{"cells":[{"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":55},"colab_type":"code","id":"VYWnEVhN2rBC","outputId":"27d57f8c-1aba-4178-afd4-16017f6d36e9","trusted":true},"cell_type":"code","source":"import gc\nimport os\nimport sys\nimport time\nimport random\nimport logging\nimport datetime as dt\n\nimport numpy as np\nimport pandas as pd\n\nimport torch\nimport torch.nn as nn\nimport torch.optim as optim\nimport torch.utils.data as data\nimport torch.nn.functional as F\nimport torchvision as vision\n\nfrom torch.optim.lr_scheduler import CosineAnnealingLR\n\nfrom pathlib import Path\nfrom PIL import Image\nfrom contextlib import contextmanager\n\nfrom joblib import Parallel, delayed\nfrom tqdm import tqdm\nfrom fastprogress import master_bar, progress_bar\n\nfrom sklearn.model_selection import KFold\nfrom sklearn.metrics import fbeta_score\n","execution_count":1,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#!mkdir -p /tmp/.torch/models/\n#!wget -O /tmp/.torch/models/se_resnet152-d17c99b7.pth http://data.lip6.fr/cadene/pretrainedmodels/se_resnet152-d17c99b7.pth\n#import pretrainedmodels\ntorch.cuda.is_available()","execution_count":2,"outputs":[{"output_type":"execute_result","execution_count":2,"data":{"text/plain":"True"},"metadata":{}}]},{"metadata":{"colab":{},"colab_type":"code","id":"Qlz4XJcr2xup","trusted":true},"cell_type":"code","source":"@contextmanager\ndef timer(name=\"Main\", logger=None):\n    t0 = time.time()\n    yield\n    msg = f\"[{name}] done in {time.time() - t0} s\"\n    if logger is not None:\n        logger.info(msg)\n    else:\n        print(msg)\n        \n\ndef get_logger(name=\"Main\", tag=\"exp\", log_dir=\"log/\"):\n    log_path = Path(log_dir)\n    path = log_path / tag\n    path.mkdir(exist_ok=True, parents=True)\n\n    logger = logging.getLogger(name)\n    logger.setLevel(logging.INFO)\n\n    fh = logging.FileHandler(\n        path / (dt.datetime.now().strftime(\"%Y-%m-%d-%H-%M-%S\") + \".log\"))\n    sh = logging.StreamHandler(sys.stdout)\n    formatter = logging.Formatter(\n        \"%(asctime)s %(name)s %(levelname)s %(message)s\")\n\n    fh.setFormatter(formatter)\n    sh.setFormatter(formatter)\n    logger.addHandler(fh)\n    logger.addHandler(sh)\n    return logger\n\n\ndef seed_torch(seed=1029):\n    random.seed(seed)\n    os.environ[\"PYTHONHASHSEED\"] = str(seed)\n    np.random.seed(seed)\n    torch.manual_seed(seed)\n    torch.cuda.manual_seed_all(seed)\n    torch.backends.cudnn.deterministic = True","execution_count":3,"outputs":[]},{"metadata":{"colab":{},"colab_type":"code","id":"IAZwBnUf28Bk","trusted":true},"cell_type":"code","source":"logger = get_logger(name=\"Main\", tag=\"Pytorch-VGG16\")","execution_count":4,"outputs":[]},{"metadata":{"colab":{},"colab_type":"code","id":"nC7RnkHd8EN6","trusted":true},"cell_type":"code","source":"@contextmanager\ndef timer(name=\"Main\", logger=None):\n    t0 = time.time()\n    yield\n    msg = f\"[{name}] done in {time.time() - t0} s\"\n    if logger is not None:\n        logger.info(msg)\n    else:\n        print(msg)\n        \n\ndef get_logger(name=\"Main\", tag=\"exp\", log_dir=\"log/\"):\n    log_path = Path(log_dir)\n    path = log_path / tag\n    path.mkdir(exist_ok=True, parents=True)\n\n    logger = logging.getLogger(name)\n    logger.setLevel(logging.INFO)\n\n    fh = logging.FileHandler(\n        path / (dt.datetime.now().strftime(\"%Y-%m-%d-%H-%M-%S\") + \".log\"))\n    sh = logging.StreamHandler(sys.stdout)\n    formatter = logging.Formatter(\n        \"%(asctime)s %(name)s %(levelname)s %(message)s\")\n\n    fh.setFormatter(formatter)\n    sh.setFormatter(formatter)\n    logger.addHandler(fh)\n    logger.addHandler(sh)\n    return logger\n\n\ndef seed_torch(seed=1029):\n    random.seed(seed)\n    os.environ[\"PYTHONHASHSEED\"] = str(seed)\n    np.random.seed(seed)\n    torch.manual_seed(seed)\n    torch.cuda.manual_seed_all(seed)\n    torch.backends.cudnn.deterministic = True","execution_count":5,"outputs":[]},{"metadata":{"colab":{},"colab_type":"code","id":"ZjwOZMBCQr3t","trusted":true},"cell_type":"code","source":"!ls ../input\nlabels = pd.read_csv(\"../input/imet-2019-fgvc6/labels.csv\")\ntrain = pd.read_csv(\"../input/imet-2019-fgvc6/train.csv\")\nsample = pd.read_csv(\"../input/imet-2019-fgvc6/sample_submission.csv\")\ntrain.head()\n\ncultures = [x for x in labels.attribute_name.values if x.startswith(\"culture\")]\ntags = [x for x in labels.attribute_name.values if x.startswith(\"tag\")]\nlen(cultures), len(tags)","execution_count":25,"outputs":[{"output_type":"stream","text":"imet-2019-fgvc6  pytorch-pretrained-image-models\r\n","name":"stdout"},{"output_type":"execute_result","execution_count":25,"data":{"text/plain":"(398, 705)"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"import cv2\ndef split_culture_tag(x):\n    cultures_ = list()\n    tags_ = list()\n    for i in x.split(\" \"):\n        if int(i) <= len(cultures):\n            cultures_.append(i)\n        else:\n            tags_.append(str(int(i) - len(cultures)))\n    if not cultures_:\n        cultures_.append(str(len(cultures)))\n    if not tags_:\n        tags_.append(str(len(tags)))\n    return \" \".join(cultures_), \" \".join(tags_)\n\nculture_ids = list()\ntag_ids = list()\n\nfor v in tqdm(train.attribute_ids.values):\n    c, t = split_culture_tag(v)\n    culture_ids.append(c)\n    tag_ids.append(t)\n\nnum_classes_c = len(cultures) + 1\nnum_classes_t = len(tags) + 1\n\ntrain[\"culture_ids\"] = culture_ids\ntrain[\"tag_ids\"] = tag_ids\n\n\ndef obtain_y_c(ids):\n    y = np.zeros(num_classes_c)\n    for idx in ids.split(\" \"):\n        y[int(idx)] = 1\n    return y\n\ndef obtain_y_t(ids):\n    y = np.zeros(num_classes_t)\n    for idx in ids.split(\" \"):\n        y[int(idx)] = 1\n    return y\n\npaths = [\"../input/imet-2019-fgvc6/train/{}.png\".format(x) for x in train.id.values]\n\ntargets_c = np.array([obtain_y_c(y) for y in train.culture_ids.values])\ntargets_t = np.array([obtain_y_t(y) for y in train.tag_ids.values])\nprint(targets_c.shape)\n\ndef rem_bkg(img):\n    y_size,x_size,col = img.shape\n    \n    for y in range(y_size):\n        for r in range(1,6):\n            col = img[y, x_size-r] \n            img[np.where((img == col).all(axis = 2))] = [255,255,255]\n        for l in range(5):\n            col = img[y, l] \n            img[np.where((img == col).all(axis = 2))] = [255,255,255]\n\n    for x in range(x_size):\n        for d in range(1,6):\n            col = img[y_size-d, x] \n            img[np.where((img == col).all(axis = 2))] = [255,255,255]\n        for u in range(5):\n            col = img[u, x] \n            img[np.where((img == col).all(axis = 2))] = [255,255,255]\n    \n    return img\n\nclass ImageDataLoader(data.DataLoader):\n    def __init__(self, root_dir: Path, \n                 df: pd.DataFrame, \n                 mode=\"train\", \n                 transforms=None):\n        self._root = root_dir\n        self.transform = transforms[mode]\n        self._img_id = (df[\"id\"] + \".png\").values\n        \n    def __len__(self):\n        return len(self._img_id)\n    \n    def __getitem__(self, idx):\n        img_id = self._img_id[idx]\n        file_name = self._root / img_id\n        img = Image.open(file_name)\n        #img = cv2.imread(file_name.absolute().as_posix())[...,[2, 1, 0]]\n        #img = rem_bkg(img)\n        #img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n        #img = Image.fromarray(img)\n        if self.transform:\n            img = self.transform(img)\n            \n        return [img]\n    \n    \ndata_transforms = {\n    'train': vision.transforms.Compose([\n        vision.transforms.RandomResizedCrop(224),\n        vision.transforms.RandomHorizontalFlip(),\n        vision.transforms.ToTensor(),\n        vision.transforms.Normalize(\n            [0.485, 0.456, 0.406], \n            [0.229, 0.224, 0.225])\n    ]),\n    'val': vision.transforms.Compose([\n        vision.transforms.Resize(256),\n        vision.transforms.CenterCrop(224),\n        vision.transforms.ToTensor(),\n        vision.transforms.Normalize(\n            [0.485, 0.456, 0.406], \n            [0.229, 0.224, 0.225])\n    ]),\n}\n\ndata_transforms[\"test\"] = data_transforms[\"val\"]","execution_count":72,"outputs":[{"output_type":"stream","text":"\n\n\n  0%|          | 0/109237 [00:00<?, ?it/s]\u001b[A\u001b[A\u001b[A\n\n\n 20%|██        | 22206/109237 [00:00<00:00, 222046.98it/s]\u001b[A\u001b[A\u001b[A\n\n\n 39%|███▉      | 42957/109237 [00:00<00:00, 217474.70it/s]\u001b[A\u001b[A\u001b[A\n\n\n 59%|█████▊    | 63910/109237 [00:00<00:00, 215015.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n 78%|███████▊  | 85587/109237 [00:00<00:00, 215537.16it/s]\u001b[A\u001b[A\u001b[A\n\n\n 97%|█████████▋| 106072/109237 [00:00<00:00, 212214.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n100%|██████████| 109237/109237 [00:00<00:00, 210271.21it/s]\u001b[A\u001b[A\u001b[A","name":"stderr"},{"output_type":"stream","text":"(109237, 399)\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"class IMetDataset(data.Dataset):\n    def __init__(self, tensor, device=\"cuda:0\", labels=None):\n        self.tensor = tensor\n        self.labels = labels\n        self.device= device\n        \n    def __len__(self):\n        return self.tensor.size(0)\n    \n    def __getitem__(self, idx):\n        tensor = self.tensor[idx, :]\n        if self.labels is not None:\n            label = self.labels[idx]\n            label_tensor = torch.zeros((1, 1103))\n            y_c = torch.FloatTensor(targets_c[idx]).to(self.device)\n            y_t = torch.FloatTensor(targets_t[idx]).to(self.device)\n            for i in label:\n                label_tensor[0, int(i)] = 1\n            label_tensor = label_tensor.to(self.device)\n            return [tensor, [y_c, y_t]]\n        else:\n            return [tensor]","execution_count":81,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"class Classifier(nn.Module):\n    def __init__(self):\n        super(Classifier, self).__init__()\n        \n    def forward(self, x):\n        return x\n\n\nclass Densenet121(nn.Module):\n    def __init__(self, pretrained: Path):\n        super(Densenet121, self).__init__()\n        self.densenet121 = vision.models.densenet121()\n        self.densenet121.load_state_dict(torch.load(pretrained))\n        self.densenet121.classifier = Classifier()\n        \n        dense = nn.Sequential(*list(self.densenet121.children())[:-1])\n        for param in dense.parameters():\n            param.requires_grad = False\n        \n    def forward(self, x):\n        return self.densenet121(x)\n    \nclass Resnet50(nn.Module):\n    def __init__(self, pretrained: Path):\n        super(Resnet50, self).__init__()\n        self.resnet50 = vision.models.resnet50()\n        self.resnet50.load_state_dict(torch.load(pretrained))\n        self.resnet50.classifier = Classifier()\n        \n        dense = nn.Sequential(*list(self.resnet50.children())[:-1])\n        for param in dense.parameters():\n            param.requires_grad = False\n        \n    def forward(self, x):\n        return self.resnet50(x)\n    \n    \nclass MultiLayerPerceptron(nn.Module):\n    def __init__(self):\n        super(MultiLayerPerceptron, self).__init__()\n        self.linear1 = nn.Linear(1024, 1024)\n        self.relu = nn.ReLU()\n        self.linear11 = nn.Linear(1024, 1024)\n        self.relu2 = nn.ReLU()\n        self.linear2 = nn.Linear(1024, 1103)\n        self.dropout = nn.Dropout(0.5)\n        self.sigmoid = nn.Sigmoid()\n        \n    def forward(self, x):\n        x = self.relu(self.linear1(x))\n        x = self.relu2(self.linear11(x))\n        x = self.dropout(x)\n        return self.sigmoid(self.linear2(x))\n    \nclass MultiLayerPerceptron1(nn.Module):\n    def __init__(self):\n        super(MultiLayerPerceptron1, self).__init__()\n        self.linear1 = nn.Linear(1024, 1024)\n        self.relu = nn.ReLU()\n        self.linear11 = nn.Linear(1024, 1024)\n        self.relu2 = nn.ReLU()\n        self.linear2 = nn.Linear(1024, 399)\n        self.dropout = nn.Dropout(0.5)\n        self.sigmoid = nn.Sigmoid()\n        \n    def forward(self, x):\n        x = self.relu(self.linear1(x))\n        x = self.relu2(self.linear11(x))\n        x = self.dropout(x)\n        return self.sigmoid(self.linear2(x))\n    \nclass MultiLayerPerceptron2(nn.Module):\n    def __init__(self):\n        super(MultiLayerPerceptron2, self).__init__()\n        self.linear1 = nn.Linear(1024, 1024)\n        self.relu = nn.ReLU()\n        self.linear11 = nn.Linear(1024, 1024)\n        self.relu2 = nn.ReLU()\n        self.linear2 = nn.Linear(1024, 706)\n        self.dropout = nn.Dropout(0.5)\n        self.sigmoid = nn.Sigmoid()\n        \n    def forward(self, x):\n        x = self.relu(self.linear1(x))\n        x = self.relu2(self.linear11(x))\n        x = self.dropout(x)\n        return self.sigmoid(self.linear2(x))","execution_count":82,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_dataset = ImageDataLoader(\n    root_dir=Path(\"../input/imet-2019-fgvc6/train/\"),\n    df=train,\n    mode=\"train\",\n    transforms=data_transforms)\ntrain_loader = data.DataLoader(dataset=train_dataset,\n                               shuffle=False,\n                               batch_size=64)\ntest_dataset = ImageDataLoader(\n    root_dir=Path(\"../input/imet-2019-fgvc6/test/\"),\n    df=sample,\n    mode=\"test\",\n    transforms=data_transforms)\ntest_loader = data.DataLoader(dataset=test_dataset,\n                              shuffle=False,\n                              batch_size=64)","execution_count":83,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from torchvision import models\ndef get_feature_vector(df, loader, device):\n    matrix = torch.zeros((df.shape[0], 1024)).to(device)\n    model = Densenet121('../input/pytorch-pretrained-image-models/densenet121.pth') #Resnet50('../input/pytorch-pretrained-image-models/resnet50.pth')\n\n    model.to(device)\n    batch = loader.batch_size\n    for i, (i_batch,) in tqdm(enumerate(loader)):\n        i_batch = i_batch.to(device)\n        pred = model(i_batch).detach()\n        matrix[i * batch:(i + 1) * batch] = pred\n    return matrix","execution_count":36,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train_tensor = get_feature_vector(train, train_loader, \"cuda:0\")\ntest_tensor = get_feature_vector(sample, test_loader, \"cuda:0\")","execution_count":37,"outputs":[{"output_type":"stream","text":"\n\n\n0it [00:00, ?it/s]\u001b[A\u001b[A\u001b[A\n\n\n1it [00:00,  1.15it/s]\u001b[A\u001b[A\u001b[A\n\n\n2it [00:01,  1.23it/s]\u001b[A\u001b[A\u001b[A\n\n\n3it [00:02,  1.29it/s]\u001b[A\u001b[A\u001b[A\n\n\n4it [00:02,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n5it [00:03,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n6it [00:04,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n7it [00:04,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n8it [00:05,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n9it [00:06,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n10it [00:06,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n11it [00:07,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n12it [00:08,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n13it [00:08,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n14it [00:09,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n15it [00:10,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n16it [00:10,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n17it [00:11,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n18it [00:12,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n19it [00:12,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n20it [00:13,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n21it [00:14,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n22it [00:15,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n23it [00:15,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n24it [00:16,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n25it [00:17,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n26it [00:17,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n27it [00:18,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n28it [00:19,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n29it [00:19,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n30it [00:20,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n31it [00:20,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n32it [00:21,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n33it [00:22,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n34it [00:22,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n35it [00:23,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n36it [00:24,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n37it [00:24,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n38it [00:25,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n39it [00:26,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n40it [00:26,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n41it [00:27,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n42it [00:28,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n43it [00:28,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n44it [00:29,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n45it [00:30,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n46it [00:30,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n47it [00:31,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n48it [00:32,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n49it [00:32,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n50it [00:33,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n51it [00:34,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n52it [00:34,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n53it [00:35,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n54it [00:36,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n55it [00:36,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n56it [00:37,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n57it [00:38,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n58it [00:38,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n59it [00:39,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n60it [00:40,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n61it [00:40,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n62it [00:41,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n63it [00:42,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n64it [00:42,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n65it [00:43,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n66it [00:44,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n67it [00:44,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n68it [00:45,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n69it [00:46,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n70it [00:46,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n71it [00:47,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n72it [00:48,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n73it [00:48,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n74it [00:49,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n75it [00:50,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n76it [00:50,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n77it [00:51,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n78it [00:52,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n79it [00:52,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n80it [00:53,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n81it [00:54,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n82it [00:54,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n83it [00:55,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n84it [00:56,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n85it [00:56,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n86it [00:57,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n87it [00:58,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n88it [00:58,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n89it [00:59,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n90it [01:00,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n91it [01:01,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n92it [01:01,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n93it [01:02,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n94it [01:02,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n95it [01:03,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n96it [01:04,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n97it [01:05,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n98it [01:05,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n99it [01:06,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n100it [01:07,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n101it [01:07,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n102it [01:08,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n103it [01:08,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n104it [01:09,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n105it [01:10,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n106it [01:11,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n107it [01:11,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n108it [01:12,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n109it [01:12,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n110it [01:13,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n111it [01:14,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n112it [01:15,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n113it [01:15,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n114it [01:16,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n115it [01:16,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n116it [01:17,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n117it [01:18,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n118it [01:18,  1.55it/s]\u001b[A\u001b[A\u001b[A\n\n\n119it [01:19,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n120it [01:20,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n121it [01:20,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n122it [01:21,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n123it [01:22,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n124it [01:22,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n125it [01:23,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n126it [01:24,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n127it [01:24,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n128it [01:25,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n129it [01:26,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n130it [01:26,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n131it [01:27,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n132it [01:28,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n133it [01:28,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n134it [01:29,  1.55it/s]\u001b[A\u001b[A\u001b[A\n\n\n135it [01:30,  1.56it/s]\u001b[A\u001b[A\u001b[A\n\n\n136it [01:30,  1.56it/s]\u001b[A\u001b[A\u001b[A\n\n\n137it [01:31,  1.56it/s]\u001b[A\u001b[A\u001b[A\n\n\n138it [01:31,  1.55it/s]\u001b[A\u001b[A\u001b[A\n\n\n139it [01:32,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n140it [01:33,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n141it [01:33,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n142it [01:34,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n143it [01:35,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n144it [01:36,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n145it [01:36,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n146it [01:37,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n147it [01:37,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n148it [01:38,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n149it [01:39,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n150it [01:40,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n151it [01:40,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n152it [01:41,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n153it [01:42,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n154it [01:42,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n155it [01:43,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n156it [01:44,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n157it [01:44,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n158it [01:45,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n159it [01:46,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n160it [01:46,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n161it [01:47,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n162it [01:48,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n163it [01:48,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n164it [01:49,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n165it [01:50,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n166it [01:50,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n167it [01:51,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n168it [01:51,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n169it [01:52,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n170it [01:53,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n171it [01:53,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n172it [01:54,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n173it [01:55,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n174it [01:56,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n175it [01:56,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n176it [01:57,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n177it [01:58,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n178it [01:58,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n179it [01:59,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n180it [01:59,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n181it [02:00,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n182it [02:01,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n183it [02:01,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n184it [02:02,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n185it [02:03,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n186it [02:03,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n187it [02:04,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n188it [02:05,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n189it [02:05,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n190it [02:06,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n191it [02:07,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n192it [02:07,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n193it [02:08,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n194it [02:09,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n195it [02:09,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n196it [02:10,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n197it [02:11,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n198it [02:11,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n199it [02:12,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n200it [02:13,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n201it [02:13,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n202it [02:14,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n203it [02:15,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n204it [02:16,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n205it [02:16,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n206it [02:17,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n207it [02:18,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n208it [02:18,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n209it [02:19,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n210it [02:20,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n211it [02:20,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n212it [02:21,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n213it [02:22,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n214it [02:22,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n215it [02:23,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n216it [02:24,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n217it [02:24,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n218it [02:25,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n219it [02:26,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n220it [02:26,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n221it [02:27,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n222it [02:28,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n223it [02:28,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n224it [02:29,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n225it [02:30,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n226it [02:30,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n227it [02:31,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n228it [02:32,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n229it [02:32,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n","name":"stderr"},{"output_type":"stream","text":"230it [02:33,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n231it [02:34,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n232it [02:34,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n233it [02:35,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n234it [02:36,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n235it [02:36,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n236it [02:37,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n237it [02:38,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n238it [02:38,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n239it [02:39,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n240it [02:40,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n241it [02:41,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n242it [02:41,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n243it [02:42,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n244it [02:43,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n245it [02:43,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n246it [02:44,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n247it [02:45,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n248it [02:45,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n249it [02:46,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n250it [02:47,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n251it [02:47,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n252it [02:48,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n253it [02:49,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n254it [02:49,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n255it [02:50,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n256it [02:51,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n257it [02:51,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n258it [02:52,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n259it [02:53,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n260it [02:53,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n261it [02:54,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n262it [02:55,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n263it [02:55,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n264it [02:56,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n265it [02:57,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n266it [02:57,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n267it [02:58,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n268it [02:59,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n269it [02:59,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n270it [03:00,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n271it [03:01,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n272it [03:01,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n273it [03:02,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n274it [03:03,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n275it [03:03,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n276it [03:04,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n277it [03:05,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n278it [03:05,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n279it [03:06,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n280it [03:07,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n281it [03:07,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n282it [03:08,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n283it [03:09,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n284it [03:09,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n285it [03:10,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n286it [03:11,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n287it [03:11,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n288it [03:12,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n289it [03:13,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n290it [03:13,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n291it [03:14,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n292it [03:15,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n293it [03:15,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n294it [03:16,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n295it [03:17,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n296it [03:17,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n297it [03:18,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n298it [03:19,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n299it [03:19,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n300it [03:20,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n301it [03:21,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n302it [03:21,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n303it [03:22,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n304it [03:23,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n305it [03:23,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n306it [03:24,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n307it [03:25,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n308it [03:25,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n309it [03:26,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n310it [03:27,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n311it [03:27,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n312it [03:28,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n313it [03:29,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n314it [03:30,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n315it [03:30,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n316it [03:31,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n317it [03:32,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n318it [03:32,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n319it [03:33,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n320it [03:34,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n321it [03:34,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n322it [03:35,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n323it [03:36,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n324it [03:36,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n325it [03:37,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n326it [03:38,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n327it [03:38,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n328it [03:39,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n329it [03:40,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n330it [03:40,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n331it [03:41,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n332it [03:42,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n333it [03:42,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n334it [03:43,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n335it [03:44,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n336it [03:44,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n337it [03:45,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n338it [03:46,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n339it [03:46,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n340it [03:47,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n341it [03:48,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n342it [03:49,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n343it [03:49,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n344it [03:50,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n345it [03:51,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n346it [03:51,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n347it [03:52,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n348it [03:53,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n349it [03:54,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n350it [03:54,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n351it [03:55,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n352it [03:56,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n353it [03:57,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n354it [03:57,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n355it [03:58,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n356it [03:59,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n357it [03:59,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n358it [04:00,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n359it [04:01,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n360it [04:01,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n361it [04:02,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n362it [04:03,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n363it [04:03,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n364it [04:04,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n365it [04:05,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n366it [04:05,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n367it [04:06,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n368it [04:07,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n369it [04:07,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n370it [04:08,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n371it [04:09,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n372it [04:09,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n373it [04:10,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n374it [04:11,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n375it [04:12,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n376it [04:12,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n377it [04:13,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n378it [04:14,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n379it [04:14,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n380it [04:15,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n381it [04:16,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n382it [04:16,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n383it [04:17,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n384it [04:18,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n385it [04:18,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n386it [04:19,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n387it [04:20,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n388it [04:20,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n389it [04:21,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n390it [04:22,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n391it [04:22,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n392it [04:23,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n393it [04:24,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n394it [04:24,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n395it [04:25,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n396it [04:26,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n397it [04:26,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n398it [04:27,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n399it [04:28,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n400it [04:28,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n401it [04:29,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n402it [04:30,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n403it [04:30,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n404it [04:31,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n405it [04:32,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n406it [04:32,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n407it [04:33,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n408it [04:34,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n409it [04:34,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n410it [04:35,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n411it [04:36,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n412it [04:36,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n413it [04:37,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n414it [04:38,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n415it [04:38,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n416it [04:39,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n417it [04:40,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n418it [04:41,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n419it [04:41,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n420it [04:42,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n421it [04:43,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n422it [04:43,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n423it [04:44,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n424it [04:45,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n425it [04:46,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n426it [04:46,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n427it [04:47,  1.31it/s]\u001b[A\u001b[A\u001b[A\n\n\n428it [04:48,  1.33it/s]\u001b[A\u001b[A\u001b[A\n\n\n429it [04:49,  1.33it/s]\u001b[A\u001b[A\u001b[A\n\n\n430it [04:49,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n431it [04:50,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n432it [04:51,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n433it [04:51,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n434it [04:52,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n435it [04:53,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n436it [04:53,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n437it [04:54,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n438it [04:55,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n439it [04:56,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n440it [04:56,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n441it [04:57,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n442it [04:57,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n443it [04:58,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n444it [04:59,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n445it [05:00,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n446it [05:00,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n447it [05:01,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n448it [05:01,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n449it [05:02,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n450it [05:03,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n451it [05:04,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n452it [05:04,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n453it [05:05,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n454it [05:06,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n455it [05:06,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n456it [05:07,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n","name":"stderr"},{"output_type":"stream","text":"457it [05:08,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n458it [05:08,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n459it [05:09,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n460it [05:10,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n461it [05:10,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n462it [05:11,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n463it [05:12,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n464it [05:13,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n465it [05:13,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n466it [05:14,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n467it [05:15,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n468it [05:16,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n469it [05:16,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n470it [05:17,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n471it [05:18,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n472it [05:18,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n473it [05:19,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n474it [05:20,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n475it [05:20,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n476it [05:21,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n477it [05:22,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n478it [05:22,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n479it [05:23,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n480it [05:24,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n481it [05:24,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n482it [05:25,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n483it [05:26,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n484it [05:26,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n485it [05:27,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n486it [05:28,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n487it [05:28,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n488it [05:29,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n489it [05:30,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n490it [05:30,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n491it [05:31,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n492it [05:32,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n493it [05:32,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n494it [05:33,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n495it [05:34,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n496it [05:34,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n497it [05:35,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n498it [05:36,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n499it [05:37,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n500it [05:37,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n501it [05:38,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n502it [05:39,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n503it [05:39,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n504it [05:40,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n505it [05:41,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n506it [05:41,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n507it [05:42,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n508it [05:43,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n509it [05:43,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n510it [05:44,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n511it [05:45,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n512it [05:45,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n513it [05:46,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n514it [05:47,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n515it [05:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n516it [05:48,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n517it [05:49,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n518it [05:49,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n519it [05:50,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n520it [05:51,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n521it [05:51,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n522it [05:52,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n523it [05:53,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n524it [05:53,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n525it [05:54,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n526it [05:55,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n527it [05:55,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n528it [05:56,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n529it [05:57,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n530it [05:57,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n531it [05:58,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n532it [05:59,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n533it [05:59,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n534it [06:00,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n535it [06:01,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n536it [06:01,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n537it [06:02,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n538it [06:03,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n539it [06:03,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n540it [06:04,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n541it [06:05,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n542it [06:05,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n543it [06:06,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n544it [06:07,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n545it [06:07,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n546it [06:08,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n547it [06:09,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n548it [06:09,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n549it [06:10,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n550it [06:11,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n551it [06:11,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n552it [06:12,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n553it [06:13,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n554it [06:14,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n555it [06:14,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n556it [06:15,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n557it [06:16,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n558it [06:16,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n559it [06:17,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n560it [06:18,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n561it [06:18,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n562it [06:19,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n563it [06:20,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n564it [06:20,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n565it [06:21,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n566it [06:22,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n567it [06:22,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n568it [06:23,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n569it [06:24,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n570it [06:24,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n571it [06:25,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n572it [06:26,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n573it [06:26,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n574it [06:27,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n575it [06:28,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n576it [06:28,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n577it [06:29,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n578it [06:30,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n579it [06:30,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n580it [06:31,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n581it [06:32,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n582it [06:32,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n583it [06:33,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n584it [06:34,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n585it [06:34,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n586it [06:35,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n587it [06:36,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n588it [06:36,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n589it [06:37,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n590it [06:38,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n591it [06:38,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n592it [06:39,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n593it [06:40,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n594it [06:40,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n595it [06:41,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n596it [06:42,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n597it [06:42,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n598it [06:43,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n599it [06:44,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n600it [06:44,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n601it [06:45,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n602it [06:46,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n603it [06:46,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n604it [06:47,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n605it [06:48,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n606it [06:49,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n607it [06:49,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n608it [06:50,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n609it [06:50,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n610it [06:51,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n611it [06:52,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n612it [06:53,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n613it [06:53,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n614it [06:54,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n615it [06:55,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n616it [06:55,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n617it [06:56,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n618it [06:57,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n619it [06:57,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n620it [06:58,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n621it [06:59,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n622it [06:59,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n623it [07:00,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n624it [07:01,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n625it [07:01,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n626it [07:02,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n627it [07:03,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n628it [07:03,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n629it [07:04,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n630it [07:05,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n631it [07:05,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n632it [07:06,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n633it [07:07,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n634it [07:07,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n635it [07:08,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n636it [07:09,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n637it [07:09,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n638it [07:10,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n639it [07:11,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n640it [07:11,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n641it [07:12,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n642it [07:13,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n643it [07:13,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n644it [07:14,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n645it [07:15,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n646it [07:15,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n647it [07:16,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n648it [07:17,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n649it [07:17,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n650it [07:18,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n651it [07:19,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n652it [07:19,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n653it [07:20,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n654it [07:21,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n655it [07:21,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n656it [07:22,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n657it [07:23,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n658it [07:23,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n659it [07:24,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n660it [07:25,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n661it [07:26,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n662it [07:26,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n663it [07:27,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n664it [07:28,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n665it [07:28,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n666it [07:29,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n667it [07:30,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n668it [07:30,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n669it [07:31,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n670it [07:32,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n671it [07:32,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n672it [07:33,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n673it [07:34,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n674it [07:34,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n675it [07:35,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n676it [07:36,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n677it [07:36,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n678it [07:37,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n679it [07:38,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n680it [07:39,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n681it [07:39,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n682it [07:40,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n683it [07:41,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n","name":"stderr"},{"output_type":"stream","text":"684it [07:41,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n685it [07:42,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n686it [07:43,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n687it [07:43,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n688it [07:44,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n689it [07:45,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n690it [07:45,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n691it [07:46,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n692it [07:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n693it [07:47,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n694it [07:48,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n695it [07:49,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n696it [07:49,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n697it [07:50,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n698it [07:51,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n699it [07:51,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n700it [07:52,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n701it [07:53,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n702it [07:54,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n703it [07:54,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n704it [07:55,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n705it [07:56,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n706it [07:56,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n707it [07:57,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n708it [07:58,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n709it [07:58,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n710it [07:59,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n711it [08:00,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n712it [08:00,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n713it [08:01,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n714it [08:02,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n715it [08:02,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n716it [08:03,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n717it [08:04,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n718it [08:05,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n719it [08:05,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n720it [08:06,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n721it [08:07,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n722it [08:07,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n723it [08:08,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n724it [08:09,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n725it [08:09,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n726it [08:10,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n727it [08:11,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n728it [08:11,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n729it [08:12,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n730it [08:13,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n731it [08:13,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n732it [08:14,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n733it [08:15,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n734it [08:15,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n735it [08:16,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n736it [08:17,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n737it [08:17,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n738it [08:18,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n739it [08:19,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n740it [08:19,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n741it [08:20,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n742it [08:21,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n743it [08:21,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n744it [08:22,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n745it [08:23,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n746it [08:23,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n747it [08:24,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n748it [08:25,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n749it [08:26,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n750it [08:26,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n751it [08:27,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n752it [08:28,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n753it [08:28,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n754it [08:29,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n755it [08:30,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n756it [08:30,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n757it [08:31,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n758it [08:32,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n759it [08:32,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n760it [08:33,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n761it [08:34,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n762it [08:34,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n763it [08:35,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n764it [08:36,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n765it [08:37,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n766it [08:37,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n767it [08:38,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n768it [08:39,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n769it [08:39,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n770it [08:40,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n771it [08:41,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n772it [08:41,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n773it [08:42,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n774it [08:43,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n775it [08:43,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n776it [08:44,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n777it [08:45,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n778it [08:45,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n779it [08:46,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n780it [08:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n781it [08:47,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n782it [08:48,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n783it [08:49,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n784it [08:49,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n785it [08:50,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n786it [08:51,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n787it [08:51,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n788it [08:52,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n789it [08:53,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n790it [08:54,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n791it [08:54,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n792it [08:55,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n793it [08:56,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n794it [08:56,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n795it [08:57,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n796it [08:58,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n797it [08:58,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n798it [08:59,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n799it [09:00,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n800it [09:00,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n801it [09:01,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n802it [09:02,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n803it [09:03,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n804it [09:03,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n805it [09:04,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n806it [09:05,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n807it [09:05,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n808it [09:06,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n809it [09:07,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n810it [09:07,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n811it [09:08,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n812it [09:09,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n813it [09:09,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n814it [09:10,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n815it [09:11,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n816it [09:11,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n817it [09:12,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n818it [09:13,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n819it [09:13,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n820it [09:14,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n821it [09:15,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n822it [09:15,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n823it [09:16,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n824it [09:17,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n825it [09:17,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n826it [09:18,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n827it [09:19,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n828it [09:19,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n829it [09:20,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n830it [09:21,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n831it [09:21,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n832it [09:22,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n833it [09:23,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n834it [09:23,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n835it [09:24,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n836it [09:25,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n837it [09:25,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n838it [09:26,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n839it [09:27,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n840it [09:27,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n841it [09:28,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n842it [09:29,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n843it [09:29,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n844it [09:30,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n845it [09:31,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n846it [09:31,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n847it [09:32,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n848it [09:33,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n849it [09:33,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n850it [09:34,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n851it [09:35,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n852it [09:35,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n853it [09:36,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n854it [09:37,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n855it [09:37,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n856it [09:38,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n857it [09:39,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n858it [09:39,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n859it [09:40,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n860it [09:41,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n861it [09:41,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n862it [09:42,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n863it [09:43,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n864it [09:43,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n865it [09:44,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n866it [09:45,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n867it [09:46,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n868it [09:46,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n869it [09:47,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n870it [09:48,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n871it [09:48,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n872it [09:49,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n873it [09:50,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n874it [09:50,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n875it [09:51,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n876it [09:52,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n877it [09:52,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n878it [09:53,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n879it [09:54,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n880it [09:54,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n881it [09:55,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n882it [09:56,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n883it [09:56,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n884it [09:57,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n885it [09:58,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n886it [09:59,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n887it [09:59,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n888it [10:00,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n889it [10:01,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n890it [10:01,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n891it [10:02,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n892it [10:03,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n893it [10:03,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n894it [10:04,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n895it [10:05,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n896it [10:05,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n897it [10:06,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n898it [10:07,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n899it [10:07,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n900it [10:08,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n901it [10:09,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n902it [10:09,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n903it [10:10,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n904it [10:11,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n905it [10:11,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n906it [10:12,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n907it [10:13,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n908it [10:14,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n909it [10:14,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n910it [10:15,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n","name":"stderr"},{"output_type":"stream","text":"911it [10:16,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n912it [10:16,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n913it [10:17,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n914it [10:18,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n915it [10:18,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n916it [10:19,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n917it [10:20,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n918it [10:20,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n919it [10:21,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n920it [10:22,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n921it [10:22,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n922it [10:23,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n923it [10:24,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n924it [10:24,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n925it [10:25,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n926it [10:26,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n927it [10:26,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n928it [10:27,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n929it [10:28,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n930it [10:28,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n931it [10:29,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n932it [10:30,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n933it [10:30,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n934it [10:31,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n935it [10:32,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n936it [10:32,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n937it [10:33,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n938it [10:34,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n939it [10:34,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n940it [10:35,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n941it [10:36,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n942it [10:36,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n943it [10:37,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n944it [10:38,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n945it [10:38,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n946it [10:39,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n947it [10:40,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n948it [10:40,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n949it [10:41,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n950it [10:42,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n951it [10:42,  1.55it/s]\u001b[A\u001b[A\u001b[A\n\n\n952it [10:43,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n953it [10:44,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n954it [10:44,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n955it [10:45,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n956it [10:46,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n957it [10:46,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n958it [10:47,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n959it [10:48,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n960it [10:48,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n961it [10:49,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n962it [10:50,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n963it [10:50,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n964it [10:51,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n965it [10:52,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n966it [10:52,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n967it [10:53,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n968it [10:54,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n969it [10:54,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n970it [10:55,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n971it [10:56,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n972it [10:56,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n973it [10:57,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n974it [10:58,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n975it [10:58,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n976it [10:59,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n977it [11:00,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n978it [11:00,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n979it [11:01,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n980it [11:02,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n981it [11:02,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n982it [11:03,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n983it [11:04,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n984it [11:04,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n985it [11:05,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n986it [11:06,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n987it [11:06,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n988it [11:07,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n989it [11:08,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n990it [11:08,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n991it [11:09,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n992it [11:10,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n993it [11:10,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n994it [11:11,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n995it [11:12,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n996it [11:12,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n997it [11:13,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n998it [11:14,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n999it [11:14,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1000it [11:15,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1001it [11:16,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1002it [11:16,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1003it [11:17,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1004it [11:18,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1005it [11:18,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1006it [11:19,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1007it [11:20,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1008it [11:20,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1009it [11:21,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1010it [11:22,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1011it [11:22,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1012it [11:23,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1013it [11:24,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1014it [11:24,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1015it [11:25,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1016it [11:26,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1017it [11:26,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1018it [11:27,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1019it [11:28,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1020it [11:28,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1021it [11:29,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1022it [11:30,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1023it [11:31,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1024it [11:31,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1025it [11:32,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1026it [11:33,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1027it [11:33,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1028it [11:34,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1029it [11:34,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1030it [11:35,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1031it [11:36,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1032it [11:36,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1033it [11:37,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1034it [11:38,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1035it [11:38,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1036it [11:39,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1037it [11:40,  1.55it/s]\u001b[A\u001b[A\u001b[A\n\n\n1038it [11:40,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n1039it [11:41,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1040it [11:42,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1041it [11:42,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1042it [11:43,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1043it [11:44,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1044it [11:44,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1045it [11:45,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1046it [11:46,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1047it [11:46,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1048it [11:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1049it [11:48,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1050it [11:48,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1051it [11:49,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1052it [11:50,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1053it [11:51,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1054it [11:51,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1055it [11:52,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1056it [11:53,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1057it [11:53,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1058it [11:54,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1059it [11:55,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1060it [11:55,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1061it [11:56,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1062it [11:57,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1063it [11:57,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1064it [11:58,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1065it [11:59,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1066it [11:59,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1067it [12:00,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1068it [12:01,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1069it [12:01,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1070it [12:02,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1071it [12:03,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1072it [12:03,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1073it [12:04,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1074it [12:05,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1075it [12:05,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1076it [12:06,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1077it [12:07,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1078it [12:08,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1079it [12:08,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1080it [12:09,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1081it [12:10,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1082it [12:10,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1083it [12:11,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n1084it [12:11,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1085it [12:12,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n1086it [12:13,  1.55it/s]\u001b[A\u001b[A\u001b[A\n\n\n1087it [12:13,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n1088it [12:14,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n1089it [12:15,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1090it [12:15,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1091it [12:16,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1092it [12:17,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1093it [12:17,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1094it [12:18,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1095it [12:19,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1096it [12:19,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1097it [12:20,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1098it [12:21,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1099it [12:21,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1100it [12:22,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1101it [12:23,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1102it [12:23,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1103it [12:24,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1104it [12:25,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1105it [12:26,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1106it [12:26,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1107it [12:27,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1108it [12:27,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1109it [12:28,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1110it [12:29,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1111it [12:29,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1112it [12:30,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1113it [12:31,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1114it [12:31,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1115it [12:32,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1116it [12:33,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1117it [12:33,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1118it [12:34,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1119it [12:35,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1120it [12:36,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1121it [12:36,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1122it [12:37,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1123it [12:38,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1124it [12:38,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1125it [12:39,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1126it [12:40,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n1127it [12:41,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1128it [12:41,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1129it [12:42,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1130it [12:43,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1131it [12:43,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1132it [12:44,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1133it [12:45,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n","name":"stderr"},{"output_type":"stream","text":"1134it [12:45,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1135it [12:46,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1136it [12:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1137it [12:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1138it [12:48,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1139it [12:49,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1140it [12:49,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1141it [12:50,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1142it [12:51,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1143it [12:51,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1144it [12:52,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1145it [12:53,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1146it [12:53,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1147it [12:54,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1148it [12:55,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1149it [12:55,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1150it [12:56,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1151it [12:57,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1152it [12:57,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1153it [12:58,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1154it [12:59,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1155it [12:59,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1156it [13:00,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1157it [13:01,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1158it [13:01,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1159it [13:02,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1160it [13:03,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1161it [13:03,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1162it [13:04,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1163it [13:05,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1164it [13:05,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1165it [13:06,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1166it [13:07,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1167it [13:07,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1168it [13:08,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1169it [13:09,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1170it [13:09,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1171it [13:10,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1172it [13:11,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1173it [13:11,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1174it [13:12,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1175it [13:13,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1176it [13:13,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1177it [13:14,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1178it [13:15,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1179it [13:16,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n1180it [13:16,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n1181it [13:17,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n1182it [13:18,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n1183it [13:19,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n1184it [13:19,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n1185it [13:20,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n1186it [13:21,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n1187it [13:21,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1188it [13:22,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1189it [13:23,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1190it [13:23,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1191it [13:24,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1192it [13:25,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1193it [13:25,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1194it [13:26,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1195it [13:27,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1196it [13:27,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1197it [13:28,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1198it [13:29,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1199it [13:29,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1200it [13:30,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1201it [13:31,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1202it [13:31,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1203it [13:32,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1204it [13:33,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1205it [13:33,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1206it [13:34,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1207it [13:35,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n1208it [13:35,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1209it [13:36,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1210it [13:37,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1211it [13:37,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1212it [13:38,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1213it [13:39,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1214it [13:39,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1215it [13:40,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1216it [13:41,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1217it [13:41,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1218it [13:42,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1219it [13:43,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1220it [13:43,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1221it [13:44,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1222it [13:45,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1223it [13:45,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1224it [13:46,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1225it [13:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1226it [13:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1227it [13:48,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1228it [13:49,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1229it [13:49,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1230it [13:50,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1231it [13:51,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1232it [13:51,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1233it [13:52,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1234it [13:53,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1235it [13:53,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1236it [13:54,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1237it [13:55,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1238it [13:55,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1239it [13:56,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1240it [13:57,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1241it [13:57,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1242it [13:58,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1243it [13:59,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1244it [14:00,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1245it [14:00,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1246it [14:01,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1247it [14:02,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1248it [14:02,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1249it [14:03,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1250it [14:04,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1251it [14:04,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1252it [14:05,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1253it [14:06,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1254it [14:06,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1255it [14:07,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1256it [14:08,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1257it [14:08,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1258it [14:09,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1259it [14:10,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1260it [14:11,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1261it [14:11,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1262it [14:12,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1263it [14:13,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1264it [14:13,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1265it [14:14,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1266it [14:15,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n1267it [14:16,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n1268it [14:16,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n1269it [14:17,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n1270it [14:18,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n1271it [14:18,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n1272it [14:19,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n1273it [14:20,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1274it [14:21,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n1275it [14:21,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1276it [14:22,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1277it [14:23,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1278it [14:23,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1279it [14:24,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1280it [14:25,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1281it [14:25,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1282it [14:26,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1283it [14:27,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1284it [14:27,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1285it [14:28,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1286it [14:29,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1287it [14:29,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1288it [14:30,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1289it [14:31,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1290it [14:31,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1291it [14:32,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1292it [14:33,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1293it [14:33,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1294it [14:34,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1295it [14:35,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1296it [14:35,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1297it [14:36,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1298it [14:37,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1299it [14:37,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1300it [14:38,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1301it [14:39,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1302it [14:39,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1303it [14:40,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1304it [14:41,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1305it [14:41,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1306it [14:42,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1307it [14:43,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1308it [14:44,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1309it [14:44,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1310it [14:45,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1311it [14:46,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1312it [14:46,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1313it [14:47,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1314it [14:48,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1315it [14:48,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1316it [14:49,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1317it [14:50,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1318it [14:50,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1319it [14:51,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1320it [14:52,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1321it [14:52,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1322it [14:53,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1323it [14:54,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1324it [14:55,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1325it [14:55,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1326it [14:56,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1327it [14:57,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1328it [14:57,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1329it [14:58,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1330it [14:59,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1331it [14:59,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1332it [15:00,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1333it [15:01,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1334it [15:01,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1335it [15:02,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1336it [15:03,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1337it [15:03,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1338it [15:04,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1339it [15:05,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1340it [15:05,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1341it [15:06,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1342it [15:07,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1343it [15:07,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1344it [15:08,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1345it [15:09,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1346it [15:09,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1347it [15:10,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1348it [15:11,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1349it [15:11,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1350it [15:12,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1351it [15:13,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1352it [15:13,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1353it [15:14,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1354it [15:15,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n","name":"stderr"},{"output_type":"stream","text":"1355it [15:15,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1356it [15:16,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1357it [15:17,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1358it [15:17,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1359it [15:18,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1360it [15:19,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1361it [15:20,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1362it [15:20,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1363it [15:21,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1364it [15:22,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1365it [15:22,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1366it [15:23,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1367it [15:24,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1368it [15:24,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1369it [15:25,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1370it [15:26,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1371it [15:26,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1372it [15:27,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1373it [15:28,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1374it [15:28,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1375it [15:29,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1376it [15:30,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1377it [15:30,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1378it [15:31,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1379it [15:32,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1380it [15:32,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1381it [15:33,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1382it [15:34,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1383it [15:34,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1384it [15:35,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1385it [15:36,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1386it [15:36,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1387it [15:37,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1388it [15:38,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1389it [15:38,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1390it [15:39,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1391it [15:40,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1392it [15:40,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1393it [15:41,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1394it [15:42,  1.42it/s]\u001b[A\u001b[A\u001b[A\n\n\n1395it [15:42,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1396it [15:43,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1397it [15:44,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1398it [15:45,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1399it [15:45,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1400it [15:46,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1401it [15:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1402it [15:47,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1403it [15:48,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1404it [15:49,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1405it [15:49,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1406it [15:50,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1407it [15:51,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1408it [15:51,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1409it [15:52,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1410it [15:53,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1411it [15:53,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1412it [15:54,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1413it [15:55,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1414it [15:55,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1415it [15:56,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1416it [15:57,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1417it [15:57,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1418it [15:58,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1419it [15:59,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1420it [15:59,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1421it [16:00,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1422it [16:01,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1423it [16:01,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1424it [16:02,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1425it [16:03,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1426it [16:03,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1427it [16:04,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1428it [16:05,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1429it [16:05,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1430it [16:06,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1431it [16:07,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1432it [16:07,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1433it [16:08,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1434it [16:09,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1435it [16:09,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1436it [16:10,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1437it [16:11,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1438it [16:11,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1439it [16:12,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1440it [16:13,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1441it [16:13,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1442it [16:14,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1443it [16:15,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1444it [16:16,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1445it [16:16,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1446it [16:17,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1447it [16:18,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1448it [16:18,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1449it [16:19,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1450it [16:20,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1451it [16:20,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1452it [16:21,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1453it [16:22,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1454it [16:22,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1455it [16:23,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1456it [16:24,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1457it [16:24,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1458it [16:25,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1459it [16:26,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1460it [16:26,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1461it [16:27,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1462it [16:28,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1463it [16:28,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1464it [16:29,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1465it [16:30,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1466it [16:30,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1467it [16:31,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1468it [16:32,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1469it [16:32,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1470it [16:33,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1471it [16:34,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1472it [16:35,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1473it [16:35,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1474it [16:36,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1475it [16:37,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1476it [16:37,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1477it [16:38,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1478it [16:39,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1479it [16:39,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1480it [16:40,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1481it [16:41,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1482it [16:41,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1483it [16:42,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1484it [16:43,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1485it [16:43,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1486it [16:44,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1487it [16:45,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1488it [16:45,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1489it [16:46,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1490it [16:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1491it [16:47,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1492it [16:48,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1493it [16:49,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1494it [16:49,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1495it [16:50,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1496it [16:51,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1497it [16:52,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1498it [16:52,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n1499it [16:53,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n1500it [16:54,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n1501it [16:55,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n1502it [16:55,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n1503it [16:56,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n1504it [16:57,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n1505it [16:57,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n1506it [16:58,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1507it [16:59,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1508it [16:59,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1509it [17:00,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1510it [17:01,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1511it [17:01,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1512it [17:02,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1513it [17:03,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1514it [17:03,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1515it [17:04,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1516it [17:05,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1517it [17:06,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1518it [17:06,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1519it [17:07,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1520it [17:08,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1521it [17:08,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1522it [17:09,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1523it [17:10,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1524it [17:10,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1525it [17:11,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1526it [17:12,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1527it [17:12,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1528it [17:13,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1529it [17:14,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1530it [17:14,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1531it [17:15,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1532it [17:16,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1533it [17:16,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1534it [17:17,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1535it [17:18,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1536it [17:18,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1537it [17:19,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1538it [17:20,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1539it [17:20,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1540it [17:21,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1541it [17:22,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1542it [17:22,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1543it [17:23,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1544it [17:24,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1545it [17:24,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1546it [17:25,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1547it [17:26,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1548it [17:27,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1549it [17:27,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1550it [17:28,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1551it [17:29,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1552it [17:29,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1553it [17:30,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1554it [17:31,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1555it [17:31,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1556it [17:32,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1557it [17:33,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1558it [17:33,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1559it [17:34,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1560it [17:35,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1561it [17:35,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1562it [17:36,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1563it [17:37,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1564it [17:37,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1565it [17:38,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1566it [17:39,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1567it [17:39,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1568it [17:40,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1569it [17:41,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1570it [17:42,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1571it [17:42,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1572it [17:43,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1573it [17:44,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1574it [17:44,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1575it [17:45,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n","name":"stderr"},{"output_type":"stream","text":"1576it [17:46,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1577it [17:46,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1578it [17:47,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1579it [17:48,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1580it [17:48,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1581it [17:49,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1582it [17:50,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1583it [17:51,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1584it [17:51,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1585it [17:52,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n1586it [17:53,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n1587it [17:53,  1.43it/s]\u001b[A\u001b[A\u001b[A\n\n\n1588it [17:54,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1589it [17:55,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1590it [17:55,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1591it [17:56,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1592it [17:57,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1593it [17:57,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1594it [17:58,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1595it [17:59,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1596it [17:59,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1597it [18:00,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1598it [18:01,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1599it [18:01,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1600it [18:02,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1601it [18:03,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1602it [18:04,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1603it [18:04,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1604it [18:05,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1605it [18:06,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1606it [18:06,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1607it [18:07,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1608it [18:08,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1609it [18:08,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1610it [18:09,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1611it [18:10,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1612it [18:10,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1613it [18:11,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1614it [18:12,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1615it [18:12,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1616it [18:13,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1617it [18:14,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1618it [18:14,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1619it [18:15,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1620it [18:16,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1621it [18:17,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1622it [18:17,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1623it [18:18,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1624it [18:19,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1625it [18:19,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1626it [18:20,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1627it [18:21,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1628it [18:21,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1629it [18:22,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1630it [18:23,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1631it [18:23,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1632it [18:24,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1633it [18:25,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1634it [18:25,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1635it [18:26,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1636it [18:27,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1637it [18:27,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1638it [18:28,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1639it [18:29,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1640it [18:29,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1641it [18:30,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1642it [18:31,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1643it [18:31,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1644it [18:32,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1645it [18:33,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1646it [18:33,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1647it [18:34,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1648it [18:35,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1649it [18:35,  1.54it/s]\u001b[A\u001b[A\u001b[A\n\n\n1650it [18:36,  1.55it/s]\u001b[A\u001b[A\u001b[A\n\n\n1651it [18:37,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n1652it [18:37,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1653it [18:38,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1654it [18:39,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1655it [18:39,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1656it [18:40,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n1657it [18:41,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n1658it [18:41,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n1659it [18:42,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1660it [18:43,  1.53it/s]\u001b[A\u001b[A\u001b[A\n\n\n1661it [18:43,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1662it [18:44,  1.52it/s]\u001b[A\u001b[A\u001b[A\n\n\n1663it [18:45,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1664it [18:45,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1665it [18:46,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1666it [18:47,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1667it [18:47,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1668it [18:48,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1669it [18:49,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1670it [18:49,  1.50it/s]\u001b[A\u001b[A\u001b[A\n\n\n1671it [18:50,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1672it [18:51,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1673it [18:51,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1674it [18:52,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1675it [18:53,  1.49it/s]\u001b[A\u001b[A\u001b[A\n\n\n1676it [18:53,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1677it [18:54,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1678it [18:55,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1679it [18:55,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1680it [18:56,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1681it [18:57,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1682it [18:58,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1683it [18:58,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1684it [18:59,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1685it [19:00,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1686it [19:00,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1687it [19:01,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1688it [19:02,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1689it [19:02,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1690it [19:03,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1691it [19:04,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1692it [19:04,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1693it [19:05,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1694it [19:06,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1695it [19:06,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1696it [19:07,  1.45it/s]\u001b[A\u001b[A\u001b[A\n\n\n1697it [19:08,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1698it [19:08,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1699it [19:09,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1700it [19:10,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1701it [19:10,  1.48it/s]\u001b[A\u001b[A\u001b[A\n\n\n1702it [19:11,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1703it [19:12,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1704it [19:13,  1.46it/s]\u001b[A\u001b[A\u001b[A\n\n\n1705it [19:13,  1.47it/s]\u001b[A\u001b[A\u001b[A\n\n\n1706it [19:14,  1.44it/s]\u001b[A\u001b[A\u001b[A\n\n\n1707it [19:14,  1.51it/s]\u001b[A\u001b[A\u001b[A\n\n\n0it [00:00, ?it/s]\u001b[A\u001b[A\u001b[A\n\n\n1it [00:00,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n2it [00:01,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n3it [00:02,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n4it [00:02,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n5it [00:03,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n6it [00:04,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n7it [00:05,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n8it [00:05,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n9it [00:06,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n10it [00:07,  1.33it/s]\u001b[A\u001b[A\u001b[A\n\n\n11it [00:08,  1.32it/s]\u001b[A\u001b[A\u001b[A\n\n\n12it [00:08,  1.31it/s]\u001b[A\u001b[A\u001b[A\n\n\n13it [00:09,  1.31it/s]\u001b[A\u001b[A\u001b[A\n\n\n14it [00:10,  1.31it/s]\u001b[A\u001b[A\u001b[A\n\n\n15it [00:11,  1.31it/s]\u001b[A\u001b[A\u001b[A\n\n\n16it [00:11,  1.30it/s]\u001b[A\u001b[A\u001b[A\n\n\n17it [00:12,  1.31it/s]\u001b[A\u001b[A\u001b[A\n\n\n18it [00:13,  1.31it/s]\u001b[A\u001b[A\u001b[A\n\n\n19it [00:14,  1.29it/s]\u001b[A\u001b[A\u001b[A\n\n\n20it [00:15,  1.29it/s]\u001b[A\u001b[A\u001b[A\n\n\n21it [00:15,  1.29it/s]\u001b[A\u001b[A\u001b[A\n\n\n22it [00:16,  1.30it/s]\u001b[A\u001b[A\u001b[A\n\n\n23it [00:17,  1.29it/s]\u001b[A\u001b[A\u001b[A\n\n\n24it [00:18,  1.32it/s]\u001b[A\u001b[A\u001b[A\n\n\n25it [00:18,  1.32it/s]\u001b[A\u001b[A\u001b[A\n\n\n26it [00:19,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n27it [00:20,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n28it [00:21,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n29it [00:21,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n30it [00:22,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n31it [00:23,  1.33it/s]\u001b[A\u001b[A\u001b[A\n\n\n32it [00:23,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n33it [00:24,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n34it [00:25,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n35it [00:26,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n36it [00:26,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n37it [00:27,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n38it [00:28,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n39it [00:29,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n40it [00:29,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n41it [00:30,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n42it [00:31,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n43it [00:32,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n44it [00:32,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n45it [00:33,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n46it [00:34,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n47it [00:34,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n48it [00:35,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n49it [00:36,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n50it [00:37,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n51it [00:37,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n52it [00:38,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n53it [00:39,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n54it [00:40,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n55it [00:40,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n56it [00:41,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n57it [00:42,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n58it [00:43,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n59it [00:43,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n60it [00:44,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n61it [00:45,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n62it [00:45,  1.40it/s]\u001b[A\u001b[A\u001b[A\n\n\n63it [00:46,  1.41it/s]\u001b[A\u001b[A\u001b[A\n\n\n64it [00:47,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n65it [00:48,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n66it [00:48,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n67it [00:49,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n68it [00:50,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n69it [00:51,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n70it [00:51,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n71it [00:52,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n72it [00:53,  1.33it/s]\u001b[A\u001b[A\u001b[A\n\n\n73it [00:54,  1.33it/s]\u001b[A\u001b[A\u001b[A\n\n\n74it [00:54,  1.33it/s]\u001b[A\u001b[A\u001b[A\n\n\n75it [00:55,  1.31it/s]\u001b[A\u001b[A\u001b[A\n\n\n76it [00:56,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n77it [00:57,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n78it [00:57,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n79it [00:58,  1.32it/s]\u001b[A\u001b[A\u001b[A\n\n\n80it [00:59,  1.33it/s]\u001b[A\u001b[A\u001b[A\n\n\n81it [01:00,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n82it [01:00,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n83it [01:01,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n84it [01:02,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n85it [01:03,  1.33it/s]\u001b[A\u001b[A\u001b[A\n\n\n86it [01:03,  1.34it/s]\u001b[A\u001b[A\u001b[A\n\n\n87it [01:04,  1.32it/s]\u001b[A\u001b[A\u001b[A\n\n\n88it [01:05,  1.32it/s]\u001b[A\u001b[A\u001b[A\n\n\n89it [01:06,  1.32it/s]\u001b[A\u001b[A\u001b[A\n\n\n90it [01:06,  1.33it/s]\u001b[A\u001b[A\u001b[A\n\n\n91it [01:07,  1.32it/s]\u001b[A\u001b[A\u001b[A\n\n\n92it [01:08,  1.32it/s]\u001b[A\u001b[A\u001b[A\n\n\n93it [01:09,  1.29it/s]\u001b[A\u001b[A\u001b[A\n\n\n94it [01:09,  1.29it/s]\u001b[A\u001b[A\u001b[A\n\n\n","name":"stderr"},{"output_type":"stream","text":"95it [01:10,  1.27it/s]\u001b[A\u001b[A\u001b[A\n\n\n96it [01:11,  1.28it/s]\u001b[A\u001b[A\u001b[A\n\n\n97it [01:12,  1.30it/s]\u001b[A\u001b[A\u001b[A\n\n\n98it [01:13,  1.27it/s]\u001b[A\u001b[A\u001b[A\n\n\n99it [01:13,  1.29it/s]\u001b[A\u001b[A\u001b[A\n\n\n100it [01:14,  1.29it/s]\u001b[A\u001b[A\u001b[A\n\n\n101it [01:15,  1.31it/s]\u001b[A\u001b[A\u001b[A\n\n\n102it [01:16,  1.28it/s]\u001b[A\u001b[A\u001b[A\n\n\n103it [01:16,  1.28it/s]\u001b[A\u001b[A\u001b[A\n\n\n104it [01:17,  1.29it/s]\u001b[A\u001b[A\u001b[A\n\n\n105it [01:18,  1.32it/s]\u001b[A\u001b[A\u001b[A\n\n\n106it [01:19,  1.35it/s]\u001b[A\u001b[A\u001b[A\n\n\n107it [01:19,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n108it [01:20,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n109it [01:21,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n110it [01:21,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n111it [01:22,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n112it [01:23,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n113it [01:24,  1.37it/s]\u001b[A\u001b[A\u001b[A\n\n\n114it [01:24,  1.39it/s]\u001b[A\u001b[A\u001b[A\n\n\n115it [01:25,  1.38it/s]\u001b[A\u001b[A\u001b[A\n\n\n116it [01:26,  1.36it/s]\u001b[A\u001b[A\u001b[A\n\n\n117it [01:26,  1.69it/s]\u001b[A\u001b[A\u001b[A","name":"stderr"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"del train_dataset, train_loader\ndel test_dataset, test_loader\ngc.collect()","execution_count":38,"outputs":[{"output_type":"execute_result","execution_count":38,"data":{"text/plain":"8954"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"from numpy.random import beta\nclass FocalLoss(nn.Module):\n    def __init__(self, alpha=0.5, gamma=2, logits=False, reduce=True):\n        super(FocalLoss, self).__init__()\n        self.alpha = alpha\n        self.gamma = gamma\n        self.logits = logits\n        self.reduce = reduce\n\n    def forward(self, inputs, targets):\n        if self.logits:\n            BCE_loss = F.binary_cross_entropy_with_logits(inputs, targets, reduce=False)\n        else:\n            BCE_loss = F.binary_cross_entropy(inputs, targets, reduce=False)\n        pt = torch.exp(-BCE_loss)\n        F_loss = self.alpha * (1-pt)**self.gamma * BCE_loss\n\n        if self.reduce:\n            return torch.sum(F_loss)\n        else:\n            return F_loss\n        \ndef mixup(input, target, gamma):\n    # target is onehot format!\n    perm = torch.randperm(input.size(0))\n    perm_input = input[perm]\n    perm_target = target[perm]\n    return input.mul_(gamma).add_(1 - gamma, perm_input), target.mul_(gamma).add_(1 - gamma, perm_target)\n\n    return mixed_x, mixed_y\n\nclass Trainer:\n    def __init__(self, \n                 model1,\n                 model2,\n                 logger,\n                 n_splits=5,\n                 seed=42,\n                 device=\"cuda:0\",\n                 train_batch=32,\n                 valid_batch=128,\n                 kwargs={}):\n        self.model1 = model1\n        self.model2 = model2\n        self.logger = logger\n        self.device = device\n        self.n_splits = n_splits\n        self.seed = seed\n        self.train_batch = train_batch\n        self.valid_batch = valid_batch\n        self.kwargs = kwargs\n        \n        self.best_score = None\n        self.tag = dt.datetime.now().strftime(\"%Y-%m-%d-%H-%M-%S\")\n        self.loss_fn = nn.BCELoss(reduction=\"mean\").to(self.device)\n        \n        path = Path(f\"bin1/{self.tag}\")\n        path.mkdir(exist_ok=True, parents=True)\n        self.path1 = path\n        path = Path(f\"bin2/{self.tag}\")\n        path.mkdir(exist_ok=True, parents=True)\n        self.path2 = path\n        \n    def fit(self, X, y, n_epochs=10):\n        train_preds1 = np.zeros((len(X), num_classes_c))\n        train_preds2 = np.zeros((len(X), num_classes_t))\n        fold = KFold(n_splits=self.n_splits, random_state=self.seed)\n        for i, (trn_idx, val_idx) in enumerate(fold.split(X)):\n            self.fold_num = i\n            self.logger.info(f\"Fold {i + 1}\")\n            X_train, X_val = X[trn_idx, :], X[val_idx, :]\n            y_train, y_val = y[trn_idx], y[val_idx]\n            \n            valid_preds1, valid_preds2 = self._fit(X_train, y_train, X_val, y_val, n_epochs)\n            #print('tp1 ' + str(train_preds1.shape[1]))\n            #print('vp1 ' + str(valid_preds1.shape[1]))\n            train_preds1[val_idx] = valid_preds1\n            train_preds2[val_idx] = valid_preds2\n        return train_preds1, train_preds2\n    \n    def _fit(self, X_train, y_train, X_val, y_val, n_epochs):\n        seed_torch(self.seed)\n        train_dataset = IMetDataset(X_train, labels=y_train, device=self.device)\n        train_loader = data.DataLoader(train_dataset, \n                                       batch_size=self.train_batch,\n                                       shuffle=True)\n\n        valid_dataset = IMetDataset(X_val, labels=y_val, device=self.device)\n        valid_loader = data.DataLoader(valid_dataset,\n                                       batch_size=self.valid_batch,\n                                       shuffle=False)\n        \n        model1 = self.model1(**self.kwargs)\n        model1.to(self.device)\n        \n        model2 = self.model2(**self.kwargs)\n        model2.to(self.device)\n        \n        optimizer1 = optim.Adam(params=model1.parameters(), \n                                lr=0.0001)\n        optimizer2 = optim.Adam(params=model2.parameters(), \n                                lr=0.0001)\n        scheduler1 = CosineAnnealingLR(optimizer1, T_max=n_epochs)\n        scheduler2 = CosineAnnealingLR(optimizer2, T_max=n_epochs)\n        best_score1 = np.inf\n        best_score2 = np.inf\n        mb = master_bar(range(n_epochs))\n        for epoch in mb:\n            model1.train()\n            model2.train()\n            avg_loss1 = 0.0\n            avg_loss2 = 0.0\n            for i_batch, y_batch in progress_bar(train_loader, parent=mb):\n                #i_batch, y_batch = mixup(i_batch, y_batch, beta(1.0, 1.0))\n                y_pred1 = model1(i_batch)\n                y_pred2 = model2(i_batch)\n                loss1 = self.loss_fn(y_pred1, y_batch[0])\n                loss2 = self.loss_fn(y_pred2, y_batch[1])\n                optimizer1.zero_grad()\n                optimizer2.zero_grad()\n                loss1.backward()\n                loss2.backward()\n                optimizer1.step()\n                optimizer2.step()\n                avg_loss1 += loss1.item() / len(train_loader)\n                avg_loss2 += loss2.item() / len(train_loader)\n            valid_preds1, avg_val_loss1, valid_preds2, avg_val_loss2 = self._val(valid_loader, model1, model2)\n            scheduler1.step()\n            scheduler2.step()\n\n            self.logger.info(\"=========================================\")\n            self.logger.info(f\"Epoch {epoch + 1} / {n_epochs}\")\n            self.logger.info(\"=========================================\")\n            self.logger.info(f\"avg_loss: {avg_loss1:.8f}\")\n            self.logger.info(f\"avg_val_loss: {avg_val_loss1:.8f}\")\n            self.logger.info(f\"avg_loss: {avg_loss2:.8f}\")\n            self.logger.info(f\"avg_val_loss: {avg_val_loss2:.8f}\")\n            \n            if best_score1 > avg_val_loss1:\n                torch.save(model1.state_dict(),\n                           self.path1 / f\"1best{self.fold_num}.pth\")\n                self.logger.info(f\"Save model at Epoch {epoch + 1}\")\n                best_score1 = avg_val_loss1\n                \n            if best_score2 > avg_val_loss2:\n                torch.save(model2.state_dict(),\n                           self.path2 / f\"2best{self.fold_num}.pth\")\n                self.logger.info(f\"Save model at Epoch {epoch + 1}\")\n                best_score2 = avg_val_loss2\n                \n        model1.load_state_dict(torch.load(self.path1 / f\"1best{self.fold_num}.pth\"))\n        model2.load_state_dict(torch.load(self.path2 / f\"2best{self.fold_num}.pth\"))\n        \n        valid_preds1, avg_val_loss1, valid_preds2, avg_val_loss2 = self._val(valid_loader, model1, model2)\n        #print('vpp'+str(valid_preds1.shape[1]))\n        #self.logger.info(f\"Best Validation Loss: {avg_val_loss:.8f}\")\n        return valid_preds1, valid_preds2\n    \n    def _val(self, loader, model1, model2):\n        model1.eval()\n        model2.eval()\n        valid_preds1 = np.zeros((len(loader.dataset), num_classes_c))\n        valid_preds2 = np.zeros((len(loader.dataset), num_classes_t))\n        avg_val_loss1 = 0.0\n        avg_val_loss2 = 0.0\n        for i, (i_batch, y_batch) in enumerate(loader):\n            with torch.no_grad():\n                y_pred1 = model1(i_batch).detach()\n                avg_val_loss1 += self.loss_fn(y_pred1, y_batch[0]).item() / len(loader)\n                valid_preds1[i * self.valid_batch:(i + 1) * self.valid_batch] = \\\n                    y_pred1.cpu().numpy()\n                y_pred2 = model2(i_batch).detach()\n                avg_val_loss2 += self.loss_fn(y_pred2, y_batch[1]).item() / len(loader)\n                valid_preds2[i * self.valid_batch:(i + 1) * self.valid_batch] = \\\n                    y_pred2.cpu().numpy()\n        #print('vp1'+str(valid_preds1.shape[1]))\n        return valid_preds1, avg_val_loss1, valid_preds2, avg_val_loss2\n    \n    def predict(self, X):\n        #print('pred')\n        dataset = IMetDataset(X, labels=None)\n        loader = data.DataLoader(dataset, \n                                 batch_size=self.valid_batch, \n                                 shuffle=False)\n        model1 = self.model1(**self.kwargs)\n        model2 = self.model2(**self.kwargs)\n        preds1 = np.zeros((X.size(0), num_classes_c))\n        #print(list(self.path1.iterdir()))\n        for path in self.path1.iterdir():\n            with timer(f\"Using {str(path)}\", self.logger):\n                model1.load_state_dict(torch.load(path))\n                model1.to(self.device)\n                model1.eval()\n                temp1 = np.zeros_like(preds1)\n                #print('try')\n                for i, (i_batch, ) in enumerate(loader):\n                    with torch.no_grad():\n                        y_pred1 = model1(i_batch).detach()\n                        #print(y_pred1[y_pred1 != 0])\n                        temp1[i * self.valid_batch:(i + 1) * self.valid_batch] = \\\n                            y_pred1.cpu().numpy()\n                preds1 += temp1 / self.n_splits\n        preds2 = np.zeros((X.size(0), num_classes_t))\n        for path in self.path2.iterdir():\n            with timer(f\"Using {str(path)}\", self.logger):\n                model2.load_state_dict(torch.load(path))\n                model2.to(self.device)\n                model2.eval()\n                temp2 = np.zeros_like(preds2)\n                for i, (i_batch, ) in enumerate(loader):\n                    with torch.no_grad():\n                        y_pred2 = model2(i_batch).detach()\n                        temp2[i * self.valid_batch:(i + 1) * self.valid_batch] = \\\n                            y_pred2.cpu().numpy()\n                preds2 += temp2 / self.n_splits\n        return preds1, preds2","execution_count":176,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"trainer = Trainer(MultiLayerPerceptron1, MultiLayerPerceptron2, logger, train_batch=64, kwargs={})","execution_count":177,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.model_selection import train_test_split\ny = train.attribute_ids.map(lambda x: x.split()).values\nvalid_preds1, valid_preds2 = trainer.fit(train_tensor, y, n_epochs=40)","execution_count":171,"outputs":[{"output_type":"stream","text":"2019-05-19 16:25:28,854 Main INFO Fold 1\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","text/html":""},"metadata":{}},{"output_type":"stream","text":"2019-05-19 16:25:57,915 Main INFO =========================================\n2019-05-19 16:25:57,916 Main INFO Epoch 1 / 1\n2019-05-19 16:25:57,917 Main INFO =========================================\n2019-05-19 16:25:57,918 Main INFO avg_loss: 0.02138878\n2019-05-19 16:25:57,922 Main INFO avg_val_loss: 0.01234947\n2019-05-19 16:25:57,923 Main INFO avg_loss: 0.02634290\n2019-05-19 16:25:57,924 Main INFO avg_val_loss: 0.01663493\n2019-05-19 16:25:57,941 Main INFO Save model at Epoch 1\n2019-05-19 16:25:57,957 Main INFO Save model at Epoch 1\n2019-05-19 16:26:02,785 Main INFO Fold 2\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","text/html":""},"metadata":{}},{"output_type":"stream","text":"2019-05-19 16:26:31,896 Main INFO =========================================\n2019-05-19 16:26:31,897 Main INFO Epoch 1 / 1\n2019-05-19 16:26:31,899 Main INFO =========================================\n2019-05-19 16:26:31,900 Main INFO avg_loss: 0.02133593\n2019-05-19 16:26:31,902 Main INFO avg_val_loss: 0.01238627\n2019-05-19 16:26:31,904 Main INFO avg_loss: 0.02629573\n2019-05-19 16:26:31,906 Main INFO avg_val_loss: 0.01663687\n2019-05-19 16:26:31,920 Main INFO Save model at Epoch 1\n2019-05-19 16:26:31,934 Main INFO Save model at Epoch 1\n2019-05-19 16:26:36,797 Main INFO Fold 3\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","text/html":""},"metadata":{}},{"output_type":"stream","text":"2019-05-19 16:27:05,762 Main INFO =========================================\n2019-05-19 16:27:05,763 Main INFO Epoch 1 / 1\n2019-05-19 16:27:05,764 Main INFO =========================================\n2019-05-19 16:27:05,767 Main INFO avg_loss: 0.02114178\n2019-05-19 16:27:05,768 Main INFO avg_val_loss: 0.01261050\n2019-05-19 16:27:05,769 Main INFO avg_loss: 0.02610728\n2019-05-19 16:27:05,773 Main INFO avg_val_loss: 0.01681555\n2019-05-19 16:27:05,789 Main INFO Save model at Epoch 1\n2019-05-19 16:27:05,804 Main INFO Save model at Epoch 1\n2019-05-19 16:27:10,612 Main INFO Fold 4\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","text/html":""},"metadata":{}},{"output_type":"stream","text":"2019-05-19 16:27:39,706 Main INFO =========================================\n2019-05-19 16:27:39,707 Main INFO Epoch 1 / 1\n2019-05-19 16:27:39,710 Main INFO =========================================\n2019-05-19 16:27:39,711 Main INFO avg_loss: 0.02080979\n2019-05-19 16:27:39,713 Main INFO avg_val_loss: 0.01306381\n2019-05-19 16:27:39,715 Main INFO avg_loss: 0.02573812\n2019-05-19 16:27:39,717 Main INFO avg_val_loss: 0.01731276\n2019-05-19 16:27:39,733 Main INFO Save model at Epoch 1\n2019-05-19 16:27:39,749 Main INFO Save model at Epoch 1\n2019-05-19 16:27:44,665 Main INFO Fold 5\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","text/html":""},"metadata":{}},{"output_type":"stream","text":"2019-05-19 16:28:13,758 Main INFO =========================================\n2019-05-19 16:28:13,759 Main INFO Epoch 1 / 1\n2019-05-19 16:28:13,760 Main INFO =========================================\n2019-05-19 16:28:13,763 Main INFO avg_loss: 0.02026352\n2019-05-19 16:28:13,765 Main INFO avg_val_loss: 0.01399481\n2019-05-19 16:28:13,768 Main INFO avg_loss: 0.02508862\n2019-05-19 16:28:13,771 Main INFO avg_val_loss: 0.01857798\n2019-05-19 16:28:13,786 Main INFO Save model at Epoch 1\n2019-05-19 16:28:13,801 Main INFO Save model at Epoch 1\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"def threshold_search(y_pred, y_true):\n    score = []\n    candidates = np.arange(0, 1.0, 0.01)\n    for th in progress_bar(candidates):\n        yp = (y_pred > th).astype(int)\n        score.append(fbeta_score(y_pred=yp, y_true=y_true, beta=2, average=\"samples\"))\n    score = np.array(score)\n    pm = score.argmax()\n    best_th, best_score = candidates[pm], score[pm]\n    return best_th, best_score","execution_count":172,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"y_true = np.zeros((train.shape[0], 1103)).astype(int)\nfor i, row in enumerate(y):\n    for idx in row:\n        y_true[i, int(idx)] = 1","execution_count":173,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"best_threshold1, best_score1 = threshold_search(valid_preds1, targets_c)\nbest_score1\nbest_threshold2, best_score2 = threshold_search(valid_preds2, targets_t)\nbest_score2","execution_count":174,"outputs":[{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","text/html":"\n    <div>\n        <style>\n            /* Turns off some styling */\n            progress {\n                /* gets rid of default border in Firefox and Opera. */\n                border: none;\n                /* Needs to be in here for Safari polyfill so background images work as expected. */\n                background-size: auto;\n            }\n            .progress-bar-interrupted, .progress-bar-interrupted::-webkit-progress-bar {\n                background: #F44336;\n            }\n        </style>\n      <progress value='100' class='' max='100', style='width:300px; height:20px; vertical-align: middle;'></progress>\n      100.00% [100/100 05:20<00:00]\n    </div>\n    "},"metadata":{}},{"output_type":"stream","text":"/opt/conda/lib/python3.6/site-packages/sklearn/metrics/classification.py:1143: UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 in samples with no predicted labels.