 30000 Steps) (loss=2.76593): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16500 / 30000 Steps) (loss=1.76284):  50%|█████     | 12/24 [00:13<00:10,  1.10it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (16488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (16488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.14s/it]\u001b[A\n",
-      "Validate (16488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.14s/it]\u001b[A\n",
-      "Validate (16488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.67s/it]\u001b[A\n",
-      "Validate (16488 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.67s/it]\u001b[A\n",
-      "Validate (16488 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.15s/it]\u001b[A\n",
-      "Validate (16488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.15s/it]\u001b[A\n",
-      "Validate (16488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.10s/it]\u001b[A\n",
-      "Validate (16488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.10s/it]\u001b[A\n",
-      "Validate (16488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (16488 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (16488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.70s/it]\u001b[A\n",
-      "Training (16500 / 30000 Steps) (loss=1.76284):  54%|█████▍    | 13/24 [00:35<01:23,  7.63s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.098011814057827\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (16511 / 30000 Steps) (loss=1.59225): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
-      "Training (16535 / 30000 Steps) (loss=1.81074): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16559 / 30000 Steps) (loss=1.66221): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (16583 / 30000 Steps) (loss=1.65103): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16607 / 30000 Steps) (loss=1.78824): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (16631 / 30000 Steps) (loss=2.45114): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16655 / 30000 Steps) (loss=2.39868): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (16679 / 30000 Steps) (loss=1.75542): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (16703 / 30000 Steps) (loss=1.64762): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16727 / 30000 Steps) (loss=1.81028): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16751 / 30000 Steps) (loss=1.71452): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16775 / 30000 Steps) (loss=1.52099): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16799 / 30000 Steps) (loss=1.52047): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (16823 / 30000 Steps) (loss=1.86302): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (16847 / 30000 Steps) (loss=1.48955): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16871 / 30000 Steps) (loss=1.52795): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (16895 / 30000 Steps) (loss=1.51909): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16919 / 30000 Steps) (loss=1.65529): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (16943 / 30000 Steps) (loss=1.79362): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16967 / 30000 Steps) (loss=1.66853): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (16991 / 30000 Steps) (loss=1.53858): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (17000 / 30000 Steps) (loss=2.06464):  33%|███▎      | 8/24 [00:09<00:14,  1.07it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (16992 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (16992 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.05s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.05s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.64s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.64s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (16992 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (17000 / 30000 Steps) (loss=2.06464):  38%|███▊      | 9/24 [00:31<01:57,  7.83s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.0814722552895546\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (17015 / 30000 Steps) (loss=2.18220): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
-      "Training (17039 / 30000 Steps) (loss=2.03658): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (17063 / 30000 Steps) (loss=2.23238): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (17087 / 30000 Steps) (loss=1.74233): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (17111 / 30000 Steps) (loss=1.88046): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (17135 / 30000 Steps) (loss=1.68661): 100%|██████████| 24/24 [00:24<00:00,  1.01s/it]\n",
-      "Training (17159 / 30000 Steps) (loss=1.75913): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (17183 / 30000 Steps) (loss=1.90502): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (17207 / 30000 Steps) (loss=1.95610): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (17231 / 30000 Steps) (loss=8.81049): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (17255 / 30000 Steps) (loss=1.61983): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (17279 / 30000 Steps) (loss=2.39689): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (17303 / 30000 Steps) (loss=1.66409): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
-      "Training (17327 / 30000 Steps) (loss=2.19589): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (17351 / 30000 Steps) (loss=2.09781): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (17375 / 30000 Steps) (loss=1.80089): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (17399 / 30000 Steps) (loss=2.05174): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (17423 / 30000 Steps) (loss=2.08487): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (17447 / 30000 Steps) (loss=1.57698): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (17471 / 30000 Steps) (loss=1.92801): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (17495 / 30000 Steps) (loss=1.85291): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (17500 / 30000 Steps) (loss=1.74886):  17%|█▋        | 4/24 [00:06<00:22,  1.13s/it]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (17496 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (17496 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.02s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.02s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.08s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (17496 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (17500 / 30000 Steps) (loss=1.74886):  21%|██        | 5/24 [00:28<02:51,  9.01s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.08850126713514328\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (17519 / 30000 Steps) (loss=1.75699): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
-      "Training (17543 / 30000 Steps) (loss=1.79229): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (17567 / 30000 Steps) (loss=1.73489): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (17591 / 30000 Steps) (loss=1.83942): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (17615 / 30000 Steps) (loss=1.62040): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (17639 / 30000 Steps) (loss=1.86634): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (17783 / 30000 Steps) (loss=1.57847): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (17807 / 30000 Steps) (loss=1.63421): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (17831 / 30000 Steps) (loss=1.93467): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (17855 / 30000 Steps) (loss=1.90310): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (17879 / 30000 Steps) (loss=1.66063): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (17903 / 30000 Steps) (loss=2.17549): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s] \n",
-      "Training (17927 / 30000 Steps) (loss=2.14292): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (17951 / 30000 Steps) (loss=1.81027): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (17975 / 30000 Steps) (loss=1.52826): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (17999 / 30000 Steps) (loss=1.61435): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18000 / 30000 Steps) (loss=1.67718):   0%|          | 0/24 [00:02<?, ?it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (18000 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (18000 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.06s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.06s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.64s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.64s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.13s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.13s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.75s/it]\u001b[A\n",
-      "Validate (18000 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.69s/it]\u001b[A\n",
-      "Training (18000 / 30000 Steps) (loss=1.67718):   4%|▍         | 1/24 [00:24<09:23, 24.48s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.08198585361242294\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (18023 / 30000 Steps) (loss=2.14477): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
-      "Training (18047 / 30000 Steps) (loss=1.62933): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (18071 / 30000 Steps) (loss=2.26788): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (18095 / 30000 Steps) (loss=2.25271): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18119 / 30000 Steps) (loss=1.76527): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18143 / 30000 Steps) (loss=1.63859): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (18167 / 30000 Steps) (loss=2.15393): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (18191 / 30000 Steps) (loss=1.63407): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (18215 / 30000 Steps) (loss=1.72186): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (18239 / 30000 Steps) (loss=1.76017): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (18263 / 30000 Steps) (loss=1.71748): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18287 / 30000 Steps) (loss=2.37135): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (18311 / 30000 Steps) (loss=2.01968): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (18335 / 30000 Steps) (loss=1.87879): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18359 / 30000 Steps) (loss=1.72722): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (18383 / 30000 Steps) (loss=1.54340): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (18407 / 30000 Steps) (loss=1.68834): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18431 / 30000 Steps) (loss=1.72274): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
-      "Training (18455 / 30000 Steps) (loss=1.71656): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s]\n",
-      "Training (18479 / 30000 Steps) (loss=1.87595): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (18500 / 30000 Steps) (loss=1.63074):  83%|████████▎ | 20/24 [00:20<00:03,  1.11it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (18480 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (18480 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:19,  3.92s/it]\u001b[A\n",
-      "Validate (18480 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:19,  3.92s/it]\u001b[A\n",
-      "Validate (18480 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.59s/it]\u001b[A\n",
-      "Validate (18480 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.59s/it]\u001b[A\n",
-      "Validate (18480 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.10s/it]\u001b[A\n",
-      "Validate (18480 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.10s/it]\u001b[A\n",
-      "Validate (18480 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.07s/it]\u001b[A\n",
-      "Validate (18480 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.07s/it]\u001b[A\n",
-      "Validate (18480 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.74s/it]\u001b[A\n",
-      "Validate (18480 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.74s/it]\u001b[A\n",
-      "Validate (18480 / 10 Steps): 100%|██████████| 6/6 [00:21<00:00,  3.66s/it]\u001b[A\n",
-      "Training (18500 / 30000 Steps) (loss=1.63074):  88%|████████▊ | 21/24 [00:42<00:22,  7.50s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.07472795993089676\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (18503 / 30000 Steps) (loss=1.72294): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it]\n",
-      "Training (18527 / 30000 Steps) (loss=1.85699): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18551 / 30000 Steps) (loss=1.88507): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s] \n",
-      "Training (18575 / 30000 Steps) (loss=2.15083): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (18599 / 30000 Steps) (loss=1.70498): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (18623 / 30000 Steps) (loss=1.96934): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18647 / 30000 Steps) (loss=1.79790): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s] \n",
-      "Training (18671 / 30000 Steps) (loss=1.71376): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18695 / 30000 Steps) (loss=2.13562): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (18719 / 30000 Steps) (loss=1.73023): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18743 / 30000 Steps) (loss=1.96963): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s] \n",
-      "Training (18767 / 30000 Steps) (loss=2.29220): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (18791 / 30000 Steps) (loss=1.87343): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (18815 / 30000 Steps) (loss=1.95431): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (18839 / 30000 Steps) (loss=1.71830): 100%|██████████| 24/24 [00:22<00:00,  1.05it/s] \n",
-      "Training (18863 / 30000 Steps) (loss=2.06890): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18887 / 30000 Steps) (loss=1.74112): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (18911 / 30000 Steps) (loss=1.82975): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (18935 / 30000 Steps) (loss=1.69898): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (18959 / 30000 Steps) (loss=1.77620): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (18983 / 30000 Steps) (loss=2.40519): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (19000 / 30000 Steps) (loss=1.78521):  67%|██████▋   | 16/24 [00:16<00:07,  1.12it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (18984 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (18984 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.03s/it]\u001b[A\n",
-      "Validate (18984 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.03s/it]\u001b[A\n",
-      "Validate (18984 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (18984 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.63s/it]\u001b[A\n",
-      "Validate (18984 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (18984 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.12s/it]\u001b[A\n",
-      "Validate (18984 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (18984 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.09s/it]\u001b[A\n",
-      "Validate (18984 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (18984 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.76s/it]\u001b[A\n",
-      "Validate (18984 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.68s/it]\u001b[A\n",
-      "Training (19000 / 30000 Steps) (loss=1.78521):  71%|███████   | 17/24 [00:38<00:52,  7.55s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.08991578966379166\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (19007 / 30000 Steps) (loss=1.66867): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
-      "Training (19031 / 30000 Steps) (loss=2.67283): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19055 / 30000 Steps) (loss=1.83468): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (19079 / 30000 Steps) (loss=2.33405): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19103 / 30000 Steps) (loss=1.68956): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (19127 / 30000 Steps) (loss=2.36081): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (19151 / 30000 Steps) (loss=2.13139): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (19175 / 30000 Steps) (loss=2.03213): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (19199 / 30000 Steps) (loss=1.69451): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (19223 / 30000 Steps) (loss=2.16839): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19247 / 30000 Steps) (loss=1.87738): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
-      "Training (19271 / 30000 Steps) (loss=1.85485): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (19295 / 30000 Steps) (loss=3.11746): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (19319 / 30000 Steps) (loss=1.84075): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
-      "Training (19343 / 30000 Steps) (loss=1.92438): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (19367 / 30000 Steps) (loss=2.05224): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (19391 / 30000 Steps) (loss=1.60599): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19415 / 30000 Steps) (loss=2.37670): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (19439 / 30000 Steps) (loss=1.68553): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19463 / 30000 Steps) (loss=1.90457): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19487 / 30000 Steps) (loss=1.74032): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19500 / 30000 Steps) (loss=1.67537):  50%|█████     | 12/24 [00:13<00:10,  1.11it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (19488 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (19488 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:20,  4.14s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:20,  4.14s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.67s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:14,  3.67s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.15s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.15s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.11s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.11s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.77s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.77s/it]\u001b[A\n",
-      "Validate (19488 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.70s/it]\u001b[A\n",
-      "Training (19500 / 30000 Steps) (loss=1.67537):  54%|█████▍    | 13/24 [00:35<01:23,  7.63s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.07180359959602356\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (19511 / 30000 Steps) (loss=1.92059): 100%|██████████| 24/24 [00:45<00:00,  1.89s/it] \n",
-      "Training (19535 / 30000 Steps) (loss=2.07178): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (19559 / 30000 Steps) (loss=2.13432): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (19583 / 30000 Steps) (loss=2.24247): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (19607 / 30000 Steps) (loss=2.14878): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (19631 / 30000 Steps) (loss=1.80989): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (19655 / 30000 Steps) (loss=2.33858): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19679 / 30000 Steps) (loss=1.78372): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (19703 / 30000 Steps) (loss=1.82741): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19727 / 30000 Steps) (loss=2.23388): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (19751 / 30000 Steps) (loss=2.10488): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (19775 / 30000 Steps) (loss=1.85092): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19799 / 30000 Steps) (loss=2.06998): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19823 / 30000 Steps) (loss=2.44791): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (19847 / 30000 Steps) (loss=1.93086): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (19871 / 30000 Steps) (loss=1.82344): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19895 / 30000 Steps) (loss=2.08381): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (19919 / 30000 Steps) (loss=2.62876): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (19943 / 30000 Steps) (loss=1.78375): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (19967 / 30000 Steps) (loss=1.94173): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (19991 / 30000 Steps) (loss=1.64909): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (20000 / 30000 Steps) (loss=1.81226):  33%|███▎      | 8/24 [00:09<00:14,  1.07it/s]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (19992 / 10 Steps):   0%|          | 0/6 [00:03<?, ?it/s]      \u001b[A\n",
-      "Validate (19992 / 10 Steps):  17%|█▋        | 1/6 [00:03<00:19,  4.00s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:19,  4.00s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:14,  3.61s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  33%|███▎      | 2/6 [00:09<00:14,  3.61s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  50%|█████     | 3/6 [00:09<00:09,  3.11s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.11s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.07s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  67%|██████▋   | 4/6 [00:18<00:08,  4.07s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  83%|████████▎ | 5/6 [00:18<00:03,  3.74s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps):  83%|████████▎ | 5/6 [00:21<00:03,  3.74s/it]\u001b[A\n",
-      "Validate (19992 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.67s/it]\u001b[A\n",
-      "Training (20000 / 30000 Steps) (loss=1.81226):  38%|███▊      | 9/24 [00:31<01:57,  7.80s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.07197052240371704\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (20015 / 30000 Steps) (loss=1.89646): 100%|██████████| 24/24 [00:45<00:00,  1.88s/it]\n",
-      "Training (20039 / 30000 Steps) (loss=1.93255): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s] \n",
-      "Training (20063 / 30000 Steps) (loss=1.77097): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20087 / 30000 Steps) (loss=1.93809): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (20111 / 30000 Steps) (loss=1.66247): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (20135 / 30000 Steps) (loss=1.84815): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (20159 / 30000 Steps) (loss=1.83373): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20183 / 30000 Steps) (loss=1.77483): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20207 / 30000 Steps) (loss=1.73208): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (20231 / 30000 Steps) (loss=1.79430): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20255 / 30000 Steps) (loss=1.97952): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20279 / 30000 Steps) (loss=2.07711): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20303 / 30000 Steps) (loss=3.14760): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (20327 / 30000 Steps) (loss=1.96483): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20351 / 30000 Steps) (loss=1.94485): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (20375 / 30000 Steps) (loss=2.13720): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (20399 / 30000 Steps) (loss=1.92308): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (20423 / 30000 Steps) (loss=1.77186): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20447 / 30000 Steps) (loss=1.71648): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s] \n",
-      "Training (20471 / 30000 Steps) (loss=4.32078): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (20495 / 30000 Steps) (loss=1.72388): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20500 / 30000 Steps) (loss=1.91721):  17%|█▋        | 4/24 [00:05<00:21,  1.09s/it]\n",
-      "Validate (X / X Steps) (dice=X.X):   0%|          | 0/6 [00:00<?, ?it/s]\u001b[A\n",
-      "Validate (20496 / 10 Steps):   0%|          | 0/6 [00:04<?, ?it/s]      \u001b[A\n",
-      "Validate (20496 / 10 Steps):  17%|█▋        | 1/6 [00:04<00:22,  4.42s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  17%|█▋        | 1/6 [00:07<00:22,  4.42s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  33%|███▎      | 2/6 [00:07<00:15,  3.79s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  33%|███▎      | 2/6 [00:10<00:15,  3.79s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  50%|█████     | 3/6 [00:10<00:09,  3.21s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  50%|█████     | 3/6 [00:15<00:09,  3.21s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  67%|██████▋   | 4/6 [00:15<00:08,  4.14s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  67%|██████▋   | 4/6 [00:19<00:08,  4.14s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  83%|████████▎ | 5/6 [00:19<00:03,  3.79s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps):  83%|████████▎ | 5/6 [00:22<00:03,  3.79s/it]\u001b[A\n",
-      "Validate (20496 / 10 Steps): 100%|██████████| 6/6 [00:22<00:00,  3.75s/it]\u001b[A\n",
-      "Training (20500 / 30000 Steps) (loss=1.91721):  21%|██        | 5/24 [00:28<02:53,  9.13s/it]"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model Was Not Saved ! Current Best Avg. Dice: 0.1141304075717926 Current Avg. Dice: 0.08904116600751877\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Training (20519 / 30000 Steps) (loss=3.33243): 100%|██████████| 24/24 [00:45<00:00,  1.90s/it]\n",
-      "Training (20543 / 30000 Steps) (loss=1.97818): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20567 / 30000 Steps) (loss=1.89210): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (20591 / 30000 Steps) (loss=1.64444): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20615 / 30000 Steps) (loss=2.15475): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (20639 / 30000 Steps) (loss=1.87721): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s] \n",
-      "Training (20663 / 30000 Steps) (loss=1.84025): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (20687 / 30000 Steps) (loss=1.59353): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (20711 / 30000 Steps) (loss=1.84514): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20735 / 30000 Steps) (loss=1.75540): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (20759 / 30000 Steps) (loss=1.79015): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (20783 / 30000 Steps) (loss=2.06733): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20807 / 30000 Steps) (loss=2.68113): 100%|██████████| 24/24 [00:23<00:00,  1.02it/s]\n",
-      "Training (20831 / 30000 Steps) (loss=1.85298): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s] \n",
-      "Training (20855 / 30000 Steps) (loss=1.60886): 100%|██████████| 24/24 [00:22<00:00,  1.04it/s]\n",
-      "Training (20879 / 30000 Steps) (loss=1.65883): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (20903 / 30000 Steps) (loss=1.73903): 100%|██████████| 24/24 [00:23<00:00,  1.03it/s]\n",
-      "Training (20927 / 30000 Steps) (loss=2.25431): 100%|██████████| 24/24 [00:23<00:00,  1.01it/s]\n",
-      "Training (20951 / 30000 Steps) (loss=2.83517): 100%|██████████| 24/24 [00:23<00:00,  1.04it/s]\n",
-      "Training (20965 / 30000 Steps) (loss=1.50596):  58%|█████▊    | 14/24 [00:14<00:10,  1.06s/it]\n",
-      "\n",
-      "KeyboardInterrupt\n",
-      "\n"
-     ]
-    }
-   ],
-   "source": [
-    "def validation(epoch_iterator_val):\n",
-    "    model.eval()\n",
-    "    with torch.no_grad():\n",
-    "        for step, batch in enumerate(epoch_iterator_val):\n",
-    "            val_inputs, val_labels = (batch[\"image\"].cuda(), batch[\"label\"].cuda())\n",
-    "            with torch.cuda.amp.autocast():\n",
-    "                val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)\n",
-    "            val_labels_list = decollate_batch(val_labels)\n",
-    "            val_labels_convert = [\n",
-    "                post_label(val_label_tensor) for val_label_tensor in val_labels_list\n",
-    "            ]\n",
-    "            val_outputs_list = decollate_batch(val_outputs)\n",
-    "            val_output_convert = [\n",
-    "                post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list\n",
-    "            ]\n",
-    "            dice_metric(y_pred=val_output_convert, y=val_labels_convert)\n",
-    "            epoch_iterator_val.set_description(\n",
-    "                \"Validate (%d / %d Steps)\" % (global_step, 10.0)\n",
-    "            )\n",
-    "        mean_dice_val = dice_metric.aggregate().item()\n",
-    "        dice_metric.reset()\n",
-    "    return mean_dice_val\n",
-    "\n",
-    "\n",
-    "def train(global_step, train_loader, dice_val_best, global_step_best):\n",
-    "    model.train()\n",
-    "    epoch_loss = 0\n",
-    "    step = 0\n",
-    "    epoch_iterator = tqdm(\n",
-    "        train_loader, desc=\"Training (X / X Steps) (loss=X.X)\", dynamic_ncols=True\n",
-    "    )\n",
-    "    for step, batch in enumerate(epoch_iterator):\n",
-    "        step += 1\n",
-    "        x, y = (batch[\"image\"].cuda(), batch[\"label\"].cuda())\n",
-    "        with torch.cuda.amp.autocast():\n",
-    "            logit_map = model(x)\n",
-    "            loss = loss_function(logit_map, y)\n",
-    "        loss.backward()\n",
-    "        epoch_loss += loss.item()\n",
-    "        optimizer.step()\n",
-    "        optimizer.zero_grad()\n",
-    "        epoch_iterator.set_description(\n",
-    "            \"Training (%d / %d Steps) (loss=%2.5f)\"\n",
-    "            % (global_step, max_iterations, loss)\n",
-    "        )\n",
-    "        if (\n",
-    "            global_step % eval_num == 0 and global_step != 0\n",
-    "        ) or global_step == max_iterations:\n",
-    "            epoch_iterator_val = tqdm(\n",
-    "                val_loader, desc=\"Validate (X / X Steps) (dice=X.X)\", dynamic_ncols=True\n",
-    "            )\n",
-    "            dice_val = validation(epoch_iterator_val)\n",
-    "            epoch_loss /= step\n",
-    "            epoch_loss_values.append(epoch_loss)\n",
-    "            metric_values.append(dice_val)\n",
-    "            if dice_val > dice_val_best:\n",
-    "                dice_val_best = dice_val\n",
-    "                global_step_best = global_step\n",
-    "                torch.save(\n",
-    "                    model.state_dict(), os.path.join(root_dir, \"best_metric_model.pth\")\n",
-    "                )\n",
-    "                print(\n",
-    "                    \"Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}\".format(\n",
-    "                        dice_val_best, dice_val\n",
-    "                    )\n",
-    "                )\n",
-    "            else:\n",
-    "                print(\n",
-    "                    \"Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}\".format(\n",
-    "                        dice_val_best, dice_val\n",
-    "                    )\n",
-    "                )\n",
-    "        global_step += 1\n",
-    "    return global_step, dice_val_best, global_step_best\n",
-    "\n",
-    "\n",
-    "max_iterations = 30000\n",
-    "eval_num = 500\n",
-    "post_label = AsDiscrete(to_onehot=14)\n",
-    "post_pred = AsDiscrete(argmax=True, to_onehot=14)\n",
-    "dice_metric = DiceMetric(include_background=True, reduction=\"mean\", get_not_nans=False)\n",
-    "global_step = 0\n",
-    "dice_val_best = 0.0\n",
-    "global_step_best = 0\n",
-    "epoch_loss_values = []\n",
-    "metric_values = []\n",
-    "begin = time.time()\n",
-    "while global_step < max_iterations:\n",
-    "    global_step, dice_val_best, global_step_best = train(\n",
-    "        global_step, train_loader, dice_val_best, global_step_best\n",
-    "    )\n",
-    "end = time.time()\n",
-    "print(f\"Total train time: {end - begin:.2f} seconds\")\n",
-    "# model.load_state_dict(torch.load(os.path.join(root_dir, \"best_metric_model.pth\")))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 12,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "train completed, best_metric: 0.8452 at iteration: 28500\n"
-     ]
-    }
-   ],
-   "source": [
-    "print(\n",
-    "    f\"train completed, best_metric: {dice_val_best:.4f} \"\n",
-    "    f\"at iteration: {global_step_best}\"\n",
-    ")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 35,
-   "metadata": {
-    "scrolled": true
-   },
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Fri Jul 15 04:29:00 2022       \n",
-      "+-----------------------------------------------------------------------------+\n",
-      "| NVIDIA-SMI 450.119.04   Driver Version: 450.119.04   CUDA Version: 11.6     |\n",
-      "|-------------------------------+----------------------+----------------------+\n",
-      "| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
-      "| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |\n",
-      "|                               |                      |               MIG M. |\n",
-      "|===============================+======================+======================|\n",
-      "|   0  Tesla V100-SXM2...  On   | 00000000:89:00.0 Off |                    0 |\n",
-      "| N/A   38C    P0    48W / 163W |  32494MiB / 32510MiB |      0%      Default |\n",
-      "|                               |                      |                  N/A |\n",
-      "+-------------------------------+----------------------+----------------------+\n",
-      "                                                                               \n",
-      "+-----------------------------------------------------------------------------+\n",
-      "| Processes:                                                                  |\n",
-      "|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |\n",
-      "|        ID   ID                                                   Usage      |\n",
-      "|=============================================================================|\n",
-      "+-----------------------------------------------------------------------------+\n"
-     ]
-    }
-   ],
-   "source": [
-    "torch.cuda.empty_cache()\n",
-    "\n",
-    "!nvidia-smi"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 11,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "'Tesla V100-SXM2-32GB-LS'"
-      ]
-     },
-     "execution_count": 11,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "torch.cuda.get_device_name(0)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Plot the loss and metric"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 13,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 864x432 with 2 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "plt.figure(\"train\", (12, 6))\n",
-    "plt.subplot(1, 2, 1)\n",
-    "plt.title(\"Iteration Average Loss\")\n",
-    "x = [eval_num * (i + 1) for i in range(len(epoch_loss_values))]\n",
-    "y = epoch_loss_values\n",
-    "plt.xlabel(\"Iteration\")\n",
-    "plt.plot(x, y)\n",
-    "plt.subplot(1, 2, 2)\n",
-    "plt.title(\"Val Mean Dice\")\n",
-    "x = [eval_num * (i + 1) for i in range(len(metric_values))]\n",
-    "y = metric_values\n",
-    "plt.xlabel(\"Iteration\")\n",
-    "plt.plot(x, y)\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Check best model output with the input image and label"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 14,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 1296x432 with 3 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "case_num = 4\n",
-    "model.load_state_dict(torch.load(os.path.join(root_dir, \"best_metric_model.pth\")))\n",
-    "model.eval()\n",
-    "with torch.no_grad():\n",
-    "    img_name = os.path.split(val_ds[case_num][\"image_meta_dict\"][\"filename_or_obj\"])[1]\n",
-    "    img = val_ds[case_num][\"image\"]\n",
-    "    label = val_ds[case_num][\"label\"]\n",
-    "    val_inputs = torch.unsqueeze(img, 1).cuda()\n",
-    "    val_labels = torch.unsqueeze(label, 1).cuda()\n",
-    "    val_outputs = sliding_window_inference(\n",
-    "        val_inputs, (96, 96, 96), 4, model, overlap=0.8\n",
-    "    )\n",
-    "    plt.figure(\"check\", (18, 6))\n",
-    "    plt.subplot(1, 3, 1)\n",
-    "    plt.title(\"image\")\n",
-    "    plt.imshow(val_inputs.cpu().numpy()[0, 0, :, :, slice_map[img_name]], cmap=\"gray\")\n",
-    "    plt.subplot(1, 3, 2)\n",
-    "    plt.title(\"label\")\n",
-    "    plt.imshow(val_labels.cpu().numpy()[0, 0, :, :, slice_map[img_name]])\n",
-    "    plt.subplot(1, 3, 3)\n",
-    "    plt.title(\"output\")\n",
-    "    plt.imshow(\n",
-    "        torch.argmax(val_outputs, dim=1).detach().cpu()[0, :, :, slice_map[img_name]]\n",
-    "    )\n",
-    "    plt.show()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Cleanup data directory\n",
-    "\n",
-    "Remove directory if a temporary was used."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 15,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "if directory is None:\n",
-    "    shutil.rmtree(root_dir)"
-   ]
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "Python 3 (ipykernel)",
-   "language": "python",
-   "name": "python3"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.8.13"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 4
-}
diff --git a/3d_segmentation/swin_unetr_profiling.py b/3d_segmentation/swin_unetr_profiling.py
deleted file mode 100644
index adac531bcd..0000000000
--- a/3d_segmentation/swin_unetr_profiling.py
+++ /dev/null
@@ -1,333 +0,0 @@
-# command line: "pip install nibabel==3.1.1; pip install tqdm==4.63.0; nsys profile --output /results/test_output --force-overwrite true --trace-fork-before-exec true python3 swin_unetr_profiling.py --epochs 5 --val_epochs 5 --batch_size 1 --thread_workers False --num_workers 0"
-
-import os
-import shutil
-import tempfile
-
-import matplotlib.pyplot as plt
-import numpy as np
-from tqdm import tqdm
-import time
-
-from monai.losses import DiceCELoss
-from monai.inferers import sliding_window_inference
-from monai.transforms import (
-    AsDiscrete,
-    AddChanneld,
-    Compose,
-    CropForegroundd,
-    LoadImaged,
-    Orientationd,
-    RandFlipd,
-    RandCropByPosNegLabeld,
-    RandShiftIntensityd,
-    ScaleIntensityRanged,
-    Spacingd,
-    RandRotate90d,
-    ToTensord,
-    EnsureTyped,
-    ToDeviced
-)
-from monai.utils import set_determinism
-
-from monai.config import print_config
-from monai.metrics import DiceMetric
-from monai.networks.nets import SwinUNETR
-
-from monai.data import (
-    ThreadDataLoader,
-    DataLoader,
-    CacheDataset,
-    load_decathlon_datalist,
-    decollate_batch,
-)
-
-import nvtx
-from monai.utils.nvtx import Range
-
-import torch
-
-import argparse
-
-print_config()
-
-parser = argparse.ArgumentParser(description='Profiling Swin UNETR.')
-parser.add_argument('--epochs', default=5, type=int)
-parser.add_argument('--val_epochs', default=-1, type=int, help='validation every X epochs; if non-positive value entered, will perform validation only once')
-parser.add_argument('--batch_size', default=1, type=int)
-parser.add_argument('--thread_workers', default=False, type=bool)
-parser.add_argument('--num_workers', default=0, type=int)
-args = parser.parse_args()
-print(args)
-
-assert args.epochs > 0
-assert args.batch_size > 0
-assert args.num_workers >= 0
-
-if not args.thread_workers:
-    args.num_workers = 0
-
-max_iterations = args.epochs * args.batch_size
-eval_num = args.val_epochs * args.batch_size - 1
-
-
-
-directory = os.environ.get("MONAI_DATA_DIRECTORY")
-root_dir = tempfile.mkdtemp() if directory is None else directory
-print(root_dir)
-
-
-num_samples = 4
-os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-
-train_transforms = Compose(
-    [
-        Range()(LoadImaged(keys=["image", "label"])),
-        Range()(AddChanneld(keys=["image", "label"])),
-        Range()(Orientationd(keys=["image", "label"], axcodes="RAS")),
-        Range()(Spacingd(
-            keys=["image", "label"],
-            pixdim=(1.5, 1.5, 2.0),
-            mode=("bilinear", "nearest"),
-        )),
-        Range()(ScaleIntensityRanged(
-            keys=["image"],
-            a_min=-175,
-            a_max=250,
-            b_min=0.0,
-            b_max=1.0,
-            clip=True,
-        )),
-        Range()(CropForegroundd(keys=["image", "label"], source_key="image")),
-#         Range()(EnsureTyped(keys=["image", "label"])),
-#         Range()(ToDeviced(keys=["image", "label"], device=device)),
-        Range()(RandCropByPosNegLabeld(
-            keys=["image", "label"],
-            label_key="label",
-            spatial_size=(96, 96, 96),
-            pos=1,
-            neg=1,
-            num_samples=num_samples,
-            image_key="image",
-            image_threshold=0,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[0],
-            prob=0.10,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[1],
-            prob=0.10,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[2],
-            prob=0.10,
-        )),
-        Range()(RandRotate90d(
-            keys=["image", "label"],
-            prob=0.10,
-            max_k=3,
-        )),
-        Range()(RandShiftIntensityd(
-            keys=["image"],
-            offsets=0.10,
-            prob=0.50,
-        )),
-        Range()(ToTensord(keys=["image", "label"])),
-    ]
-)
-val_transforms = Compose(
-    [
-        LoadImaged(keys=["image", "label"]),
-        AddChanneld(keys=["image", "label"]),
-        Orientationd(keys=["image", "label"], axcodes="RAS"),
-        Spacingd(
-            keys=["image", "label"],
-            pixdim=(1.5, 1.5, 2.0),
-            mode=("bilinear", "nearest"),
-        ),
-        ScaleIntensityRanged(
-            keys=["image"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True
-        ),
-        CropForegroundd(keys=["image", "label"], source_key="image"),
-#        ToTensord(keys=["image", "label"]),
-        EnsureTyped(keys=["image", "label"]),
-        ToDeviced(keys=["image", "label"], device=device),
-    ]
-)
-
-
-torch.cuda.empty_cache()
-data_dir = "data/"
-split_JSON = "dataset_0.json"
-
-datasets = data_dir + split_JSON
-datalist = load_decathlon_datalist(datasets, True, "training")
-val_files = load_decathlon_datalist(datasets, True, "validation")
-
-# TODO: try thread_workers
-train_ds = CacheDataset(
-    data=datalist,
-    transform=train_transforms,
-    cache_num=24,
-    cache_rate=1.0,
-    num_workers=8,
-)
-# train_loader = ThreadDataLoader(train_ds, batch_size=1, shuffle=True, num_workers=0)
-train_loader = ThreadDataLoader(train_ds, batch_size=args.batch_size, shuffle=True, use_thread_workers=args.thread_workers, num_workers=args.num_workers)
-
-val_ds = CacheDataset(
-    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4
-)
-val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)
-
-
-model = SwinUNETR(
-    img_size=(96, 96, 96),
-    in_channels=1,
-    out_channels=14,
-    feature_size=48,
-    use_checkpoint=True,
-).to(device)
-
-
-weight = torch.load("./model_swinvit.pt")
-model.load_from(weights=weight)
-print("Using pretrained self-supervied Swin UNETR backbone weights !")
-
-
-torch.backends.cudnn.benchmark = True
-loss_function = DiceCELoss(to_onehot_y=True, softmax=True)
-optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)
-
-post_label = AsDiscrete(to_onehot=14)
-post_pred = AsDiscrete(argmax=True, to_onehot=14)
-dice_metric = DiceMetric(include_background=True, reduction="mean", get_not_nans=False)
-
-def validation(epoch_iterable_val):
-    model.eval()
-    epoch_iterator_val = iter(epoch_iterable_val)
-    with torch.no_grad():
-        for _ in range(len(epoch_iterable_val)):
-            with nvtx.annotate("val dataload", color="red"):
-                batch = next(epoch_iterator_val)
-                val_inputs, val_labels = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
-
-            with nvtx.annotate("sliding window", color="green"):
-#                 with torch.cuda.amp.autocast():
-                val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
-
-            with nvtx.annotate("decollate batch", color="blue"):
-                val_labels_list = decollate_batch(val_labels)
-                val_labels_convert = [
-                    post_label(val_label_tensor) for val_label_tensor in val_labels_list
-                ]
-                val_outputs_list = decollate_batch(val_outputs)
-                val_output_convert = [
-                    post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list
-                ]
-
-            with nvtx.annotate("compute metric", color="yellow"):
-                dice_metric(y_pred=val_output_convert, y=val_labels_convert)
-
-            epoch_iterable_val.set_description(
-                "Validate (%d / %d Steps)" % (global_step, 10.0)
-            )
-
-        mean_dice_val = dice_metric.aggregate().item()
-        dice_metric.reset()
-    return mean_dice_val
-
-
-def train(global_step, train_loader, dice_val_best, global_step_best):
-    model.train()
-    epoch_loss = 0
-    step = 0
-    epoch_iterable = tqdm(
-        train_loader, desc="Training (X / X Steps) (loss=X.X)", dynamic_ncols=True
-    )
-    epoch_iterator = iter(epoch_iterable)
-
-    for step in range(len(epoch_iterable)):
-        step += 1
-
-        with nvtx.annotate("dataload", color="red"):
-            batch = next(epoch_iterator)
-            x, y = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
-#         with torch.cuda.amp.autocast():
-        with nvtx.annotate("forward", color="green"):
-            logit_map = model(x)
-            loss = loss_function(logit_map, y)
-
-        with nvtx.annotate("backward", color="blue"):
-            loss.backward()
-            epoch_loss += loss.item()
-
-        with nvtx.annotate("update", color="yellow"):
-            optimizer.step()
-            optimizer.zero_grad()
-
-        epoch_iterable.set_description(
-            "Training (%d / %d Steps) (loss=%2.5f)"
-            % (global_step, max_iterations, loss)
-        )
-        if (
-            global_step % eval_num == 0 and global_step != 0
-        ) or global_step == max_iterations:
-            epoch_iterable_val = tqdm(
-                val_loader, desc="Validate (X / X Steps) (dice=X.X)", dynamic_ncols=True
-            )
-            dice_val = validation(epoch_iterable_val)
-            # FIXME: epoch_loss is a running average at time of validation??
-            epoch_loss /= step
-            epoch_loss_values.append(epoch_loss)
-            metric_values.append(dice_val)
-            if dice_val > dice_val_best:
-                dice_val_best = dice_val
-                global_step_best = global_step
-                torch.save(
-                    model.state_dict(), os.path.join(root_dir, "best_metric_model.pth")
-                )
-                print(
-                    "Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
-                        dice_val_best, dice_val
-                    )
-                )
-            else:
-                print(
-                    "Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
-                        dice_val_best, dice_val
-                    )
-                )
-        global_step += 1
-    return global_step, dice_val_best, global_step_best
-
-# max_iterations = 30000
-# eval_num = 500
-set_determinism(seed=0)
-global_step = 0
-dice_val_best = 0.0
-global_step_best = 0
-epoch_loss_values = []
-metric_values = []
-
-begin = time.time()
-while global_step < max_iterations:
-    with nvtx.annotate("epoch", color="red"):
-        global_step, dice_val_best, global_step_best = train(
-            global_step, train_loader, dice_val_best, global_step_best
-        )
-print(f"Total train time: {time.time() - begin:.2f} seconds")
-
-
-print(
-    f"train completed, best_metric: {dice_val_best:.4f} "
-    f"at iteration: {global_step_best}"
-)
-
-if directory is None:
-    shutil.rmtree(root_dir)

From 5783c330214d1e06ca0a5f47fd21870f515cae00 Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Tue, 26 Jul 2022 21:44:59 -0700
Subject: [PATCH 10/17] updated with set_track_meta

---
 3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
index 1c2a77f398..9e1567095d 100644
--- a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
+++ b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
@@ -148,6 +148,7 @@
     "    CacheDataset,\n",
     "    load_decathlon_datalist,\n",
     "    decollate_batch,\n",
+    "    set_track_meta,\n",
     ")\n",
     "\n",
     "\n",
@@ -318,7 +319,9 @@
     "val_ds = CacheDataset(\n",
     "    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4\n",
     ")\n",
-    "val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)"
+    "val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)\n",
+    "\n",
+    "set_track_meta(False)"
    ]
   },
   {

From 3b35850c3866420a22667b1bc93d1b93a89b0f8f Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Tue, 26 Jul 2022 21:45:33 -0700
Subject: [PATCH 11/17] tmp backup for profiling code

---
 3d_segmentation/swin_unetr_profiling.py     | 338 ++++++++++++++++++++
 3d_segmentation/swin_unetr_profiling_new.py | 331 +++++++++++++++++++
 2 files changed, 669 insertions(+)
 create mode 100644 3d_segmentation/swin_unetr_profiling.py
 create mode 100644 3d_segmentation/swin_unetr_profiling_new.py

diff --git a/3d_segmentation/swin_unetr_profiling.py b/3d_segmentation/swin_unetr_profiling.py
new file mode 100644
index 0000000000..75acf6033d
--- /dev/null
+++ b/3d_segmentation/swin_unetr_profiling.py
@@ -0,0 +1,338 @@
+# command line: "pip install monai==0.9.1; nsys profile --output /results/orig --force-overwrite true --trace-fork-before-exec true python3 swin_unetr_profiling.py --epochs 5 --val_epochs 5 --batch_size 1 --thread_workers False --num_workers 0"
+
+import os
+import shutil
+import tempfile
+
+import matplotlib.pyplot as plt
+import numpy as np
+from tqdm import tqdm
+import time
+
+from monai.losses import DiceCELoss
+from monai.inferers import sliding_window_inference
+from monai.transforms import (
+    AsDiscrete,
+    AddChanneld,
+    Compose,
+    CropForegroundd,
+    LoadImaged,
+    Orientationd,
+    RandFlipd,
+    RandCropByPosNegLabeld,
+    RandShiftIntensityd,
+    ScaleIntensityRanged,
+    Spacingd,
+    RandRotate90d,
+    ToTensord,
+    EnsureTyped,
+    ToDeviced
+)
+from monai.utils import set_determinism
+
+from monai.config import print_config
+from monai.metrics import DiceMetric
+from monai.networks.nets import SwinUNETR
+
+from monai.data import (
+#    ThreadDataLoader,
+    DataLoader,
+    CacheDataset,
+    load_decathlon_datalist,
+    decollate_batch,
+)
+
+import nvtx
+from monai.utils.nvtx import Range
+
+import torch
+
+import argparse
+
+print_config()
+
+parser = argparse.ArgumentParser(description='Profiling Swin UNETR.')
+parser.add_argument('--epochs', default=5, type=int)
+parser.add_argument('--val_epochs', default=-1, type=int, help='validation every X epochs; if non-positive value entered, will perform validation only once')
+parser.add_argument('--batch_size', default=1, type=int)
+parser.add_argument('--thread_workers', default=False, type=bool)
+parser.add_argument('--num_workers', default=0, type=int)
+args = parser.parse_args()
+print(args)
+
+assert args.epochs > 0
+assert args.batch_size > 0
+assert args.num_workers >= 0
+
+if not args.thread_workers:
+    args.num_workers = 0
+
+max_iterations = args.epochs * args.batch_size
+eval_num = args.val_epochs * args.batch_size - 1
+
+
+
+directory = os.environ.get("MONAI_DATA_DIRECTORY")
+root_dir = tempfile.mkdtemp() if directory is None else directory
+print(root_dir)
+
+
+num_samples = 4
+os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+train_transforms = Compose(
+    [
+        Range()(LoadImaged(keys=["image", "label"])),
+        Range()(AddChanneld(keys=["image", "label"])),
+        Range()(Orientationd(keys=["image", "label"], axcodes="RAS")),
+        Range()(Spacingd(
+            keys=["image", "label"],
+            pixdim=(1.5, 1.5, 2.0),
+            mode=("bilinear", "nearest"),
+        )),
+        Range()(ScaleIntensityRanged(
+            keys=["image"],
+            a_min=-175,
+            a_max=250,
+            b_min=0.0,
+            b_max=1.0,
+            clip=True,
+        )),
+        Range()(CropForegroundd(keys=["image", "label"], source_key="image")),
+#         Range()(EnsureTyped(keys=["image", "label"])),
+#         Range()(ToDeviced(keys=["image", "label"], device=device)),
+        Range()(RandCropByPosNegLabeld(
+            keys=["image", "label"],
+            label_key="label",
+            spatial_size=(96, 96, 96),
+            pos=1,
+            neg=1,
+            num_samples=num_samples,
+            image_key="image",
+            image_threshold=0,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[0],
+            prob=0.10,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[1],
+            prob=0.10,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[2],
+            prob=0.10,
+        )),
+        Range()(RandRotate90d(
+            keys=["image", "label"],
+            prob=0.10,
+            max_k=3,
+        )),
+        Range()(RandShiftIntensityd(
+            keys=["image"],
+            offsets=0.10,
+            prob=0.50,
+        )),
+#        Range()(ToTensord(keys=["image", "label"])),
+    ]
+)
+val_transforms = Compose(
+    [
+        LoadImaged(keys=["image", "label"]),
+        AddChanneld(keys=["image", "label"]),
+        Orientationd(keys=["image", "label"], axcodes="RAS"),
+        Spacingd(
+            keys=["image", "label"],
+            pixdim=(1.5, 1.5, 2.0),
+            mode=("bilinear", "nearest"),
+        ),
+        ScaleIntensityRanged(
+            keys=["image"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True
+        ),
+        CropForegroundd(keys=["image", "label"], source_key="image"),
+#        ToTensord(keys=["image", "label"]),
+#        EnsureTyped(keys=["image", "label"]),
+#        ToDeviced(keys=["image", "label"], device=device),
+    ]
+)
+
+
+torch.cuda.empty_cache()
+data_dir = "data/"
+split_JSON = "dataset_0.json"
+
+datasets = data_dir + split_JSON
+datalist = load_decathlon_datalist(datasets, True, "training")
+val_files = load_decathlon_datalist(datasets, True, "validation")
+
+# TODO: try thread_workers
+train_ds = CacheDataset(
+    data=datalist,
+    transform=train_transforms,
+    cache_num=24,
+    cache_rate=1.0,
+    num_workers=8,
+)
+# train_loader = ThreadDataLoader(train_ds, batch_size=1, shuffle=True, num_workers=0)
+# train_loader = ThreadDataLoader(train_ds, batch_size=args.batch_size, shuffle=True, use_thread_workers=args.thread_workers, num_workers=args.num_workers)
+train_loader = DataLoader(
+    train_ds, batch_size=args.batch_size, shuffle=True, num_workers=8, pin_memory=True
+)
+
+val_ds = CacheDataset(
+    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4
+)
+# val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)
+val_loader = DataLoader(
+    val_ds, batch_size=args.batch_size, shuffle=False, num_workers=4, pin_memory=True
+)
+
+model = SwinUNETR(
+    img_size=(96, 96, 96),
+    in_channels=1,
+    out_channels=14,
+    feature_size=48,
+    use_checkpoint=True,
+).to(device)
+
+
+weight = torch.load("./model_swinvit.pt")
+model.load_from(weights=weight)
+print("Using pretrained self-supervied Swin UNETR backbone weights !")
+
+
+torch.backends.cudnn.benchmark = True
+loss_function = DiceCELoss(to_onehot_y=True, softmax=True)
+optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)
+
+post_label = AsDiscrete(to_onehot=14)
+post_pred = AsDiscrete(argmax=True, to_onehot=14)
+dice_metric = DiceMetric(include_background=True, reduction="mean", get_not_nans=False)
+
+def validation(epoch_iterable_val):
+    model.eval()
+    epoch_iterator_val = iter(epoch_iterable_val)
+    with torch.no_grad():
+        for _ in range(len(epoch_iterable_val)):
+            with nvtx.annotate("val dataload", color="red"):
+                batch = next(epoch_iterator_val)
+                val_inputs, val_labels = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
+
+            with nvtx.annotate("sliding window", color="green"):
+#                 with torch.cuda.amp.autocast():
+                val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
+
+            with nvtx.annotate("decollate batch", color="blue"):
+                val_labels_list = decollate_batch(val_labels)
+                val_labels_convert = [
+                    post_label(val_label_tensor) for val_label_tensor in val_labels_list
+                ]
+                val_outputs_list = decollate_batch(val_outputs)
+                val_output_convert = [
+                    post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list
+                ]
+
+            with nvtx.annotate("compute metric", color="yellow"):
+                dice_metric(y_pred=val_output_convert, y=val_labels_convert)
+
+            epoch_iterable_val.set_description(
+                "Validate (%d / %d Steps)" % (global_step, 10.0)
+            )
+
+        mean_dice_val = dice_metric.aggregate().item()
+        dice_metric.reset()
+    return mean_dice_val
+
+
+def train(global_step, train_loader, dice_val_best, global_step_best):
+    model.train()
+    epoch_loss = 0
+    step = 0
+    epoch_iterable = tqdm(
+        train_loader, desc="Training (X / X Steps) (loss=X.X)", dynamic_ncols=True
+    )
+    epoch_iterator = iter(epoch_iterable)
+
+    for step in range(len(epoch_iterable)):
+        step += 1
+
+        with nvtx.annotate("dataload", color="red"):
+            batch = next(epoch_iterator)
+            x, y = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
+#         with torch.cuda.amp.autocast():
+        with nvtx.annotate("forward", color="green"):
+            logit_map = model(x)
+            loss = loss_function(logit_map, y)
+
+        with nvtx.annotate("backward", color="blue"):
+            loss.backward()
+            epoch_loss += loss.item()
+
+        with nvtx.annotate("update", color="yellow"):
+            optimizer.step()
+            optimizer.zero_grad()
+
+        epoch_iterable.set_description(
+            "Training (%d / %d Steps) (loss=%2.5f)"
+            % (global_step, max_iterations, loss)
+        )
+        if (
+            global_step % eval_num == 0 and global_step != 0
+        ) or global_step == max_iterations:
+            epoch_iterable_val = tqdm(
+                val_loader, desc="Validate (X / X Steps) (dice=X.X)", dynamic_ncols=True
+            )
+            dice_val = validation(epoch_iterable_val)
+            # FIXME: epoch_loss is a running average at time of validation??
+            epoch_loss /= step
+            epoch_loss_values.append(epoch_loss)
+            metric_values.append(dice_val)
+            if dice_val > dice_val_best:
+                dice_val_best = dice_val
+                global_step_best = global_step
+                torch.save(
+                    model.state_dict(), os.path.join(root_dir, "best_metric_model.pth")
+                )
+                print(
+                    "Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
+                        dice_val_best, dice_val
+                    )
+                )
+            else:
+                print(
+                    "Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
+                        dice_val_best, dice_val
+                    )
+                )
+        global_step += 1
+    return global_step, dice_val_best, global_step_best
+
+# max_iterations = 30000
+# eval_num = 500
+set_determinism(seed=0)
+global_step = 0
+dice_val_best = 0.0
+global_step_best = 0
+epoch_loss_values = []
+metric_values = []
+
+begin = time.time()
+while global_step < max_iterations:
+    with nvtx.annotate("epoch", color="red"):
+        global_step, dice_val_best, global_step_best = train(
+            global_step, train_loader, dice_val_best, global_step_best
+        )
+print(f"Total train time: {time.time() - begin:.2f} seconds")
+
+
+print(
+    f"train completed, best_metric: {dice_val_best:.4f} "
+    f"at iteration: {global_step_best}"
+)
+
+if directory is None:
+    shutil.rmtree(root_dir)
diff --git a/3d_segmentation/swin_unetr_profiling_new.py b/3d_segmentation/swin_unetr_profiling_new.py
new file mode 100644
index 0000000000..85eb114b88
--- /dev/null
+++ b/3d_segmentation/swin_unetr_profiling_new.py
@@ -0,0 +1,331 @@
+# command line: "pip install monai==0.9.1; nsys profile --output /results/orig --force-overwrite true --trace-fork-before-exec true python3 swin_unetr_profiling.py --epochs 5 --val_epochs 5 --batch_size 1 --thread_workers False --num_workers 0"
+
+import os
+import shutil
+import tempfile
+
+import matplotlib.pyplot as plt
+import numpy as np
+from tqdm import tqdm
+import time
+
+from monai.losses import DiceCELoss
+from monai.inferers import sliding_window_inference
+from monai.transforms import (
+    AsDiscrete,
+    AddChanneld,
+    Compose,
+    CropForegroundd,
+    LoadImaged,
+    Orientationd,
+    RandFlipd,
+    RandCropByPosNegLabeld,
+    RandShiftIntensityd,
+    ScaleIntensityRanged,
+    Spacingd,
+    RandRotate90d,
+    EnsureTyped,
+)
+from monai.utils import set_determinism
+
+from monai.config import print_config
+from monai.metrics import DiceMetric
+from monai.networks.nets import SwinUNETR
+
+from monai.data import (
+    ThreadDataLoader,
+    CacheDataset,
+    load_decathlon_datalist,
+    decollate_batch,
+    set_track_meta,
+    get_track_meta,
+)
+
+import nvtx
+from monai.utils.nvtx import Range
+
+import torch
+
+import argparse
+
+print_config()
+
+parser = argparse.ArgumentParser(description='Profiling Swin UNETR.')
+parser.add_argument('--epochs', default=5, type=int)
+parser.add_argument('--val_epochs', default=-1, type=int, help='validation every X epochs; if non-positive value entered, will perform validation only once')
+parser.add_argument('--batch_size', default=1, type=int)
+parser.add_argument('--thread_workers', default=False, type=bool)
+parser.add_argument('--num_workers', default=0, type=int)
+args = parser.parse_args()
+print(args)
+
+assert args.epochs > 0
+assert args.batch_size > 0
+assert args.num_workers >= 0
+
+if not args.thread_workers:
+    args.num_workers = 0
+
+max_iterations = args.epochs * args.batch_size
+eval_num = args.val_epochs * args.batch_size - 1
+
+
+
+directory = os.environ.get("MONAI_DATA_DIRECTORY")
+root_dir = tempfile.mkdtemp() if directory is None else directory
+print(root_dir)
+
+
+num_samples = 4
+os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+train_transforms = Compose(
+    [
+        Range()(LoadImaged(keys=["image", "label"])),
+        Range()(AddChanneld(keys=["image", "label"])),
+        Range()(Orientationd(keys=["image", "label"], axcodes="RAS")),
+        Range()(Spacingd(
+            keys=["image", "label"],
+            pixdim=(1.5, 1.5, 2.0),
+            mode=("bilinear", "nearest"),
+        )),
+        Range()(ScaleIntensityRanged(
+            keys=["image"],
+            a_min=-175,
+            a_max=250,
+            b_min=0.0,
+            b_max=1.0,
+            clip=True,
+        )),
+        Range()(CropForegroundd(keys=["image", "label"], source_key="image")),
+#         Range()(EnsureTyped(keys=["image", "label"])),
+#         Range()(ToDeviced(keys=["image", "label"], device=device)),
+        Range()(EnsureTyped(keys=["image", "label"], device=device, track_meta=False)),
+        Range()(RandCropByPosNegLabeld(
+            keys=["image", "label"],
+            label_key="label",
+            spatial_size=(96, 96, 96),
+            pos=1,
+            neg=1,
+            num_samples=num_samples,
+            image_key="image",
+            image_threshold=0,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[0],
+            prob=0.10,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[1],
+            prob=0.10,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[2],
+            prob=0.10,
+        )),
+        Range()(RandRotate90d(
+            keys=["image", "label"],
+            prob=0.10,
+            max_k=3,
+        )),
+        Range()(RandShiftIntensityd(
+            keys=["image"],
+            offsets=0.10,
+            prob=0.50,
+        )),
+    ]
+)
+val_transforms = Compose(
+    [
+        LoadImaged(keys=["image", "label"]),
+        AddChanneld(keys=["image", "label"]),
+        Orientationd(keys=["image", "label"], axcodes="RAS"),
+        Spacingd(
+            keys=["image", "label"],
+            pixdim=(1.5, 1.5, 2.0),
+            mode=("bilinear", "nearest"),
+        ),
+        ScaleIntensityRanged(
+            keys=["image"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True
+        ),
+        CropForegroundd(keys=["image", "label"], source_key="image"),
+        EnsureTyped(keys=["image", "label"], device=device, track_meta=True),
+    ]
+)
+
+
+torch.cuda.empty_cache()
+data_dir = "data/"
+split_JSON = "dataset_0.json"
+
+datasets = data_dir + split_JSON
+datalist = load_decathlon_datalist(datasets, True, "training")
+val_files = load_decathlon_datalist(datasets, True, "validation")
+train_ds = CacheDataset(
+    data=datalist,
+    transform=train_transforms,
+    cache_num=24,
+    cache_rate=1.0,
+    num_workers=8,
+)
+train_loader = ThreadDataLoader(train_ds, batch_size=args.batch_size, shuffle=True, use_thread_workers=args.thread_workers, num_workers=args.num_workers)
+val_ds = CacheDataset(
+    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4
+)
+val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)
+
+set_track_meta(False)
+
+model = SwinUNETR(
+    img_size=(96, 96, 96),
+    in_channels=1,
+    out_channels=14,
+    feature_size=48,
+    use_checkpoint=True,
+).to(device)
+
+
+weight = torch.load("./model_swinvit.pt")
+model.load_from(weights=weight)
+print("Using pretrained self-supervied Swin UNETR backbone weights !")
+
+
+torch.backends.cudnn.benchmark = True
+loss_function = DiceCELoss(to_onehot_y=True, softmax=True)
+optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)
+scaler = torch.cuda.amp.GradScaler()
+
+post_label = AsDiscrete(to_onehot=14)
+post_pred = AsDiscrete(argmax=True, to_onehot=14)
+dice_metric = DiceMetric(include_background=True, reduction="mean", get_not_nans=False)
+
+def validation(epoch_iterable_val):
+    model.eval()
+    epoch_iterator_val = iter(epoch_iterable_val)
+    with torch.no_grad():
+        for _ in range(len(epoch_iterable_val)):
+            with nvtx.annotate("val dataload", color="red"):
+                batch = next(epoch_iterator_val)
+                val_inputs, val_labels = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
+
+            with nvtx.annotate("sliding window", color="green"):
+                with torch.cuda.amp.autocast():
+                    val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
+
+            with nvtx.annotate("decollate batch", color="blue"):
+                val_labels_list = decollate_batch(val_labels)
+                val_labels_convert = [
+                    post_label(val_label_tensor) for val_label_tensor in val_labels_list
+                ]
+                val_outputs_list = decollate_batch(val_outputs)
+                val_output_convert = [
+                    post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list
+                ]
+
+            with nvtx.annotate("compute metric", color="yellow"):
+                dice_metric(y_pred=val_output_convert, y=val_labels_convert)
+
+            epoch_iterable_val.set_description(
+                "Validate (%d / %d Steps)" % (global_step, 10.0)
+            )
+
+        mean_dice_val = dice_metric.aggregate().item()
+        dice_metric.reset()
+    return mean_dice_val
+
+
+def train(global_step, train_loader, dice_val_best, global_step_best):
+    model.train()
+    epoch_loss = 0
+    step = 0
+    epoch_iterable = tqdm(
+        train_loader, desc="Training (X / X Steps) (loss=X.X)", dynamic_ncols=True
+    )
+    epoch_iterator = iter(epoch_iterable)
+
+    for step in range(len(epoch_iterable)):
+        step += 1
+
+        with nvtx.annotate("dataload", color="red"):
+            batch = next(epoch_iterator)
+            x, y = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
+
+        with nvtx.annotate("forward", color="green"):
+            with torch.cuda.amp.autocast():
+                logit_map = model(x)
+                loss = loss_function(logit_map, y)
+
+        with nvtx.annotate("backward", color="blue"):
+            scaler.scale(loss).backward()
+            epoch_loss += loss.item()
+
+        with nvtx.annotate("update", color="yellow"):
+            scaler.unscale_(optimizer)
+            scaler.step(optimizer)
+            scaler.update()
+            optimizer.zero_grad()
+
+        epoch_iterable.set_description(
+            "Training (%d / %d Steps) (loss=%2.5f)"
+            % (global_step, max_iterations, loss)
+        )
+        if (
+            global_step % eval_num == 0 and global_step != 0
+        ) or global_step == max_iterations:
+            epoch_iterable_val = tqdm(
+                val_loader, desc="Validate (X / X Steps) (dice=X.X)", dynamic_ncols=True
+            )
+            dice_val = validation(epoch_iterable_val)
+            # FIXME: epoch_loss is a running average at time of validation??
+            epoch_loss /= step
+            epoch_loss_values.append(epoch_loss)
+            metric_values.append(dice_val)
+            if dice_val > dice_val_best:
+                dice_val_best = dice_val
+                global_step_best = global_step
+                torch.save(
+                    model.state_dict(), os.path.join(root_dir, "best_metric_model.pth")
+                )
+                print(
+                    "Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
+                        dice_val_best, dice_val
+                    )
+                )
+            else:
+                print(
+                    "Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
+                        dice_val_best, dice_val
+                    )
+                )
+        global_step += 1
+    return global_step, dice_val_best, global_step_best
+
+# max_iterations = 30000
+# eval_num = 500
+set_determinism(seed=0)
+global_step = 0
+dice_val_best = 0.0
+global_step_best = 0
+epoch_loss_values = []
+metric_values = []
+
+begin = time.time()
+while global_step < max_iterations:
+    with nvtx.annotate("epoch", color="red"):
+        global_step, dice_val_best, global_step_best = train(
+            global_step, train_loader, dice_val_best, global_step_best
+        )
+print(f"Total train time: {time.time() - begin:.2f} seconds")
+
+
+print(
+    f"train completed, best_metric: {dice_val_best:.4f} "
+    f"at iteration: {global_step_best}"
+)
+
+if directory is None:
+    shutil.rmtree(root_dir)

From 2fbcbc54eeef3a03d54ed7a942ae76cecd2d7d69 Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Tue, 26 Jul 2022 21:46:20 -0700
Subject: [PATCH 12/17] removed test files

---
 3d_segmentation/swin_unetr_profiling.py     | 338 --------------------
 3d_segmentation/swin_unetr_profiling_new.py | 331 -------------------
 2 files changed, 669 deletions(-)
 delete mode 100644 3d_segmentation/swin_unetr_profiling.py
 delete mode 100644 3d_segmentation/swin_unetr_profiling_new.py

diff --git a/3d_segmentation/swin_unetr_profiling.py b/3d_segmentation/swin_unetr_profiling.py
deleted file mode 100644
index 75acf6033d..0000000000
--- a/3d_segmentation/swin_unetr_profiling.py
+++ /dev/null
@@ -1,338 +0,0 @@
-# command line: "pip install monai==0.9.1; nsys profile --output /results/orig --force-overwrite true --trace-fork-before-exec true python3 swin_unetr_profiling.py --epochs 5 --val_epochs 5 --batch_size 1 --thread_workers False --num_workers 0"
-
-import os
-import shutil
-import tempfile
-
-import matplotlib.pyplot as plt
-import numpy as np
-from tqdm import tqdm
-import time
-
-from monai.losses import DiceCELoss
-from monai.inferers import sliding_window_inference
-from monai.transforms import (
-    AsDiscrete,
-    AddChanneld,
-    Compose,
-    CropForegroundd,
-    LoadImaged,
-    Orientationd,
-    RandFlipd,
-    RandCropByPosNegLabeld,
-    RandShiftIntensityd,
-    ScaleIntensityRanged,
-    Spacingd,
-    RandRotate90d,
-    ToTensord,
-    EnsureTyped,
-    ToDeviced
-)
-from monai.utils import set_determinism
-
-from monai.config import print_config
-from monai.metrics import DiceMetric
-from monai.networks.nets import SwinUNETR
-
-from monai.data import (
-#    ThreadDataLoader,
-    DataLoader,
-    CacheDataset,
-    load_decathlon_datalist,
-    decollate_batch,
-)
-
-import nvtx
-from monai.utils.nvtx import Range
-
-import torch
-
-import argparse
-
-print_config()
-
-parser = argparse.ArgumentParser(description='Profiling Swin UNETR.')
-parser.add_argument('--epochs', default=5, type=int)
-parser.add_argument('--val_epochs', default=-1, type=int, help='validation every X epochs; if non-positive value entered, will perform validation only once')
-parser.add_argument('--batch_size', default=1, type=int)
-parser.add_argument('--thread_workers', default=False, type=bool)
-parser.add_argument('--num_workers', default=0, type=int)
-args = parser.parse_args()
-print(args)
-
-assert args.epochs > 0
-assert args.batch_size > 0
-assert args.num_workers >= 0
-
-if not args.thread_workers:
-    args.num_workers = 0
-
-max_iterations = args.epochs * args.batch_size
-eval_num = args.val_epochs * args.batch_size - 1
-
-
-
-directory = os.environ.get("MONAI_DATA_DIRECTORY")
-root_dir = tempfile.mkdtemp() if directory is None else directory
-print(root_dir)
-
-
-num_samples = 4
-os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-
-train_transforms = Compose(
-    [
-        Range()(LoadImaged(keys=["image", "label"])),
-        Range()(AddChanneld(keys=["image", "label"])),
-        Range()(Orientationd(keys=["image", "label"], axcodes="RAS")),
-        Range()(Spacingd(
-            keys=["image", "label"],
-            pixdim=(1.5, 1.5, 2.0),
-            mode=("bilinear", "nearest"),
-        )),
-        Range()(ScaleIntensityRanged(
-            keys=["image"],
-            a_min=-175,
-            a_max=250,
-            b_min=0.0,
-            b_max=1.0,
-            clip=True,
-        )),
-        Range()(CropForegroundd(keys=["image", "label"], source_key="image")),
-#         Range()(EnsureTyped(keys=["image", "label"])),
-#         Range()(ToDeviced(keys=["image", "label"], device=device)),
-        Range()(RandCropByPosNegLabeld(
-            keys=["image", "label"],
-            label_key="label",
-            spatial_size=(96, 96, 96),
-            pos=1,
-            neg=1,
-            num_samples=num_samples,
-            image_key="image",
-            image_threshold=0,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[0],
-            prob=0.10,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[1],
-            prob=0.10,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[2],
-            prob=0.10,
-        )),
-        Range()(RandRotate90d(
-            keys=["image", "label"],
-            prob=0.10,
-            max_k=3,
-        )),
-        Range()(RandShiftIntensityd(
-            keys=["image"],
-            offsets=0.10,
-            prob=0.50,
-        )),
-#        Range()(ToTensord(keys=["image", "label"])),
-    ]
-)
-val_transforms = Compose(
-    [
-        LoadImaged(keys=["image", "label"]),
-        AddChanneld(keys=["image", "label"]),
-        Orientationd(keys=["image", "label"], axcodes="RAS"),
-        Spacingd(
-            keys=["image", "label"],
-            pixdim=(1.5, 1.5, 2.0),
-            mode=("bilinear", "nearest"),
-        ),
-        ScaleIntensityRanged(
-            keys=["image"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True
-        ),
-        CropForegroundd(keys=["image", "label"], source_key="image"),
-#        ToTensord(keys=["image", "label"]),
-#        EnsureTyped(keys=["image", "label"]),
-#        ToDeviced(keys=["image", "label"], device=device),
-    ]
-)
-
-
-torch.cuda.empty_cache()
-data_dir = "data/"
-split_JSON = "dataset_0.json"
-
-datasets = data_dir + split_JSON
-datalist = load_decathlon_datalist(datasets, True, "training")
-val_files = load_decathlon_datalist(datasets, True, "validation")
-
-# TODO: try thread_workers
-train_ds = CacheDataset(
-    data=datalist,
-    transform=train_transforms,
-    cache_num=24,
-    cache_rate=1.0,
-    num_workers=8,
-)
-# train_loader = ThreadDataLoader(train_ds, batch_size=1, shuffle=True, num_workers=0)
-# train_loader = ThreadDataLoader(train_ds, batch_size=args.batch_size, shuffle=True, use_thread_workers=args.thread_workers, num_workers=args.num_workers)
-train_loader = DataLoader(
-    train_ds, batch_size=args.batch_size, shuffle=True, num_workers=8, pin_memory=True
-)
-
-val_ds = CacheDataset(
-    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4
-)
-# val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)
-val_loader = DataLoader(
-    val_ds, batch_size=args.batch_size, shuffle=False, num_workers=4, pin_memory=True
-)
-
-model = SwinUNETR(
-    img_size=(96, 96, 96),
-    in_channels=1,
-    out_channels=14,
-    feature_size=48,
-    use_checkpoint=True,
-).to(device)
-
-
-weight = torch.load("./model_swinvit.pt")
-model.load_from(weights=weight)
-print("Using pretrained self-supervied Swin UNETR backbone weights !")
-
-
-torch.backends.cudnn.benchmark = True
-loss_function = DiceCELoss(to_onehot_y=True, softmax=True)
-optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)
-
-post_label = AsDiscrete(to_onehot=14)
-post_pred = AsDiscrete(argmax=True, to_onehot=14)
-dice_metric = DiceMetric(include_background=True, reduction="mean", get_not_nans=False)
-
-def validation(epoch_iterable_val):
-    model.eval()
-    epoch_iterator_val = iter(epoch_iterable_val)
-    with torch.no_grad():
-        for _ in range(len(epoch_iterable_val)):
-            with nvtx.annotate("val dataload", color="red"):
-                batch = next(epoch_iterator_val)
-                val_inputs, val_labels = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
-
-            with nvtx.annotate("sliding window", color="green"):
-#                 with torch.cuda.amp.autocast():
-                val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
-
-            with nvtx.annotate("decollate batch", color="blue"):
-                val_labels_list = decollate_batch(val_labels)
-                val_labels_convert = [
-                    post_label(val_label_tensor) for val_label_tensor in val_labels_list
-                ]
-                val_outputs_list = decollate_batch(val_outputs)
-                val_output_convert = [
-                    post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list
-                ]
-
-            with nvtx.annotate("compute metric", color="yellow"):
-                dice_metric(y_pred=val_output_convert, y=val_labels_convert)
-
-            epoch_iterable_val.set_description(
-                "Validate (%d / %d Steps)" % (global_step, 10.0)
-            )
-
-        mean_dice_val = dice_metric.aggregate().item()
-        dice_metric.reset()
-    return mean_dice_val
-
-
-def train(global_step, train_loader, dice_val_best, global_step_best):
-    model.train()
-    epoch_loss = 0
-    step = 0
-    epoch_iterable = tqdm(
-        train_loader, desc="Training (X / X Steps) (loss=X.X)", dynamic_ncols=True
-    )
-    epoch_iterator = iter(epoch_iterable)
-
-    for step in range(len(epoch_iterable)):
-        step += 1
-
-        with nvtx.annotate("dataload", color="red"):
-            batch = next(epoch_iterator)
-            x, y = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
-#         with torch.cuda.amp.autocast():
-        with nvtx.annotate("forward", color="green"):
-            logit_map = model(x)
-            loss = loss_function(logit_map, y)
-
-        with nvtx.annotate("backward", color="blue"):
-            loss.backward()
-            epoch_loss += loss.item()
-
-        with nvtx.annotate("update", color="yellow"):
-            optimizer.step()
-            optimizer.zero_grad()
-
-        epoch_iterable.set_description(
-            "Training (%d / %d Steps) (loss=%2.5f)"
-            % (global_step, max_iterations, loss)
-        )
-        if (
-            global_step % eval_num == 0 and global_step != 0
-        ) or global_step == max_iterations:
-            epoch_iterable_val = tqdm(
-                val_loader, desc="Validate (X / X Steps) (dice=X.X)", dynamic_ncols=True
-            )
-            dice_val = validation(epoch_iterable_val)
-            # FIXME: epoch_loss is a running average at time of validation??
-            epoch_loss /= step
-            epoch_loss_values.append(epoch_loss)
-            metric_values.append(dice_val)
-            if dice_val > dice_val_best:
-                dice_val_best = dice_val
-                global_step_best = global_step
-                torch.save(
-                    model.state_dict(), os.path.join(root_dir, "best_metric_model.pth")
-                )
-                print(
-                    "Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
-                        dice_val_best, dice_val
-                    )
-                )
-            else:
-                print(
-                    "Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
-                        dice_val_best, dice_val
-                    )
-                )
-        global_step += 1
-    return global_step, dice_val_best, global_step_best
-
-# max_iterations = 30000
-# eval_num = 500
-set_determinism(seed=0)
-global_step = 0
-dice_val_best = 0.0
-global_step_best = 0
-epoch_loss_values = []
-metric_values = []
-
-begin = time.time()
-while global_step < max_iterations:
-    with nvtx.annotate("epoch", color="red"):
-        global_step, dice_val_best, global_step_best = train(
-            global_step, train_loader, dice_val_best, global_step_best
-        )
-print(f"Total train time: {time.time() - begin:.2f} seconds")
-
-
-print(
-    f"train completed, best_metric: {dice_val_best:.4f} "
-    f"at iteration: {global_step_best}"
-)
-
-if directory is None:
-    shutil.rmtree(root_dir)
diff --git a/3d_segmentation/swin_unetr_profiling_new.py b/3d_segmentation/swin_unetr_profiling_new.py
deleted file mode 100644
index 85eb114b88..0000000000
--- a/3d_segmentation/swin_unetr_profiling_new.py
+++ /dev/null
@@ -1,331 +0,0 @@
-# command line: "pip install monai==0.9.1; nsys profile --output /results/orig --force-overwrite true --trace-fork-before-exec true python3 swin_unetr_profiling.py --epochs 5 --val_epochs 5 --batch_size 1 --thread_workers False --num_workers 0"
-
-import os
-import shutil
-import tempfile
-
-import matplotlib.pyplot as plt
-import numpy as np
-from tqdm import tqdm
-import time
-
-from monai.losses import DiceCELoss
-from monai.inferers import sliding_window_inference
-from monai.transforms import (
-    AsDiscrete,
-    AddChanneld,
-    Compose,
-    CropForegroundd,
-    LoadImaged,
-    Orientationd,
-    RandFlipd,
-    RandCropByPosNegLabeld,
-    RandShiftIntensityd,
-    ScaleIntensityRanged,
-    Spacingd,
-    RandRotate90d,
-    EnsureTyped,
-)
-from monai.utils import set_determinism
-
-from monai.config import print_config
-from monai.metrics import DiceMetric
-from monai.networks.nets import SwinUNETR
-
-from monai.data import (
-    ThreadDataLoader,
-    CacheDataset,
-    load_decathlon_datalist,
-    decollate_batch,
-    set_track_meta,
-    get_track_meta,
-)
-
-import nvtx
-from monai.utils.nvtx import Range
-
-import torch
-
-import argparse
-
-print_config()
-
-parser = argparse.ArgumentParser(description='Profiling Swin UNETR.')
-parser.add_argument('--epochs', default=5, type=int)
-parser.add_argument('--val_epochs', default=-1, type=int, help='validation every X epochs; if non-positive value entered, will perform validation only once')
-parser.add_argument('--batch_size', default=1, type=int)
-parser.add_argument('--thread_workers', default=False, type=bool)
-parser.add_argument('--num_workers', default=0, type=int)
-args = parser.parse_args()
-print(args)
-
-assert args.epochs > 0
-assert args.batch_size > 0
-assert args.num_workers >= 0
-
-if not args.thread_workers:
-    args.num_workers = 0
-
-max_iterations = args.epochs * args.batch_size
-eval_num = args.val_epochs * args.batch_size - 1
-
-
-
-directory = os.environ.get("MONAI_DATA_DIRECTORY")
-root_dir = tempfile.mkdtemp() if directory is None else directory
-print(root_dir)
-
-
-num_samples = 4
-os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-
-train_transforms = Compose(
-    [
-        Range()(LoadImaged(keys=["image", "label"])),
-        Range()(AddChanneld(keys=["image", "label"])),
-        Range()(Orientationd(keys=["image", "label"], axcodes="RAS")),
-        Range()(Spacingd(
-            keys=["image", "label"],
-            pixdim=(1.5, 1.5, 2.0),
-            mode=("bilinear", "nearest"),
-        )),
-        Range()(ScaleIntensityRanged(
-            keys=["image"],
-            a_min=-175,
-            a_max=250,
-            b_min=0.0,
-            b_max=1.0,
-            clip=True,
-        )),
-        Range()(CropForegroundd(keys=["image", "label"], source_key="image")),
-#         Range()(EnsureTyped(keys=["image", "label"])),
-#         Range()(ToDeviced(keys=["image", "label"], device=device)),
-        Range()(EnsureTyped(keys=["image", "label"], device=device, track_meta=False)),
-        Range()(RandCropByPosNegLabeld(
-            keys=["image", "label"],
-            label_key="label",
-            spatial_size=(96, 96, 96),
-            pos=1,
-            neg=1,
-            num_samples=num_samples,
-            image_key="image",
-            image_threshold=0,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[0],
-            prob=0.10,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[1],
-            prob=0.10,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[2],
-            prob=0.10,
-        )),
-        Range()(RandRotate90d(
-            keys=["image", "label"],
-            prob=0.10,
-            max_k=3,
-        )),
-        Range()(RandShiftIntensityd(
-            keys=["image"],
-            offsets=0.10,
-            prob=0.50,
-        )),
-    ]
-)
-val_transforms = Compose(
-    [
-        LoadImaged(keys=["image", "label"]),
-        AddChanneld(keys=["image", "label"]),
-        Orientationd(keys=["image", "label"], axcodes="RAS"),
-        Spacingd(
-            keys=["image", "label"],
-            pixdim=(1.5, 1.5, 2.0),
-            mode=("bilinear", "nearest"),
-        ),
-        ScaleIntensityRanged(
-            keys=["image"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True
-        ),
-        CropForegroundd(keys=["image", "label"], source_key="image"),
-        EnsureTyped(keys=["image", "label"], device=device, track_meta=True),
-    ]
-)
-
-
-torch.cuda.empty_cache()
-data_dir = "data/"
-split_JSON = "dataset_0.json"
-
-datasets = data_dir + split_JSON
-datalist = load_decathlon_datalist(datasets, True, "training")
-val_files = load_decathlon_datalist(datasets, True, "validation")
-train_ds = CacheDataset(
-    data=datalist,
-    transform=train_transforms,
-    cache_num=24,
-    cache_rate=1.0,
-    num_workers=8,
-)
-train_loader = ThreadDataLoader(train_ds, batch_size=args.batch_size, shuffle=True, use_thread_workers=args.thread_workers, num_workers=args.num_workers)
-val_ds = CacheDataset(
-    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4
-)
-val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)
-
-set_track_meta(False)
-
-model = SwinUNETR(
-    img_size=(96, 96, 96),
-    in_channels=1,
-    out_channels=14,
-    feature_size=48,
-    use_checkpoint=True,
-).to(device)
-
-
-weight = torch.load("./model_swinvit.pt")
-model.load_from(weights=weight)
-print("Using pretrained self-supervied Swin UNETR backbone weights !")
-
-
-torch.backends.cudnn.benchmark = True
-loss_function = DiceCELoss(to_onehot_y=True, softmax=True)
-optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)
-scaler = torch.cuda.amp.GradScaler()
-
-post_label = AsDiscrete(to_onehot=14)
-post_pred = AsDiscrete(argmax=True, to_onehot=14)
-dice_metric = DiceMetric(include_background=True, reduction="mean", get_not_nans=False)
-
-def validation(epoch_iterable_val):
-    model.eval()
-    epoch_iterator_val = iter(epoch_iterable_val)
-    with torch.no_grad():
-        for _ in range(len(epoch_iterable_val)):
-            with nvtx.annotate("val dataload", color="red"):
-                batch = next(epoch_iterator_val)
-                val_inputs, val_labels = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
-
-            with nvtx.annotate("sliding window", color="green"):
-                with torch.cuda.amp.autocast():
-                    val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
-
-            with nvtx.annotate("decollate batch", color="blue"):
-                val_labels_list = decollate_batch(val_labels)
-                val_labels_convert = [
-                    post_label(val_label_tensor) for val_label_tensor in val_labels_list
-                ]
-                val_outputs_list = decollate_batch(val_outputs)
-                val_output_convert = [
-                    post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list
-                ]
-
-            with nvtx.annotate("compute metric", color="yellow"):
-                dice_metric(y_pred=val_output_convert, y=val_labels_convert)
-
-            epoch_iterable_val.set_description(
-                "Validate (%d / %d Steps)" % (global_step, 10.0)
-            )
-
-        mean_dice_val = dice_metric.aggregate().item()
-        dice_metric.reset()
-    return mean_dice_val
-
-
-def train(global_step, train_loader, dice_val_best, global_step_best):
-    model.train()
-    epoch_loss = 0
-    step = 0
-    epoch_iterable = tqdm(
-        train_loader, desc="Training (X / X Steps) (loss=X.X)", dynamic_ncols=True
-    )
-    epoch_iterator = iter(epoch_iterable)
-
-    for step in range(len(epoch_iterable)):
-        step += 1
-
-        with nvtx.annotate("dataload", color="red"):
-            batch = next(epoch_iterator)
-            x, y = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
-
-        with nvtx.annotate("forward", color="green"):
-            with torch.cuda.amp.autocast():
-                logit_map = model(x)
-                loss = loss_function(logit_map, y)
-
-        with nvtx.annotate("backward", color="blue"):
-            scaler.scale(loss).backward()
-            epoch_loss += loss.item()
-
-        with nvtx.annotate("update", color="yellow"):
-            scaler.unscale_(optimizer)
-            scaler.step(optimizer)
-            scaler.update()
-            optimizer.zero_grad()
-
-        epoch_iterable.set_description(
-            "Training (%d / %d Steps) (loss=%2.5f)"
-            % (global_step, max_iterations, loss)
-        )
-        if (
-            global_step % eval_num == 0 and global_step != 0
-        ) or global_step == max_iterations:
-            epoch_iterable_val = tqdm(
-                val_loader, desc="Validate (X / X Steps) (dice=X.X)", dynamic_ncols=True
-            )
-            dice_val = validation(epoch_iterable_val)
-            # FIXME: epoch_loss is a running average at time of validation??
-            epoch_loss /= step
-            epoch_loss_values.append(epoch_loss)
-            metric_values.append(dice_val)
-            if dice_val > dice_val_best:
-                dice_val_best = dice_val
-                global_step_best = global_step
-                torch.save(
-                    model.state_dict(), os.path.join(root_dir, "best_metric_model.pth")
-                )
-                print(
-                    "Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
-                        dice_val_best, dice_val
-                    )
-                )
-            else:
-                print(
-                    "Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
-                        dice_val_best, dice_val
-                    )
-                )
-        global_step += 1
-    return global_step, dice_val_best, global_step_best
-
-# max_iterations = 30000
-# eval_num = 500
-set_determinism(seed=0)
-global_step = 0
-dice_val_best = 0.0
-global_step_best = 0
-epoch_loss_values = []
-metric_values = []
-
-begin = time.time()
-while global_step < max_iterations:
-    with nvtx.annotate("epoch", color="red"):
-        global_step, dice_val_best, global_step_best = train(
-            global_step, train_loader, dice_val_best, global_step_best
-        )
-print(f"Total train time: {time.time() - begin:.2f} seconds")
-
-
-print(
-    f"train completed, best_metric: {dice_val_best:.4f} "
-    f"at iteration: {global_step_best}"
-)
-
-if directory is None:
-    shutil.rmtree(root_dir)

From bf7e9f9c5cb0a6904070ca0799e02e3fda38d567 Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Wed, 27 Jul 2022 14:13:44 -0700
Subject: [PATCH 13/17] updated json download

---
 .../swin_unetr_btcv_segmentation_3d.ipynb            | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
index 9e1567095d..9eb60d0205 100644
--- a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
+++ b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
@@ -293,7 +293,17 @@
     "\n",
     "3. Make a JSON file to define train/val split and other relevant parameters. Place the JSON file at `./data/dataset_0.json`.\n",
     "\n",
-    "    You can download an example of the JSON file [here](https://drive.google.com/file/d/1t4fIQQkONv7ArTSZe4Nucwkk1KfdUDvW/view?usp=sharing). If you would like to use this directly, please move it into the `./data` folder."
+    "    You can download an example of the JSON file [here](https://drive.google.com/file/d/1t4fIQQkONv7ArTSZe4Nucwkk1KfdUDvW/view?usp=sharing), or, equivalently, use the following `wget` command. If you would like to use this directly, please move it into the `./data` folder."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# uncomment this command to download the JSON file directly\n",
+    "# wget -O data/dataset_0.json 'https://drive.google.com/uc?export=download&id=1t4fIQQkONv7ArTSZe4Nucwkk1KfdUDvW'"
    ]
   },
   {

From e4d7004c7aa15de431cb0d19500698bd1b30bae6 Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Wed, 27 Jul 2022 14:26:23 -0700
Subject: [PATCH 14/17] updated explanation

---
 3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
index 9eb60d0205..1114184bcb 100644
--- a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
+++ b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
@@ -186,11 +186,17 @@
     "## Setup transforms for training and validation\n",
     "To save on GPU memory utilization, the num_samples can be reduced to 2. \n",
     "\n",
-    "A note on design:\n",
+    "A note on design related to MetaTensors:\n",
+    "\n",
+    "- Summary: using `EnsureTyped(..., track_meta=False)` and `set_track_meta(False)` (later) speeds up training significantly.\n",
     "\n",
     "- We are moving towards the use of MONAI's MetaTensor in place of numpy arrays or PyTorch tensors. MetaTensors have the benefit of carrying the metadata directly with the tensor, but in some use cases (like here with training, where training data are only used for computing loss and metadata is not useful), we can safely disregard the metadata to improve speed.\n",
     "\n",
-    "- Hence, you will see `EnsureTyped` being used before the first random transform in the training transform chain, which caches the result of deterministic transforms on GPU, with `track_meta = False`. On the other hand, in the following demos we will display example validation images, which uses metadata, so we use `EnsureTyped` with `track_meta = True`. Since there are no random transforms during validation, tracking metadata for validation images causes virtually no slowdown (~0.5%)."
+    "- Hence, you will see `EnsureTyped` being used before the first random transform in the training transform chain, which caches the result of deterministic transforms on GPU as Tensors (rather than MetaTensors), with `track_meta = False`. \n",
+    "\n",
+    "- On the other hand, in the following demos we will display example validation images, which uses metadata, so we use `EnsureTyped` with `track_meta = True`. Since there are no random transforms during validation, tracking metadata for validation images causes virtually no slowdown (~0.5%).\n",
+    "\n",
+    "- In the next section, you will see `set_track_meta(False)`. This is a global API introduced in MONAI 0.9.1, and it makes sure that random transforms will also be performed using Tensors rather than MetaTensors. Used together with `track_meta=False` in `EnsureTyped`, it results in all transforms being performed on Tensors, which we have found to speed up training."
    ]
   },
   {

From b282b296a85c4157facc04b6de36a11dcd87b62d Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Wed, 27 Jul 2022 14:28:15 -0700
Subject: [PATCH 15/17] final profiling code backup

---
 3d_segmentation/swin_unetr_profiling.py     | 335 +++++++++++++++++++
 3d_segmentation/swin_unetr_profiling_new.py | 338 ++++++++++++++++++++
 2 files changed, 673 insertions(+)
 create mode 100644 3d_segmentation/swin_unetr_profiling.py
 create mode 100644 3d_segmentation/swin_unetr_profiling_new.py

diff --git a/3d_segmentation/swin_unetr_profiling.py b/3d_segmentation/swin_unetr_profiling.py
new file mode 100644
index 0000000000..0bb4bb22a6
--- /dev/null
+++ b/3d_segmentation/swin_unetr_profiling.py
@@ -0,0 +1,335 @@
+# command line: "pip install monai==0.9.1; nsys profile --output /results/orig --force-overwrite true --trace-fork-before-exec true python3 swin_unetr_profiling.py --epochs 5 --val_epochs 5 --batch_size 1 --thread_workers False --num_workers 0"
+
+import os
+import shutil
+import tempfile
+
+import matplotlib.pyplot as plt
+import numpy as np
+from tqdm import tqdm
+import time
+
+from monai.losses import DiceCELoss
+from monai.inferers import sliding_window_inference
+from monai.transforms import (
+    AsDiscrete,
+    AddChanneld,
+    Compose,
+    CropForegroundd,
+    LoadImaged,
+    Orientationd,
+    RandFlipd,
+    RandCropByPosNegLabeld,
+    RandShiftIntensityd,
+    ScaleIntensityRanged,
+    Spacingd,
+    RandRotate90d,
+    ToTensord,
+    EnsureTyped,
+    ToDeviced
+)
+from monai.utils import set_determinism
+
+from monai.config import print_config
+from monai.metrics import DiceMetric
+from monai.networks.nets import SwinUNETR
+
+from monai.data import (
+#    ThreadDataLoader,
+    DataLoader,
+    Dataset,
+    load_decathlon_datalist,
+    decollate_batch,
+)
+
+import nvtx
+from monai.utils.nvtx import Range
+
+import torch
+
+import argparse
+
+print_config()
+
+parser = argparse.ArgumentParser(description='Profiling Swin UNETR.')
+parser.add_argument('--epochs', default=5, type=int)
+parser.add_argument('--val_epochs', default=-1, type=int, help='validation every X epochs; if non-positive value entered, will perform validation only once')
+parser.add_argument('--batch_size', default=1, type=int)
+parser.add_argument('--thread_workers', default=False, type=bool)
+parser.add_argument('--num_workers', default=0, type=int)
+args = parser.parse_args()
+print(args)
+
+assert args.epochs >= 0
+assert args.batch_size > 0
+assert args.num_workers >= 0
+
+if not args.thread_workers:
+    args.num_workers = 0
+
+# +
+max_iterations = args.epochs * args.batch_size
+max_iterations = max(max_iterations, 24)
+
+eval_num = args.val_epochs * args.batch_size - 1
+if eval_num <= 0:
+    eval_num = max_iterations - 1
+# -
+
+
+
+directory = os.environ.get("MONAI_DATA_DIRECTORY")
+root_dir = tempfile.mkdtemp() if directory is None else directory
+print(root_dir)
+
+
+num_samples = 4
+os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+train_transforms = Compose(
+    [
+        Range()(LoadImaged(keys=["image", "label"])),
+        Range()(AddChanneld(keys=["image", "label"])),
+        Range()(Orientationd(keys=["image", "label"], axcodes="RAS")),
+        Range()(Spacingd(
+            keys=["image", "label"],
+            pixdim=(1.5, 1.5, 2.0),
+            mode=("bilinear", "nearest"),
+        )),
+        Range()(ScaleIntensityRanged(
+            keys=["image"],
+            a_min=-175,
+            a_max=250,
+            b_min=0.0,
+            b_max=1.0,
+            clip=True,
+        )),
+        Range()(CropForegroundd(keys=["image", "label"], source_key="image")),
+#         Range()(EnsureTyped(keys=["image", "label"])),
+#         Range()(ToDeviced(keys=["image", "label"], device=device)),
+        Range()(RandCropByPosNegLabeld(
+            keys=["image", "label"],
+            label_key="label",
+            spatial_size=(96, 96, 96),
+            pos=1,
+            neg=1,
+            num_samples=num_samples,
+            image_key="image",
+            image_threshold=0,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[0],
+            prob=0.10,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[1],
+            prob=0.10,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[2],
+            prob=0.10,
+        )),
+        Range()(RandRotate90d(
+            keys=["image", "label"],
+            prob=0.10,
+            max_k=3,
+        )),
+        Range()(RandShiftIntensityd(
+            keys=["image"],
+            offsets=0.10,
+            prob=0.50,
+        )),
+#        Range()(ToTensord(keys=["image", "label"])),
+    ]
+)
+val_transforms = Compose(
+    [
+        LoadImaged(keys=["image", "label"]),
+        AddChanneld(keys=["image", "label"]),
+        Orientationd(keys=["image", "label"], axcodes="RAS"),
+        Spacingd(
+            keys=["image", "label"],
+            pixdim=(1.5, 1.5, 2.0),
+            mode=("bilinear", "nearest"),
+        ),
+        ScaleIntensityRanged(
+            keys=["image"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True
+        ),
+        CropForegroundd(keys=["image", "label"], source_key="image"),
+#        ToTensord(keys=["image", "label"]),
+#        EnsureTyped(keys=["image", "label"]),
+#        ToDeviced(keys=["image", "label"], device=device),
+    ]
+)
+
+
+torch.cuda.empty_cache()
+data_dir = "data/"
+split_JSON = "dataset_0.json"
+
+datasets = data_dir + split_JSON
+datalist = load_decathlon_datalist(datasets, True, "training")
+val_files = load_decathlon_datalist(datasets, True, "validation")
+
+# TODO: try thread_workers
+train_ds = Dataset(data=datalist, transform=train_transforms)
+# train_ds = CacheDataset(data=datalist, transform=train_transforms, cache_num=24, cache_rate=1.0, num_workers=8,)
+# train_loader = ThreadDataLoader(train_ds, batch_size=1, shuffle=True, num_workers=0)
+# train_loader = ThreadDataLoader(train_ds, batch_size=args.batch_size, shuffle=True, use_thread_workers=args.thread_workers, num_workers=args.num_workers)
+train_loader = DataLoader(train_ds, batch_size=args.batch_size, shuffle=True, num_workers=8, pin_memory=True)
+
+# val_ds = CacheDataset(data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4)
+val_ds = Dataset(data=val_files, transform=val_transforms)
+# val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)
+val_loader = DataLoader(val_ds, batch_size=args.batch_size, shuffle=False, num_workers=4, pin_memory=True)
+
+model = SwinUNETR(
+    img_size=(96, 96, 96),
+    in_channels=1,
+    out_channels=14,
+    feature_size=48,
+    use_checkpoint=True,
+).to(device)
+
+
+weight = torch.load("./model_swinvit.pt")
+model.load_from(weights=weight)
+print("Using pretrained self-supervied Swin UNETR backbone weights !")
+
+
+torch.backends.cudnn.benchmark = True
+loss_function = DiceCELoss(to_onehot_y=True, softmax=True)
+optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)
+
+post_label = AsDiscrete(to_onehot=14)
+post_pred = AsDiscrete(argmax=True, to_onehot=14)
+dice_metric = DiceMetric(include_background=True, reduction="mean", get_not_nans=False)
+
+def validation(epoch_iterable_val):
+    model.eval()
+    epoch_iterator_val = iter(epoch_iterable_val)
+    with torch.no_grad():
+        for _ in range(len(epoch_iterable_val)):
+            with nvtx.annotate("val dataload", color="red"):
+                batch = next(epoch_iterator_val)
+                val_inputs, val_labels = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
+
+            with nvtx.annotate("sliding window", color="green"):
+#                 with torch.cuda.amp.autocast():
+                val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
+
+            with nvtx.annotate("decollate batch", color="blue"):
+                val_labels_list = decollate_batch(val_labels)
+                val_labels_convert = [
+                    post_label(val_label_tensor) for val_label_tensor in val_labels_list
+                ]
+                val_outputs_list = decollate_batch(val_outputs)
+                val_output_convert = [
+                    post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list
+                ]
+
+            with nvtx.annotate("compute metric", color="yellow"):
+                dice_metric(y_pred=val_output_convert, y=val_labels_convert)
+
+            epoch_iterable_val.set_description(
+                "Validate (%d / %d Steps)" % (global_step, 10.0)
+            )
+
+        mean_dice_val = dice_metric.aggregate().item()
+        dice_metric.reset()
+    return mean_dice_val
+
+
+def train(global_step, train_loader, dice_val_best, global_step_best):
+    model.train()
+    epoch_loss = 0
+    step = 0
+    epoch_iterable = tqdm(
+        train_loader, desc="Training (X / X Steps) (loss=X.X)", dynamic_ncols=True
+    )
+    epoch_iterator = iter(epoch_iterable)
+
+    for step in range(len(epoch_iterable)):
+        step += 1
+
+        with nvtx.annotate("dataload", color="red"):
+            batch = next(epoch_iterator)
+            x, y = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
+#         with torch.cuda.amp.autocast():
+        with nvtx.annotate("forward", color="green"):
+            logit_map = model(x)
+            loss = loss_function(logit_map, y)
+
+        with nvtx.annotate("backward", color="blue"):
+            loss.backward()
+            epoch_loss += loss.item()
+
+        with nvtx.annotate("update", color="yellow"):
+            optimizer.step()
+            optimizer.zero_grad()
+
+        epoch_iterable.set_description(
+            "Training (%d / %d Steps) (loss=%2.5f)"
+            % (global_step, max_iterations, loss)
+        )
+        if (
+            global_step % eval_num == 0 and global_step != 0
+        ) or global_step == max_iterations:
+            epoch_iterable_val = tqdm(
+                val_loader, desc="Validate (X / X Steps) (dice=X.X)", dynamic_ncols=True
+            )
+            dice_val = validation(epoch_iterable_val)
+            # FIXME: epoch_loss is a running average at time of validation??
+            with nvtx.annotate("post-validation processing", color="blue"):
+                epoch_loss /= step
+                epoch_loss_values.append(epoch_loss)
+                metric_values.append(dice_val)
+                if dice_val > dice_val_best:
+                    dice_val_best = dice_val
+                    global_step_best = global_step
+                    torch.save(
+                        model.state_dict(), os.path.join(root_dir, "best_metric_model.pth")
+                    )
+                    print(
+                        "Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
+                            dice_val_best, dice_val
+                        )
+                    )
+                else:
+                    print(
+                        "Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
+                            dice_val_best, dice_val
+                        )
+                    )
+        global_step += 1
+    return global_step, dice_val_best, global_step_best
+
+# max_iterations = 30000
+# eval_num = 500
+set_determinism(seed=0)
+global_step = 0
+dice_val_best = 0.0
+global_step_best = 0
+epoch_loss_values = []
+metric_values = []
+
+begin = time.time()
+while global_step < max_iterations:
+    with nvtx.annotate("epoch", color="red"):
+        global_step, dice_val_best, global_step_best = train(
+            global_step, train_loader, dice_val_best, global_step_best
+        )
+print(f"Total train time: {time.time() - begin:.2f} seconds")
+
+
+print(
+    f"train completed, best_metric: {dice_val_best:.4f} "
+    f"at iteration: {global_step_best}"
+)
+
+if directory is None:
+    shutil.rmtree(root_dir)
diff --git a/3d_segmentation/swin_unetr_profiling_new.py b/3d_segmentation/swin_unetr_profiling_new.py
new file mode 100644
index 0000000000..5186f9640d
--- /dev/null
+++ b/3d_segmentation/swin_unetr_profiling_new.py
@@ -0,0 +1,338 @@
+# command line: "pip install monai==0.9.1; nsys profile --output /results/orig --force-overwrite true --trace-fork-before-exec true python3 swin_unetr_profiling.py --epochs 0 --val_epochs 0 --batch_size 1 --thread_workers False --num_workers 0"
+
+import os
+import shutil
+import tempfile
+
+import matplotlib.pyplot as plt
+import numpy as np
+from tqdm import tqdm
+import time
+
+from monai.losses import DiceCELoss
+from monai.inferers import sliding_window_inference
+from monai.transforms import (
+    AsDiscrete,
+    AddChanneld,
+    Compose,
+    CropForegroundd,
+    LoadImaged,
+    Orientationd,
+    RandFlipd,
+    RandCropByPosNegLabeld,
+    RandShiftIntensityd,
+    ScaleIntensityRanged,
+    Spacingd,
+    RandRotate90d,
+    EnsureTyped,
+)
+from monai.utils import set_determinism
+
+from monai.config import print_config
+from monai.metrics import DiceMetric
+from monai.networks.nets import SwinUNETR
+
+from monai.data import (
+    ThreadDataLoader,
+    CacheDataset,
+    load_decathlon_datalist,
+    decollate_batch,
+    set_track_meta,
+    get_track_meta,
+)
+
+import nvtx
+from monai.utils.nvtx import Range
+
+import torch
+
+import argparse
+
+print_config()
+
+parser = argparse.ArgumentParser(description='Profiling Swin UNETR.')
+parser.add_argument('--epochs', default=5, type=int)
+parser.add_argument('--val_epochs', default=-1, type=int, help='validation every X epochs; if non-positive value entered, will perform validation only once')
+parser.add_argument('--batch_size', default=1, type=int)
+parser.add_argument('--thread_workers', default=False, type=bool)
+parser.add_argument('--num_workers', default=0, type=int)
+args = parser.parse_args()
+print(args)
+
+assert args.epochs >= 0
+assert args.batch_size > 0
+assert args.num_workers >= 0
+
+if not args.thread_workers:
+    args.num_workers = 0
+
+# +
+max_iterations = args.epochs * args.batch_size
+max_iterations = max(max_iterations, 24)
+
+eval_num = args.val_epochs * args.batch_size - 1
+if eval_num <= 0:
+    eval_num = max_iterations - 1
+# -
+
+
+
+directory = os.environ.get("MONAI_DATA_DIRECTORY")
+root_dir = tempfile.mkdtemp() if directory is None else directory
+print(root_dir)
+
+
+num_samples = 4
+os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+train_transforms = Compose(
+    [
+        Range()(LoadImaged(keys=["image", "label"])),
+        Range()(AddChanneld(keys=["image", "label"])),
+        Range()(Orientationd(keys=["image", "label"], axcodes="RAS")),
+        Range()(Spacingd(
+            keys=["image", "label"],
+            pixdim=(1.5, 1.5, 2.0),
+            mode=("bilinear", "nearest"),
+        )),
+        Range()(ScaleIntensityRanged(
+            keys=["image"],
+            a_min=-175,
+            a_max=250,
+            b_min=0.0,
+            b_max=1.0,
+            clip=True,
+        )),
+        Range()(CropForegroundd(keys=["image", "label"], source_key="image")),
+#         Range()(EnsureTyped(keys=["image", "label"])),
+#         Range()(ToDeviced(keys=["image", "label"], device=device)),
+        Range()(EnsureTyped(keys=["image", "label"], device=device, track_meta=False)),
+        Range()(RandCropByPosNegLabeld(
+            keys=["image", "label"],
+            label_key="label",
+            spatial_size=(96, 96, 96),
+            pos=1,
+            neg=1,
+            num_samples=num_samples,
+            image_key="image",
+            image_threshold=0,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[0],
+            prob=0.10,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[1],
+            prob=0.10,
+        )),
+        Range()(RandFlipd(
+            keys=["image", "label"],
+            spatial_axis=[2],
+            prob=0.10,
+        )),
+        Range()(RandRotate90d(
+            keys=["image", "label"],
+            prob=0.10,
+            max_k=3,
+        )),
+        Range()(RandShiftIntensityd(
+            keys=["image"],
+            offsets=0.10,
+            prob=0.50,
+        )),
+    ]
+)
+val_transforms = Compose(
+    [
+        LoadImaged(keys=["image", "label"]),
+        AddChanneld(keys=["image", "label"]),
+        Orientationd(keys=["image", "label"], axcodes="RAS"),
+        Spacingd(
+            keys=["image", "label"],
+            pixdim=(1.5, 1.5, 2.0),
+            mode=("bilinear", "nearest"),
+        ),
+        ScaleIntensityRanged(
+            keys=["image"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True
+        ),
+        CropForegroundd(keys=["image", "label"], source_key="image"),
+        EnsureTyped(keys=["image", "label"], device=device, track_meta=True),
+    ]
+)
+
+
+torch.cuda.empty_cache()
+data_dir = "data/"
+split_JSON = "dataset_0.json"
+
+datasets = data_dir + split_JSON
+datalist = load_decathlon_datalist(datasets, True, "training")
+val_files = load_decathlon_datalist(datasets, True, "validation")
+train_ds = CacheDataset(
+    data=datalist,
+    transform=train_transforms,
+    cache_num=24,
+    cache_rate=1.0,
+    num_workers=8,
+)
+train_loader = ThreadDataLoader(train_ds, batch_size=args.batch_size, shuffle=True, use_thread_workers=args.thread_workers, num_workers=args.num_workers)
+val_ds = CacheDataset(
+    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4
+)
+val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)
+
+set_track_meta(False)
+
+model = SwinUNETR(
+    img_size=(96, 96, 96),
+    in_channels=1,
+    out_channels=14,
+    feature_size=48,
+    use_checkpoint=True,
+).to(device)
+
+
+weight = torch.load("./model_swinvit.pt")
+model.load_from(weights=weight)
+print("Using pretrained self-supervied Swin UNETR backbone weights !")
+
+
+torch.backends.cudnn.benchmark = True
+loss_function = DiceCELoss(to_onehot_y=True, softmax=True)
+optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)
+scaler = torch.cuda.amp.GradScaler()
+
+post_label = AsDiscrete(to_onehot=14)
+post_pred = AsDiscrete(argmax=True, to_onehot=14)
+dice_metric = DiceMetric(include_background=True, reduction="mean", get_not_nans=False)
+
+def validation(epoch_iterable_val):
+    model.eval()
+    epoch_iterator_val = iter(epoch_iterable_val)
+    with torch.no_grad():
+        for _ in range(len(epoch_iterable_val)):
+            with nvtx.annotate("val dataload", color="red"):
+                batch = next(epoch_iterator_val)
+                val_inputs, val_labels = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
+
+            with nvtx.annotate("sliding window", color="green"):
+                with torch.cuda.amp.autocast():
+                    val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
+
+            with nvtx.annotate("decollate batch", color="blue"):
+                val_labels_list = decollate_batch(val_labels)
+                val_labels_convert = [
+                    post_label(val_label_tensor) for val_label_tensor in val_labels_list
+                ]
+                val_outputs_list = decollate_batch(val_outputs)
+                val_output_convert = [
+                    post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list
+                ]
+
+            with nvtx.annotate("compute metric", color="yellow"):
+                dice_metric(y_pred=val_output_convert, y=val_labels_convert)
+
+            epoch_iterable_val.set_description(
+                "Validate (%d / %d Steps)" % (global_step, 10.0)
+            )
+
+        mean_dice_val = dice_metric.aggregate().item()
+        dice_metric.reset()
+    return mean_dice_val
+
+
+def train(global_step, train_loader, dice_val_best, global_step_best):
+    model.train()
+    epoch_loss = 0
+    step = 0
+    epoch_iterable = tqdm(
+        train_loader, desc="Training (X / X Steps) (loss=X.X)", dynamic_ncols=True
+    )
+    epoch_iterator = iter(epoch_iterable)
+
+    for step in range(len(epoch_iterable)):
+        step += 1
+
+        with nvtx.annotate("dataload", color="red"):
+            batch = next(epoch_iterator)
+            x, y = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
+
+        with nvtx.annotate("forward", color="green"):
+            with torch.cuda.amp.autocast():
+                logit_map = model(x)
+                loss = loss_function(logit_map, y)
+
+        with nvtx.annotate("backward", color="blue"):
+            scaler.scale(loss).backward()
+            epoch_loss += loss.item()
+
+        with nvtx.annotate("update", color="yellow"):
+            scaler.unscale_(optimizer)
+            scaler.step(optimizer)
+            scaler.update()
+            optimizer.zero_grad()
+
+        epoch_iterable.set_description(
+            "Training (%d / %d Steps) (loss=%2.5f)"
+            % (global_step, max_iterations, loss)
+        )
+        if (
+            global_step % eval_num == 0 and global_step != 0
+        ) or global_step == max_iterations:
+            epoch_iterable_val = tqdm(
+                val_loader, desc="Validate (X / X Steps) (dice=X.X)", dynamic_ncols=True
+            )
+            dice_val = validation(epoch_iterable_val)
+            # FIXME: epoch_loss is a running average at time of validation??
+            with nvtx.annotate("post-validation processing", color="blue"):
+                epoch_loss /= step
+                epoch_loss_values.append(epoch_loss)
+                metric_values.append(dice_val)
+                if dice_val > dice_val_best:
+                    dice_val_best = dice_val
+                    global_step_best = global_step
+                    torch.save(
+                        model.state_dict(), os.path.join(root_dir, "best_metric_model.pth")
+                    )
+                    print(
+                        "Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
+                            dice_val_best, dice_val
+                        )
+                    )
+                else:
+                    print(
+                        "Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
+                            dice_val_best, dice_val
+                        )
+                    )
+        global_step += 1
+    return global_step, dice_val_best, global_step_best
+
+# max_iterations = 30000
+# eval_num = 500
+set_determinism(seed=0)
+global_step = 0
+dice_val_best = 0.0
+global_step_best = 0
+epoch_loss_values = []
+metric_values = []
+
+begin = time.time()
+while global_step < max_iterations:
+    with nvtx.annotate("epoch", color="red"):
+        global_step, dice_val_best, global_step_best = train(
+            global_step, train_loader, dice_val_best, global_step_best
+        )
+print(f"Total train time: {time.time() - begin:.2f} seconds")
+
+
+print(
+    f"train completed, best_metric: {dice_val_best:.4f} "
+    f"at iteration: {global_step_best}"
+)
+
+if directory is None:
+    shutil.rmtree(root_dir)

From e5f4c8eb79af814899ca3b8b079b447c68596908 Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Wed, 27 Jul 2022 14:28:55 -0700
Subject: [PATCH 16/17] remove test files

---
 3d_segmentation/swin_unetr_profiling.py     | 335 -------------------
 3d_segmentation/swin_unetr_profiling_new.py | 338 --------------------
 2 files changed, 673 deletions(-)
 delete mode 100644 3d_segmentation/swin_unetr_profiling.py
 delete mode 100644 3d_segmentation/swin_unetr_profiling_new.py

diff --git a/3d_segmentation/swin_unetr_profiling.py b/3d_segmentation/swin_unetr_profiling.py
deleted file mode 100644
index 0bb4bb22a6..0000000000
--- a/3d_segmentation/swin_unetr_profiling.py
+++ /dev/null
@@ -1,335 +0,0 @@
-# command line: "pip install monai==0.9.1; nsys profile --output /results/orig --force-overwrite true --trace-fork-before-exec true python3 swin_unetr_profiling.py --epochs 5 --val_epochs 5 --batch_size 1 --thread_workers False --num_workers 0"
-
-import os
-import shutil
-import tempfile
-
-import matplotlib.pyplot as plt
-import numpy as np
-from tqdm import tqdm
-import time
-
-from monai.losses import DiceCELoss
-from monai.inferers import sliding_window_inference
-from monai.transforms import (
-    AsDiscrete,
-    AddChanneld,
-    Compose,
-    CropForegroundd,
-    LoadImaged,
-    Orientationd,
-    RandFlipd,
-    RandCropByPosNegLabeld,
-    RandShiftIntensityd,
-    ScaleIntensityRanged,
-    Spacingd,
-    RandRotate90d,
-    ToTensord,
-    EnsureTyped,
-    ToDeviced
-)
-from monai.utils import set_determinism
-
-from monai.config import print_config
-from monai.metrics import DiceMetric
-from monai.networks.nets import SwinUNETR
-
-from monai.data import (
-#    ThreadDataLoader,
-    DataLoader,
-    Dataset,
-    load_decathlon_datalist,
-    decollate_batch,
-)
-
-import nvtx
-from monai.utils.nvtx import Range
-
-import torch
-
-import argparse
-
-print_config()
-
-parser = argparse.ArgumentParser(description='Profiling Swin UNETR.')
-parser.add_argument('--epochs', default=5, type=int)
-parser.add_argument('--val_epochs', default=-1, type=int, help='validation every X epochs; if non-positive value entered, will perform validation only once')
-parser.add_argument('--batch_size', default=1, type=int)
-parser.add_argument('--thread_workers', default=False, type=bool)
-parser.add_argument('--num_workers', default=0, type=int)
-args = parser.parse_args()
-print(args)
-
-assert args.epochs >= 0
-assert args.batch_size > 0
-assert args.num_workers >= 0
-
-if not args.thread_workers:
-    args.num_workers = 0
-
-# +
-max_iterations = args.epochs * args.batch_size
-max_iterations = max(max_iterations, 24)
-
-eval_num = args.val_epochs * args.batch_size - 1
-if eval_num <= 0:
-    eval_num = max_iterations - 1
-# -
-
-
-
-directory = os.environ.get("MONAI_DATA_DIRECTORY")
-root_dir = tempfile.mkdtemp() if directory is None else directory
-print(root_dir)
-
-
-num_samples = 4
-os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-
-train_transforms = Compose(
-    [
-        Range()(LoadImaged(keys=["image", "label"])),
-        Range()(AddChanneld(keys=["image", "label"])),
-        Range()(Orientationd(keys=["image", "label"], axcodes="RAS")),
-        Range()(Spacingd(
-            keys=["image", "label"],
-            pixdim=(1.5, 1.5, 2.0),
-            mode=("bilinear", "nearest"),
-        )),
-        Range()(ScaleIntensityRanged(
-            keys=["image"],
-            a_min=-175,
-            a_max=250,
-            b_min=0.0,
-            b_max=1.0,
-            clip=True,
-        )),
-        Range()(CropForegroundd(keys=["image", "label"], source_key="image")),
-#         Range()(EnsureTyped(keys=["image", "label"])),
-#         Range()(ToDeviced(keys=["image", "label"], device=device)),
-        Range()(RandCropByPosNegLabeld(
-            keys=["image", "label"],
-            label_key="label",
-            spatial_size=(96, 96, 96),
-            pos=1,
-            neg=1,
-            num_samples=num_samples,
-            image_key="image",
-            image_threshold=0,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[0],
-            prob=0.10,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[1],
-            prob=0.10,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[2],
-            prob=0.10,
-        )),
-        Range()(RandRotate90d(
-            keys=["image", "label"],
-            prob=0.10,
-            max_k=3,
-        )),
-        Range()(RandShiftIntensityd(
-            keys=["image"],
-            offsets=0.10,
-            prob=0.50,
-        )),
-#        Range()(ToTensord(keys=["image", "label"])),
-    ]
-)
-val_transforms = Compose(
-    [
-        LoadImaged(keys=["image", "label"]),
-        AddChanneld(keys=["image", "label"]),
-        Orientationd(keys=["image", "label"], axcodes="RAS"),
-        Spacingd(
-            keys=["image", "label"],
-            pixdim=(1.5, 1.5, 2.0),
-            mode=("bilinear", "nearest"),
-        ),
-        ScaleIntensityRanged(
-            keys=["image"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True
-        ),
-        CropForegroundd(keys=["image", "label"], source_key="image"),
-#        ToTensord(keys=["image", "label"]),
-#        EnsureTyped(keys=["image", "label"]),
-#        ToDeviced(keys=["image", "label"], device=device),
-    ]
-)
-
-
-torch.cuda.empty_cache()
-data_dir = "data/"
-split_JSON = "dataset_0.json"
-
-datasets = data_dir + split_JSON
-datalist = load_decathlon_datalist(datasets, True, "training")
-val_files = load_decathlon_datalist(datasets, True, "validation")
-
-# TODO: try thread_workers
-train_ds = Dataset(data=datalist, transform=train_transforms)
-# train_ds = CacheDataset(data=datalist, transform=train_transforms, cache_num=24, cache_rate=1.0, num_workers=8,)
-# train_loader = ThreadDataLoader(train_ds, batch_size=1, shuffle=True, num_workers=0)
-# train_loader = ThreadDataLoader(train_ds, batch_size=args.batch_size, shuffle=True, use_thread_workers=args.thread_workers, num_workers=args.num_workers)
-train_loader = DataLoader(train_ds, batch_size=args.batch_size, shuffle=True, num_workers=8, pin_memory=True)
-
-# val_ds = CacheDataset(data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4)
-val_ds = Dataset(data=val_files, transform=val_transforms)
-# val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)
-val_loader = DataLoader(val_ds, batch_size=args.batch_size, shuffle=False, num_workers=4, pin_memory=True)
-
-model = SwinUNETR(
-    img_size=(96, 96, 96),
-    in_channels=1,
-    out_channels=14,
-    feature_size=48,
-    use_checkpoint=True,
-).to(device)
-
-
-weight = torch.load("./model_swinvit.pt")
-model.load_from(weights=weight)
-print("Using pretrained self-supervied Swin UNETR backbone weights !")
-
-
-torch.backends.cudnn.benchmark = True
-loss_function = DiceCELoss(to_onehot_y=True, softmax=True)
-optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)
-
-post_label = AsDiscrete(to_onehot=14)
-post_pred = AsDiscrete(argmax=True, to_onehot=14)
-dice_metric = DiceMetric(include_background=True, reduction="mean", get_not_nans=False)
-
-def validation(epoch_iterable_val):
-    model.eval()
-    epoch_iterator_val = iter(epoch_iterable_val)
-    with torch.no_grad():
-        for _ in range(len(epoch_iterable_val)):
-            with nvtx.annotate("val dataload", color="red"):
-                batch = next(epoch_iterator_val)
-                val_inputs, val_labels = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
-
-            with nvtx.annotate("sliding window", color="green"):
-#                 with torch.cuda.amp.autocast():
-                val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
-
-            with nvtx.annotate("decollate batch", color="blue"):
-                val_labels_list = decollate_batch(val_labels)
-                val_labels_convert = [
-                    post_label(val_label_tensor) for val_label_tensor in val_labels_list
-                ]
-                val_outputs_list = decollate_batch(val_outputs)
-                val_output_convert = [
-                    post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list
-                ]
-
-            with nvtx.annotate("compute metric", color="yellow"):
-                dice_metric(y_pred=val_output_convert, y=val_labels_convert)
-
-            epoch_iterable_val.set_description(
-                "Validate (%d / %d Steps)" % (global_step, 10.0)
-            )
-
-        mean_dice_val = dice_metric.aggregate().item()
-        dice_metric.reset()
-    return mean_dice_val
-
-
-def train(global_step, train_loader, dice_val_best, global_step_best):
-    model.train()
-    epoch_loss = 0
-    step = 0
-    epoch_iterable = tqdm(
-        train_loader, desc="Training (X / X Steps) (loss=X.X)", dynamic_ncols=True
-    )
-    epoch_iterator = iter(epoch_iterable)
-
-    for step in range(len(epoch_iterable)):
-        step += 1
-
-        with nvtx.annotate("dataload", color="red"):
-            batch = next(epoch_iterator)
-            x, y = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
-#         with torch.cuda.amp.autocast():
-        with nvtx.annotate("forward", color="green"):
-            logit_map = model(x)
-            loss = loss_function(logit_map, y)
-
-        with nvtx.annotate("backward", color="blue"):
-            loss.backward()
-            epoch_loss += loss.item()
-
-        with nvtx.annotate("update", color="yellow"):
-            optimizer.step()
-            optimizer.zero_grad()
-
-        epoch_iterable.set_description(
-            "Training (%d / %d Steps) (loss=%2.5f)"
-            % (global_step, max_iterations, loss)
-        )
-        if (
-            global_step % eval_num == 0 and global_step != 0
-        ) or global_step == max_iterations:
-            epoch_iterable_val = tqdm(
-                val_loader, desc="Validate (X / X Steps) (dice=X.X)", dynamic_ncols=True
-            )
-            dice_val = validation(epoch_iterable_val)
-            # FIXME: epoch_loss is a running average at time of validation??
-            with nvtx.annotate("post-validation processing", color="blue"):
-                epoch_loss /= step
-                epoch_loss_values.append(epoch_loss)
-                metric_values.append(dice_val)
-                if dice_val > dice_val_best:
-                    dice_val_best = dice_val
-                    global_step_best = global_step
-                    torch.save(
-                        model.state_dict(), os.path.join(root_dir, "best_metric_model.pth")
-                    )
-                    print(
-                        "Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
-                            dice_val_best, dice_val
-                        )
-                    )
-                else:
-                    print(
-                        "Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
-                            dice_val_best, dice_val
-                        )
-                    )
-        global_step += 1
-    return global_step, dice_val_best, global_step_best
-
-# max_iterations = 30000
-# eval_num = 500
-set_determinism(seed=0)
-global_step = 0
-dice_val_best = 0.0
-global_step_best = 0
-epoch_loss_values = []
-metric_values = []
-
-begin = time.time()
-while global_step < max_iterations:
-    with nvtx.annotate("epoch", color="red"):
-        global_step, dice_val_best, global_step_best = train(
-            global_step, train_loader, dice_val_best, global_step_best
-        )
-print(f"Total train time: {time.time() - begin:.2f} seconds")
-
-
-print(
-    f"train completed, best_metric: {dice_val_best:.4f} "
-    f"at iteration: {global_step_best}"
-)
-
-if directory is None:
-    shutil.rmtree(root_dir)
diff --git a/3d_segmentation/swin_unetr_profiling_new.py b/3d_segmentation/swin_unetr_profiling_new.py
deleted file mode 100644
index 5186f9640d..0000000000
--- a/3d_segmentation/swin_unetr_profiling_new.py
+++ /dev/null
@@ -1,338 +0,0 @@
-# command line: "pip install monai==0.9.1; nsys profile --output /results/orig --force-overwrite true --trace-fork-before-exec true python3 swin_unetr_profiling.py --epochs 0 --val_epochs 0 --batch_size 1 --thread_workers False --num_workers 0"
-
-import os
-import shutil
-import tempfile
-
-import matplotlib.pyplot as plt
-import numpy as np
-from tqdm import tqdm
-import time
-
-from monai.losses import DiceCELoss
-from monai.inferers import sliding_window_inference
-from monai.transforms import (
-    AsDiscrete,
-    AddChanneld,
-    Compose,
-    CropForegroundd,
-    LoadImaged,
-    Orientationd,
-    RandFlipd,
-    RandCropByPosNegLabeld,
-    RandShiftIntensityd,
-    ScaleIntensityRanged,
-    Spacingd,
-    RandRotate90d,
-    EnsureTyped,
-)
-from monai.utils import set_determinism
-
-from monai.config import print_config
-from monai.metrics import DiceMetric
-from monai.networks.nets import SwinUNETR
-
-from monai.data import (
-    ThreadDataLoader,
-    CacheDataset,
-    load_decathlon_datalist,
-    decollate_batch,
-    set_track_meta,
-    get_track_meta,
-)
-
-import nvtx
-from monai.utils.nvtx import Range
-
-import torch
-
-import argparse
-
-print_config()
-
-parser = argparse.ArgumentParser(description='Profiling Swin UNETR.')
-parser.add_argument('--epochs', default=5, type=int)
-parser.add_argument('--val_epochs', default=-1, type=int, help='validation every X epochs; if non-positive value entered, will perform validation only once')
-parser.add_argument('--batch_size', default=1, type=int)
-parser.add_argument('--thread_workers', default=False, type=bool)
-parser.add_argument('--num_workers', default=0, type=int)
-args = parser.parse_args()
-print(args)
-
-assert args.epochs >= 0
-assert args.batch_size > 0
-assert args.num_workers >= 0
-
-if not args.thread_workers:
-    args.num_workers = 0
-
-# +
-max_iterations = args.epochs * args.batch_size
-max_iterations = max(max_iterations, 24)
-
-eval_num = args.val_epochs * args.batch_size - 1
-if eval_num <= 0:
-    eval_num = max_iterations - 1
-# -
-
-
-
-directory = os.environ.get("MONAI_DATA_DIRECTORY")
-root_dir = tempfile.mkdtemp() if directory is None else directory
-print(root_dir)
-
-
-num_samples = 4
-os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-
-train_transforms = Compose(
-    [
-        Range()(LoadImaged(keys=["image", "label"])),
-        Range()(AddChanneld(keys=["image", "label"])),
-        Range()(Orientationd(keys=["image", "label"], axcodes="RAS")),
-        Range()(Spacingd(
-            keys=["image", "label"],
-            pixdim=(1.5, 1.5, 2.0),
-            mode=("bilinear", "nearest"),
-        )),
-        Range()(ScaleIntensityRanged(
-            keys=["image"],
-            a_min=-175,
-            a_max=250,
-            b_min=0.0,
-            b_max=1.0,
-            clip=True,
-        )),
-        Range()(CropForegroundd(keys=["image", "label"], source_key="image")),
-#         Range()(EnsureTyped(keys=["image", "label"])),
-#         Range()(ToDeviced(keys=["image", "label"], device=device)),
-        Range()(EnsureTyped(keys=["image", "label"], device=device, track_meta=False)),
-        Range()(RandCropByPosNegLabeld(
-            keys=["image", "label"],
-            label_key="label",
-            spatial_size=(96, 96, 96),
-            pos=1,
-            neg=1,
-            num_samples=num_samples,
-            image_key="image",
-            image_threshold=0,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[0],
-            prob=0.10,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[1],
-            prob=0.10,
-        )),
-        Range()(RandFlipd(
-            keys=["image", "label"],
-            spatial_axis=[2],
-            prob=0.10,
-        )),
-        Range()(RandRotate90d(
-            keys=["image", "label"],
-            prob=0.10,
-            max_k=3,
-        )),
-        Range()(RandShiftIntensityd(
-            keys=["image"],
-            offsets=0.10,
-            prob=0.50,
-        )),
-    ]
-)
-val_transforms = Compose(
-    [
-        LoadImaged(keys=["image", "label"]),
-        AddChanneld(keys=["image", "label"]),
-        Orientationd(keys=["image", "label"], axcodes="RAS"),
-        Spacingd(
-            keys=["image", "label"],
-            pixdim=(1.5, 1.5, 2.0),
-            mode=("bilinear", "nearest"),
-        ),
-        ScaleIntensityRanged(
-            keys=["image"], a_min=-175, a_max=250, b_min=0.0, b_max=1.0, clip=True
-        ),
-        CropForegroundd(keys=["image", "label"], source_key="image"),
-        EnsureTyped(keys=["image", "label"], device=device, track_meta=True),
-    ]
-)
-
-
-torch.cuda.empty_cache()
-data_dir = "data/"
-split_JSON = "dataset_0.json"
-
-datasets = data_dir + split_JSON
-datalist = load_decathlon_datalist(datasets, True, "training")
-val_files = load_decathlon_datalist(datasets, True, "validation")
-train_ds = CacheDataset(
-    data=datalist,
-    transform=train_transforms,
-    cache_num=24,
-    cache_rate=1.0,
-    num_workers=8,
-)
-train_loader = ThreadDataLoader(train_ds, batch_size=args.batch_size, shuffle=True, use_thread_workers=args.thread_workers, num_workers=args.num_workers)
-val_ds = CacheDataset(
-    data=val_files, transform=val_transforms, cache_num=6, cache_rate=1.0, num_workers=4
-)
-val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)
-
-set_track_meta(False)
-
-model = SwinUNETR(
-    img_size=(96, 96, 96),
-    in_channels=1,
-    out_channels=14,
-    feature_size=48,
-    use_checkpoint=True,
-).to(device)
-
-
-weight = torch.load("./model_swinvit.pt")
-model.load_from(weights=weight)
-print("Using pretrained self-supervied Swin UNETR backbone weights !")
-
-
-torch.backends.cudnn.benchmark = True
-loss_function = DiceCELoss(to_onehot_y=True, softmax=True)
-optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=1e-5)
-scaler = torch.cuda.amp.GradScaler()
-
-post_label = AsDiscrete(to_onehot=14)
-post_pred = AsDiscrete(argmax=True, to_onehot=14)
-dice_metric = DiceMetric(include_background=True, reduction="mean", get_not_nans=False)
-
-def validation(epoch_iterable_val):
-    model.eval()
-    epoch_iterator_val = iter(epoch_iterable_val)
-    with torch.no_grad():
-        for _ in range(len(epoch_iterable_val)):
-            with nvtx.annotate("val dataload", color="red"):
-                batch = next(epoch_iterator_val)
-                val_inputs, val_labels = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
-
-            with nvtx.annotate("sliding window", color="green"):
-                with torch.cuda.amp.autocast():
-                    val_outputs = sliding_window_inference(val_inputs, (96, 96, 96), 4, model)
-
-            with nvtx.annotate("decollate batch", color="blue"):
-                val_labels_list = decollate_batch(val_labels)
-                val_labels_convert = [
-                    post_label(val_label_tensor) for val_label_tensor in val_labels_list
-                ]
-                val_outputs_list = decollate_batch(val_outputs)
-                val_output_convert = [
-                    post_pred(val_pred_tensor) for val_pred_tensor in val_outputs_list
-                ]
-
-            with nvtx.annotate("compute metric", color="yellow"):
-                dice_metric(y_pred=val_output_convert, y=val_labels_convert)
-
-            epoch_iterable_val.set_description(
-                "Validate (%d / %d Steps)" % (global_step, 10.0)
-            )
-
-        mean_dice_val = dice_metric.aggregate().item()
-        dice_metric.reset()
-    return mean_dice_val
-
-
-def train(global_step, train_loader, dice_val_best, global_step_best):
-    model.train()
-    epoch_loss = 0
-    step = 0
-    epoch_iterable = tqdm(
-        train_loader, desc="Training (X / X Steps) (loss=X.X)", dynamic_ncols=True
-    )
-    epoch_iterator = iter(epoch_iterable)
-
-    for step in range(len(epoch_iterable)):
-        step += 1
-
-        with nvtx.annotate("dataload", color="red"):
-            batch = next(epoch_iterator)
-            x, y = (batch["image"].cuda(device=device), batch["label"].cuda(device=device))
-
-        with nvtx.annotate("forward", color="green"):
-            with torch.cuda.amp.autocast():
-                logit_map = model(x)
-                loss = loss_function(logit_map, y)
-
-        with nvtx.annotate("backward", color="blue"):
-            scaler.scale(loss).backward()
-            epoch_loss += loss.item()
-
-        with nvtx.annotate("update", color="yellow"):
-            scaler.unscale_(optimizer)
-            scaler.step(optimizer)
-            scaler.update()
-            optimizer.zero_grad()
-
-        epoch_iterable.set_description(
-            "Training (%d / %d Steps) (loss=%2.5f)"
-            % (global_step, max_iterations, loss)
-        )
-        if (
-            global_step % eval_num == 0 and global_step != 0
-        ) or global_step == max_iterations:
-            epoch_iterable_val = tqdm(
-                val_loader, desc="Validate (X / X Steps) (dice=X.X)", dynamic_ncols=True
-            )
-            dice_val = validation(epoch_iterable_val)
-            # FIXME: epoch_loss is a running average at time of validation??
-            with nvtx.annotate("post-validation processing", color="blue"):
-                epoch_loss /= step
-                epoch_loss_values.append(epoch_loss)
-                metric_values.append(dice_val)
-                if dice_val > dice_val_best:
-                    dice_val_best = dice_val
-                    global_step_best = global_step
-                    torch.save(
-                        model.state_dict(), os.path.join(root_dir, "best_metric_model.pth")
-                    )
-                    print(
-                        "Model Was Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
-                            dice_val_best, dice_val
-                        )
-                    )
-                else:
-                    print(
-                        "Model Was Not Saved ! Current Best Avg. Dice: {} Current Avg. Dice: {}".format(
-                            dice_val_best, dice_val
-                        )
-                    )
-        global_step += 1
-    return global_step, dice_val_best, global_step_best
-
-# max_iterations = 30000
-# eval_num = 500
-set_determinism(seed=0)
-global_step = 0
-dice_val_best = 0.0
-global_step_best = 0
-epoch_loss_values = []
-metric_values = []
-
-begin = time.time()
-while global_step < max_iterations:
-    with nvtx.annotate("epoch", color="red"):
-        global_step, dice_val_best, global_step_best = train(
-            global_step, train_loader, dice_val_best, global_step_best
-        )
-print(f"Total train time: {time.time() - begin:.2f} seconds")
-
-
-print(
-    f"train completed, best_metric: {dice_val_best:.4f} "
-    f"at iteration: {global_step_best}"
-)
-
-if directory is None:
-    shutil.rmtree(root_dir)

From a521149a3dd11e3e186776984e0bcd547e4d3c6b Mon Sep 17 00:00:00 2001
From: Yuchen Xu <yuchenx@nvidia.com>
Date: Thu, 28 Jul 2022 09:22:37 -0700
Subject: [PATCH 17/17] updated json link and comments

---
 .../swin_unetr_btcv_segmentation_3d.ipynb            | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
index 1114184bcb..46a7fdb299 100644
--- a/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
+++ b/3d_segmentation/swin_unetr_btcv_segmentation_3d.ipynb
@@ -20,7 +20,7 @@
     "\n",
     "For this tutorial, the dataset needs to be downloaded from: https://www.synapse.org/#!Synapse:syn3193805/wiki/217752. More details are provided in the \"Download dataset\" section below.\n",
     "\n",
-    "In addition, the json file for data splits needs to be downloaded from this [link](https://drive.google.com/file/d/1t4fIQQkONv7ArTSZe4Nucwkk1KfdUDvW/view?usp=sharing). Once downloaded, place the json file in the same folder as the dataset. \n",
+    "In addition, the json file for data splits needs to be downloaded from this [link](https://drive.google.com/file/d/1qcGh41p-rI3H_sQ0JwOAhNiQSXriQqGi/view?usp=sharing). Once downloaded, place the json file in the same folder as the dataset. \n",
     "\n",
     "For BTCV dataset, under Institutional Review Board (IRB) supervision, 50 abdomen CT scans of were randomly selected from a combination of an ongoing colorectal cancer chemotherapy trial, and a retrospective ventral hernia study. The 50 scans were captured during portal venous contrast phase with variable volume sizes (512 x 512 x 85 - 512 x 512 x 198) and field of views (approx. 280 x 280 x 280 mm3 - 500 x 500 x 650 mm3). The in-plane resolution varies from 0.54 x 0.54 mm2 to 0.98 x 0.98 mm2, while the slice thickness ranges from 2.5 mm to 5.0 mm. \n",
     "\n",
@@ -188,7 +188,7 @@
     "\n",
     "A note on design related to MetaTensors:\n",
     "\n",
-    "- Summary: using `EnsureTyped(..., track_meta=False)` and `set_track_meta(False)` (later) speeds up training significantly.\n",
+    "- Summary: using `EnsureTyped(..., track_meta=False)` (caching) and `set_track_meta(False)` (during training) speeds up training significantly.\n",
     "\n",
     "- We are moving towards the use of MONAI's MetaTensor in place of numpy arrays or PyTorch tensors. MetaTensors have the benefit of carrying the metadata directly with the tensor, but in some use cases (like here with training, where training data are only used for computing loss and metadata is not useful), we can safely disregard the metadata to improve speed.\n",
     "\n",
@@ -299,7 +299,7 @@
     "\n",
     "3. Make a JSON file to define train/val split and other relevant parameters. Place the JSON file at `./data/dataset_0.json`.\n",
     "\n",
-    "    You can download an example of the JSON file [here](https://drive.google.com/file/d/1t4fIQQkONv7ArTSZe4Nucwkk1KfdUDvW/view?usp=sharing), or, equivalently, use the following `wget` command. If you would like to use this directly, please move it into the `./data` folder."
+    "    You can download an example of the JSON file [here](https://drive.google.com/file/d/1qcGh41p-rI3H_sQ0JwOAhNiQSXriQqGi/view?usp=sharing), or, equivalently, use the following `wget` command. If you would like to use this directly, please move it into the `./data` folder."
    ]
   },
   {
@@ -309,7 +309,7 @@
    "outputs": [],
    "source": [
     "# uncomment this command to download the JSON file directly\n",
-    "# wget -O data/dataset_0.json 'https://drive.google.com/uc?export=download&id=1t4fIQQkONv7ArTSZe4Nucwkk1KfdUDvW'"
+    "# wget -O data/dataset_0.json 'https://drive.google.com/uc?export=download&id=1qcGh41p-rI3H_sQ0JwOAhNiQSXriQqGi'"
    ]
   },
   {
@@ -337,6 +337,10 @@
     ")\n",
     "val_loader = ThreadDataLoader(val_ds, num_workers=0, batch_size=1)\n",
     "\n",
+    "# as explained in the \"Setup transforms\" section above, we want cached training images to not have metadata, and validations to have metadata\n",
+    "# the EnsureTyped transforms allow us to make this distinction\n",
+    "# on the other hand, set_track_meta is a global API; doing so here makes sure subsequent transforms (i.e., random transforms for training)\n",
+    "# will be carried out as Tensors, not MetaTensors\n",
     "set_track_meta(False)"
    ]
   },