\n  'precision', 'predicted', average, warn_for)\n","name":"stderr"},{"output_type":"display_data","data":{"text/plain":"<IPython.core.display.HTML object>","text/html":"\n    <div>\n        <style>\n            /* Turns off some styling */\n            progress {\n                /* gets rid of default border in Firefox and Opera. */\n                border: none;\n                /* Needs to be in here for Safari polyfill so background images work as expected. */\n                background-size: auto;\n            }\n            .progress-bar-interrupted, .progress-bar-interrupted::-webkit-progress-bar {\n                background: #F44336;\n            }\n        </style>\n      <progress value='100' class='' max='100', style='width:300px; height:20px; vertical-align: middle;'></progress>\n      100.00% [100/100 09:39<00:00]\n    </div>\n    "},"metadata":{}},{"output_type":"execute_result","execution_count":174,"data":{"text/plain":"0.2195233321503681"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"test_preds1, test_preds2  = trainer.predict(test_tensor)","execution_count":178,"outputs":[{"output_type":"stream","text":"pred\n[PosixPath('bin1/2019-05-19-16-25-25/1best3.pth'), PosixPath('bin1/2019-05-19-16-25-25/1best2.pth'), PosixPath('bin1/2019-05-19-16-25-25/1best1.pth'), PosixPath('bin1/2019-05-19-16-25-25/1best0.pth'), PosixPath('bin1/2019-05-19-16-25-25/1best4.pth')]\ntry\ntensor([2.5248e-04, 1.0047e-03, 2.0699e-04,  ..., 2.0171e-05, 1.8681e-04,\n        1.1756e-01], device='cuda:0')\ntensor([1.0845e-04, 6.5119e-04, 9.8716e-05,  ..., 2.2761e-06, 7.9371e-05,\n        2.3472e-01], device='cuda:0')\ntensor([3.1960e-04, 1.0539e-03, 3.1949e-04,  ..., 1.4618e-06, 4.1715e-05,\n        3.9006e-02], device='cuda:0')\ntensor([1.0588e-04, 5.9860e-04, 9.3018e-05,  ..., 1.0097e-05, 1.6525e-04,\n        4.9028e-02], device='cuda:0')\ntensor([0.0002, 0.0007, 0.0002,  ..., 0.0001, 0.0009, 0.0607], device='cuda:0')\ntensor([1.4686e-04, 7.4334e-04, 1.7538e-04,  ..., 2.4546e-06, 6.3998e-05,\n        9.9638e-02], device='cuda:0')\ntensor([1.5932e-04, 6.3283e-04, 1.7209e-04,  ..., 2.1084e-05, 3.3003e-04,\n        2.0034e-01], device='cuda:0')\ntensor([2.8858e-04, 1.0258e-03, 3.5458e-04,  ..., 3.7954e-05, 5.1888e-04,\n        6.4321e-02], device='cuda:0')\ntensor([1.9039e-04, 6.6209e-04, 1.8362e-04,  ..., 7.2135e-06, 1.3152e-04,\n        4.2092e-02], device='cuda:0')\ntensor([5.3481e-05, 2.5657e-04, 4.6897e-05,  ..., 1.7495e-05, 2.8260e-04,\n        1.0420e-01], device='cuda:0')\ntensor([2.7387e-04, 8.5615e-04, 2.9702e-04,  ..., 1.0033e-05, 1.6898e-04,\n        1.1531e-01], device='cuda:0')\ntensor([7.6554e-05, 4.4963e-04, 6.6964e-05,  ..., 3.1791e-05, 4.9156e-04,\n        7.4337e-02], device='cuda:0')\ntensor([3.7065e-05, 1.4663e-04, 2.8307e-05,  ..., 4.3524e-06, 6.6658e-05,\n        5.2802e-02], device='cuda:0')\ntensor([3.4579e-05, 1.3163e-04, 2.7818e-05,  ..., 1.7526e-05, 2.0881e-04,\n        7.6669e-02], device='cuda:0')\ntensor([2.5447e-04, 8.2094e-04, 2.8411e-04,  ..., 3.5043e-05, 3.3535e-04,\n        7.2430e-02], device='cuda:0')\ntensor([1.2132e-04, 5.1377e-04, 1.0916e-04,  ..., 1.5421e-05, 2.2945e-04,\n        9.4971e-02], device='cuda:0')\ntensor([3.6284e-05, 2.5659e-04, 5.2877e-05,  ..., 3.9846e-06, 9.0495e-05,\n        1.5185e-01], device='cuda:0')\ntensor([4.3448e-05, 1.2990e-04, 3.0193e-05,  ..., 6.8278e-06, 9.7441e-05,\n        1.0050e-01], device='cuda:0')\ntensor([2.5437e-04, 1.3198e-03, 2.4391e-04,  ..., 3.9611e-06, 8.8353e-05,\n        3.0863e-02], device='cuda:0')\ntensor([5.0923e-05, 2.0941e-04, 6.4631e-05,  ..., 5.6607e-05, 6.1455e-04,\n        1.2271e-01], device='cuda:0')\ntensor([1.5854e-04, 8.4954e-04, 1.6492e-04,  ..., 2.8308e-05, 4.0462e-04,\n        1.2450e-01], device='cuda:0')\ntensor([1.6438e-04, 6.4801e-04, 1.5804e-04,  ..., 8.9967e-06, 1.7940e-04,\n        2.4677e-02], device='cuda:0')\ntensor([1.8999e-04, 8.5593e-04, 1.4773e-04,  ..., 2.6971e-06, 7.5039e-05,\n        5.1446e-02], device='cuda:0')\ntensor([5.8442e-05, 3.0594e-04, 8.6486e-05,  ..., 4.1223e-05, 4.3887e-04,\n        8.2267e-02], device='cuda:0')\ntensor([1.4860e-04, 5.8503e-04, 1.5655e-04,  ..., 3.2254e-06, 5.8943e-05,\n        5.6555e-02], device='cuda:0')\ntensor([3.5780e-04, 1.5263e-03, 3.8908e-04,  ..., 4.0408e-05, 4.1415e-04,\n        7.1516e-02], device='cuda:0')\ntensor([5.5417e-05, 1.5300e-04, 4.4462e-05,  ..., 1.1746e-06, 4.6727e-05,\n        5.1847e-02], device='cuda:0')\ntensor([1.4192e-04, 6.8496e-04, 1.6960e-04,  ..., 9.4836e-06, 1.4273e-04,\n        6.4575e-02], device='cuda:0')\ntensor([1.7471e-04, 6.5600e-04, 1.9917e-04,  ..., 7.7117e-05, 6.9255e-04,\n        5.5992e-02], device='cuda:0')\ntensor([1.7801e-04, 6.9098e-04, 2.2215e-04,  ..., 2.8840e-05, 3.6590e-04,\n        6.1227e-02], device='cuda:0')\ntensor([1.7260e-04, 1.0440e-03, 1.8017e-04,  ..., 6.7057e-07, 3.0053e-05,\n        3.2198e-02], device='cuda:0')\ntensor([0.0001, 0.0008, 0.0002,  ..., 0.0001, 0.0011, 0.0739], device='cuda:0')\ntensor([5.3338e-04, 1.2803e-03, 5.1491e-04,  ..., 1.5534e-05, 1.9538e-04,\n        1.3638e-01], device='cuda:0')\ntensor([6.7797e-05, 2.0875e-04, 4.8446e-05,  ..., 1.7294e-05, 2.0979e-04,\n        1.1949e-01], device='cuda:0')\ntensor([2.0441e-04, 6.9628e-04, 2.2245e-04,  ..., 1.8985e-06, 4.6852e-05,\n        7.5620e-02], device='cuda:0')\ntensor([7.5600e-05, 4.1625e-04, 7.8562e-05,  ..., 3.0279e-06, 9.3555e-05,\n        5.9604e-02], device='cuda:0')\ntensor([0.0002, 0.0007, 0.0001,  ..., 0.0001, 0.0012, 0.0632], device='cuda:0')\ntensor([3.5326e-05, 1.3882e-04, 3.8747e-05,  ..., 2.0331e-05, 2.2088e-04,\n        8.5661e-02], device='cuda:0')\ntensor([1.4964e-04, 5.0688e-04, 1.5978e-04,  ..., 1.0560e-05, 1.7017e-04,\n        9.9336e-02], device='cuda:0')\ntensor([3.9710e-04, 1.7999e-03, 3.7926e-04,  ..., 2.0297e-05, 2.9067e-04,\n        6.5886e-02], device='cuda:0')\ntensor([1.4123e-04, 7.1513e-04, 1.3134e-04,  ..., 8.1506e-06, 1.3327e-04,\n        4.7030e-02], device='cuda:0')\ntensor([1.0012e-04, 4.0364e-04, 1.0389e-04,  ..., 1.8997e-05, 1.6616e-04,\n        9.3157e-02], device='cuda:0')\ntensor([1.3297e-04, 3.8410e-04, 9.6362e-05,  ..., 8.0020e-06, 1.3763e-04,\n        2.7109e-02], device='cuda:0')\ntensor([2.5244e-05, 1.0864e-04, 2.8050e-05,  ..., 3.0680e-05, 4.2500e-04,\n        4.7448e-02], device='cuda:0')\ntensor([1.0478e-04, 4.3339e-04, 8.8060e-05,  ..., 1.2196e-05, 1.8846e-04,\n        5.8060e-02], device='cuda:0')\ntensor([3.3777e-04, 9.5882e-04, 3.1143e-04,  ..., 2.1186e-06, 7.3715e-05,\n        1.6493e-01], device='cuda:0')\ntensor([3.5899e-05, 1.8176e-04, 3.5860e-05,  ..., 1.2150e-05, 1.0620e-04,\n        6.4251e-02], device='cuda:0')\ntensor([2.1831e-04, 6.1446e-04, 1.7443e-04,  ..., 2.7546e-05, 1.8366e-04,\n        5.9859e-02], device='cuda:0')\ntensor([3.0905e-04, 9.9655e-04, 4.5445e-04,  ..., 1.6570e-06, 5.1263e-05,\n        6.0754e-02], device='cuda:0')\ntensor([3.9234e-05, 1.7121e-04, 4.8767e-05,  ..., 1.6799e-06, 6.7315e-05,\n        1.4209e-01], device='cuda:0')\ntensor([8.9190e-05, 3.4621e-04, 1.2171e-04,  ..., 3.9602e-06, 9.3781e-05,\n        5.1829e-02], device='cuda:0')\ntensor([1.4131e-04, 5.8898e-04, 1.6611e-04,  ..., 1.0043e-05, 1.6247e-04,\n        1.2887e-01], device='cuda:0')\ntensor([1.0381e-04, 5.7826e-04, 1.0089e-04,  ..., 3.3578e-05, 4.0993e-04,\n        1.0412e-01], device='cuda:0')\ntensor([1.2221e-04, 7.0937e-04, 1.2988e-04,  ..., 1.6313e-05, 2.0708e-04,\n        1.7244e-01], device='cuda:0')\ntensor([5.6622e-05, 2.2667e-04, 5.5484e-05,  ..., 1.1434e-05, 2.2229e-04,\n        2.1011e-01], device='cuda:0')\ntensor([5.8998e-05, 2.8665e-04, 8.3317e-05,  ..., 7.9705e-05, 5.1811e-04,\n        8.4913e-02], device='cuda:0')\ntensor([3.4093e-04, 1.3696e-03, 4.4566e-04,  ..., 9.0744e-06, 1.5212e-04,\n        9.6815e-02], device='cuda:0')\ntensor([1.5274e-04, 5.4864e-04, 1.3980e-04,  ..., 5.7973e-06, 1.2917e-04,\n        1.6763e-01], device='cuda:0')\ntensor([2.5111e-04, 6.5339e-04, 2.3425e-04,  ..., 4.3289e-05, 4.7662e-04,\n        1.2444e-01], device='cuda:0')\n2019-05-19 16:49:36,203 Main INFO [Using bin1/2019-05-19-16-25-25/1best3.pth] done in 0.20346999168395996 s\ntry\ntensor([2.0127e-04, 8.5388e-04, 1.7882e-04,  ..., 2.0886e-05, 1.8192e-04,\n        1.0291e-01], device='cuda:0')\ntensor([1.5046e-04, 7.7087e-04, 1.3033e-04,  ..., 8.1086e-06, 1.2829e-04,\n        1.6476e-01], device='cuda:0')\ntensor([3.3545e-04, 1.1160e-03, 3.1214e-04,  ..., 5.3422e-06, 7.3942e-05,\n        6.3478e-02], device='cuda:0')\ntensor([1.0279e-04, 5.6734e-04, 9.4352e-05,  ..., 1.3368e-05, 1.6409e-04,\n        6.7512e-02], device='cuda:0')\ntensor([1.7777e-04, 6.2408e-04, 1.7489e-04,  ..., 8.6313e-05, 6.2691e-04,\n        8.1696e-02], device='cuda:0')\ntensor([1.5702e-04, 6.6963e-04, 1.5815e-04,  ..., 1.0603e-05, 1.3870e-04,\n        9.2424e-02], device='cuda:0')\ntensor([2.5477e-04, 8.7447e-04, 2.4365e-04,  ..., 3.8175e-05, 3.9133e-04,\n        1.6509e-01], device='cuda:0')\ntensor([2.6192e-04, 9.4510e-04, 2.6018e-04,  ..., 5.0940e-05, 5.5792e-04,\n        8.5770e-02], device='cuda:0')\ntensor([1.7507e-04, 6.7148e-04, 1.5870e-04,  ..., 2.2488e-05, 2.3981e-04,\n        7.8054e-02], device='cuda:0')\ntensor([1.1399e-04, 4.2165e-04, 1.0402e-04,  ..., 2.6190e-05, 2.9028e-04,\n        1.0872e-01], device='cuda:0')\ntensor([2.3753e-04, 7.2715e-04, 2.1929e-04,  ..., 2.3353e-05, 2.5972e-04,\n        1.2204e-01], device='cuda:0')\ntensor([9.9793e-05, 5.0651e-04, 7.7268e-05,  ..., 2.9064e-05, 3.9500e-04,\n        7.8303e-02], device='cuda:0')\ntensor([5.3852e-05, 2.1307e-04, 4.8548e-05,  ..., 6.9045e-06, 8.0072e-05,\n        6.8695e-02], device='cuda:0')\ntensor([4.5104e-05, 1.5751e-04, 3.4652e-05,  ..., 1.9565e-05, 1.9199e-04,\n        9.8079e-02], device='cuda:0')\ntensor([3.1430e-04, 9.3050e-04, 2.9126e-04,  ..., 3.4069e-05, 3.1295e-04,\n        8.6097e-02], device='cuda:0')\ntensor([1.0981e-04, 4.9383e-04, 1.0185e-04,  ..., 2.4009e-05, 2.9466e-04,\n        1.0589e-01], device='cuda:0')\ntensor([7.6895e-05, 3.7544e-04, 7.9195e-05,  ..., 9.9944e-06, 1.2085e-04,\n        1.2829e-01], device='cuda:0')\ntensor([4.7549e-05, 1.5613e-04, 3.5672e-05,  ..., 2.0278e-05, 1.9257e-04,\n        1.0876e-01], device='cuda:0')\ntensor([2.7926e-04, 1.3559e-03, 2.7169e-04,  ..., 8.5976e-06, 1.4293e-04,\n        5.6661e-02], device='cuda:0')\ntensor([1.0345e-04, 3.4579e-04, 9.8486e-05,  ..., 5.9713e-05, 5.1425e-04,\n        1.1822e-01], device='cuda:0')\n","name":"stdout"},{"output_type":"stream","text":"tensor([1.4523e-04, 7.5082e-04, 1.3365e-04,  ..., 4.0698e-05, 4.5111e-04,\n        1.2616e-01], device='cuda:0')\ntensor([2.0306e-04, 6.7481e-04, 1.6195e-04,  ..., 1.1499e-05, 1.8538e-04,\n        4.0496e-02], device='cuda:0')\ntensor([1.7492e-04, 8.4895e-04, 1.5897e-04,  ..., 1.1469e-05, 1.7444e-04,\n        7.5672e-02], device='cuda:0')\ntensor([1.3471e-04, 5.0354e-04, 1.3491e-04,  ..., 3.3929e-05, 3.3235e-04,\n        8.6545e-02], device='cuda:0')\ntensor([1.7555e-04, 6.6057e-04, 1.6349e-04,  ..., 7.5237e-06, 9.6806e-05,\n        8.3095e-02], device='cuda:0')\ntensor([2.2843e-04, 1.0647e-03, 2.2277e-04,  ..., 5.1600e-05, 4.0816e-04,\n        9.2395e-02], device='cuda:0')\ntensor([7.5698e-05, 2.0190e-04, 6.1976e-05,  ..., 5.4642e-06, 1.1409e-04,\n        8.6126e-02], device='cuda:0')\ntensor([1.4171e-04, 6.8432e-04, 1.5590e-04,  ..., 1.4096e-05, 1.6877e-04,\n        7.7436e-02], device='cuda:0')\ntensor([1.6728e-04, 6.6626e-04, 1.9981e-04,  ..., 7.8692e-05, 6.3388e-04,\n        7.4540e-02], device='cuda:0')\ntensor([1.6520e-04, 5.9420e-04, 1.6756e-04,  ..., 2.9286e-05, 3.1266e-04,\n        8.1591e-02], device='cuda:0')\ntensor([2.0190e-04, 1.0074e-03, 1.8838e-04,  ..., 3.2537e-06, 7.2762e-05,\n        6.0951e-02], device='cuda:0')\ntensor([1.3322e-04, 8.0839e-04, 1.6368e-04,  ..., 8.3553e-05, 6.8768e-04,\n        9.1532e-02], device='cuda:0')\ntensor([3.2184e-04, 1.0589e-03, 3.0153e-04,  ..., 2.3160e-05, 2.2415e-04,\n        1.1353e-01], device='cuda:0')\ntensor([8.6299e-05, 2.7160e-04, 7.3342e-05,  ..., 2.5276e-05, 2.6618e-04,\n        1.0805e-01], device='cuda:0')\ntensor([2.3767e-04, 7.4789e-04, 2.1829e-04,  ..., 5.2213e-06, 6.7314e-05,\n        7.7775e-02], device='cuda:0')\ntensor([6.7190e-05, 3.2611e-04, 7.1301e-05,  ..., 6.7564e-06, 1.2828e-04,\n        7.8069e-02], device='cuda:0')\ntensor([1.6411e-04, 7.2696e-04, 1.5193e-04,  ..., 9.7519e-05, 7.9435e-04,\n        8.4654e-02], device='cuda:0')\ntensor([5.0856e-05, 1.8459e-04, 4.5904e-05,  ..., 2.1404e-05, 2.0220e-04,\n        9.2556e-02], device='cuda:0')\ntensor([1.4225e-04, 5.2702e-04, 1.3838e-04,  ..., 1.6461e-05, 1.9632e-04,\n        9.3773e-02], device='cuda:0')\ntensor([4.3113e-04, 1.7219e-03, 3.8703e-04,  ..., 3.1647e-05, 3.2400e-04,\n        8.8178e-02], device='cuda:0')\ntensor([1.6843e-04, 7.4088e-04, 1.7253e-04,  ..., 2.2091e-05, 2.2922e-04,\n        7.1645e-02], device='cuda:0')\ntensor([1.0384e-04, 4.1567e-04, 8.8838e-05,  ..., 2.7341e-05, 2.2124e-04,\n        9.1844e-02], device='cuda:0')\ntensor([1.3960e-04, 4.7437e-04, 1.1982e-04,  ..., 1.6797e-05, 2.2338e-04,\n        5.5489e-02], device='cuda:0')\ntensor([5.8708e-05, 1.9144e-04, 5.1569e-05,  ..., 3.5367e-05, 3.9872e-04,\n        7.5347e-02], device='cuda:0')\ntensor([1.6716e-04, 5.5149e-04, 1.3236e-04,  ..., 2.0679e-05, 2.2877e-04,\n        8.2441e-02], device='cuda:0')\ntensor([3.3414e-04, 9.7099e-04, 3.1936e-04,  ..., 7.7144e-06, 1.3338e-04,\n        1.5019e-01], device='cuda:0')\ntensor([8.7971e-05, 3.1432e-04, 6.3508e-05,  ..., 1.1391e-05, 1.0729e-04,\n        7.1149e-02], device='cuda:0')\ntensor([1.8934e-04, 5.5996e-04, 1.4641e-04,  ..., 3.2799e-05, 2.2434e-04,\n        6.9532e-02], device='cuda:0')\ntensor([3.7470e-04, 1.1106e-03, 4.3046e-04,  ..., 3.7889e-06, 6.8690e-05,\n        8.3656e-02], device='cuda:0')\ntensor([7.8944e-05, 2.7695e-04, 7.7077e-05,  ..., 4.3801e-06, 9.5188e-05,\n        1.3365e-01], device='cuda:0')\ntensor([1.1755e-04, 4.1467e-04, 1.2631e-04,  ..., 1.2959e-05, 1.7823e-04,\n        7.4657e-02], device='cuda:0')\ntensor([1.3630e-04, 5.9404e-04, 1.4544e-04,  ..., 1.4850e-05, 1.8116e-04,\n        1.2114e-01], device='cuda:0')\ntensor([1.0748e-04, 5.2367e-04, 9.7092e-05,  ..., 3.6603e-05, 3.5506e-04,\n        9.7478e-02], device='cuda:0')\ntensor([1.4987e-04, 7.9341e-04, 1.4227e-04,  ..., 2.6034e-05, 2.5400e-04,\n        1.2044e-01], device='cuda:0')\ntensor([1.0214e-04, 3.5412e-04, 8.2884e-05,  ..., 2.0520e-05, 2.4791e-04,\n        1.5898e-01], device='cuda:0')\ntensor([1.2081e-04, 4.5112e-04, 1.2355e-04,  ..., 8.1696e-05, 4.8428e-04,\n        9.5893e-02], device='cuda:0')\ntensor([2.3242e-04, 1.1160e-03, 2.9675e-04,  ..., 1.8674e-05, 1.9859e-04,\n        1.0150e-01], device='cuda:0')\ntensor([2.1003e-04, 7.1963e-04, 1.7412e-04,  ..., 1.5104e-05, 1.8573e-04,\n        1.5184e-01], device='cuda:0')\ntensor([2.1559e-04, 6.3657e-04, 1.8504e-04,  ..., 4.7663e-05, 4.3488e-04,\n        1.1173e-01], device='cuda:0')\n2019-05-19 16:49:36,398 Main INFO [Using bin1/2019-05-19-16-25-25/1best2.pth] done in 0.1943821907043457 s\ntry\ntensor([1.6865e-04, 5.0433e-04, 8.7156e-05,  ..., 2.8947e-05, 1.7228e-04,\n        1.1144e-01], device='cuda:0')\ntensor([1.9327e-04, 5.2135e-04, 9.6983e-05,  ..., 1.7860e-05, 1.3524e-04,\n        1.2635e-01], device='cuda:0')\ntensor([3.2209e-04, 7.3285e-04, 1.9091e-04,  ..., 1.0142e-05, 8.0802e-05,\n        9.1828e-02], device='cuda:0')\ntensor([9.9596e-05, 3.7753e-04, 5.2592e-05,  ..., 1.5026e-05, 1.1544e-04,\n        9.6684e-02], device='cuda:0')\ntensor([1.9082e-04, 5.2397e-04, 1.2008e-04,  ..., 6.0368e-05, 3.2976e-04,\n        1.1383e-01], device='cuda:0')\ntensor([1.7238e-04, 5.2388e-04, 1.0341e-04,  ..., 1.5050e-05, 1.1481e-04,\n        1.0562e-01], device='cuda:0')\ntensor([2.7727e-04, 7.6121e-04, 1.5807e-04,  ..., 6.3297e-05, 3.9519e-04,\n        1.5704e-01], device='cuda:0')\ntensor([3.4112e-04, 8.7250e-04, 2.0517e-04,  ..., 7.0677e-05, 4.9335e-04,\n        1.3883e-01], device='cuda:0')\ntensor([2.2120e-04, 6.5579e-04, 1.3236e-04,  ..., 2.7554e-05, 1.9294e-04,\n        1.0463e-01], device='cuda:0')\ntensor([2.2135e-04, 6.3684e-04, 1.1328e-04,  ..., 2.7178e-05, 2.1007e-04,\n        1.2945e-01], device='cuda:0')\ntensor([2.4142e-04, 6.4157e-04, 1.2907e-04,  ..., 2.5219e-05, 1.7757e-04,\n        1.3698e-01], device='cuda:0')\ntensor([1.1462e-04, 3.8891e-04, 5.7325e-05,  ..., 3.3819e-05, 2.9502e-04,\n        1.1101e-01], device='cuda:0')\ntensor([9.9319e-05, 2.7028e-04, 5.1764e-05,  ..., 7.6979e-06, 5.9764e-05,\n        8.5472e-02], device='cuda:0')\ntensor([6.8096e-05, 1.9597e-04, 3.5199e-05,  ..., 2.0519e-05, 1.4173e-04,\n        1.1060e-01], device='cuda:0')\ntensor([3.8801e-04, 9.0540e-04, 2.2691e-04,  ..., 2.9937e-05, 1.9439e-04,\n        1.0572e-01], device='cuda:0')\ntensor([1.3735e-04, 4.2980e-04, 7.0603e-05,  ..., 3.9282e-05, 2.7881e-04,\n        1.2086e-01], device='cuda:0')\ntensor([1.0290e-04, 3.5021e-04, 5.6831e-05,  ..., 1.0106e-05, 8.1266e-05,\n        1.1077e-01], device='cuda:0')\ntensor([5.6854e-05, 1.7955e-04, 3.2844e-05,  ..., 2.4779e-05, 1.4905e-04,\n        1.1881e-01], device='cuda:0')\ntensor([2.8665e-04, 8.7528e-04, 1.7134e-04,  ..., 1.4105e-05, 1.2609e-04,\n        9.8726e-02], device='cuda:0')\ntensor([1.2791e-04, 3.6140e-04, 7.0888e-05,  ..., 6.4238e-05, 3.8159e-04,\n        1.3698e-01], device='cuda:0')\ntensor([1.5660e-04, 4.4693e-04, 7.9669e-05,  ..., 5.7687e-05, 3.6661e-04,\n        1.5156e-01], device='cuda:0')\ntensor([2.1040e-04, 5.1420e-04, 9.3088e-05,  ..., 1.5402e-05, 1.5190e-04,\n        8.1191e-02], device='cuda:0')\ntensor([1.5488e-04, 5.4291e-04, 8.2914e-05,  ..., 1.9206e-05, 1.5183e-04,\n        1.0180e-01], device='cuda:0')\ntensor([1.6408e-04, 4.8719e-04, 9.7505e-05,  ..., 4.4209e-05, 2.8678e-04,\n        1.3143e-01], device='cuda:0')\ntensor([2.1065e-04, 5.5437e-04, 1.1502e-04,  ..., 9.7353e-06, 7.3791e-05,\n        9.9300e-02], device='cuda:0')\ntensor([2.5836e-04, 7.8094e-04, 1.5011e-04,  ..., 4.1676e-05, 2.5656e-04,\n        1.1429e-01], device='cuda:0')\ntensor([1.1908e-04, 2.7709e-04, 6.0339e-05,  ..., 1.1274e-05, 1.2345e-04,\n        1.1666e-01], device='cuda:0')\ntensor([1.8557e-04, 5.7876e-04, 1.2298e-04,  ..., 1.6627e-05, 1.2886e-04,\n        1.0045e-01], device='cuda:0')\ntensor([1.8658e-04, 5.4360e-04, 1.1744e-04,  ..., 5.3531e-05, 3.4413e-04,\n        1.0664e-01], device='cuda:0')\ntensor([1.9895e-04, 4.7611e-04, 1.1356e-04,  ..., 2.9231e-05, 2.0754e-04,\n        1.0949e-01], device='cuda:0')\ntensor([2.0632e-04, 6.3903e-04, 1.1181e-04,  ..., 7.6454e-06, 7.8536e-05,\n        9.7271e-02], device='cuda:0')\ntensor([1.2744e-04, 4.5526e-04, 8.8068e-05,  ..., 6.9815e-05, 4.2666e-04,\n        1.1861e-01], device='cuda:0')\ntensor([2.5476e-04, 6.8752e-04, 1.2943e-04,  ..., 2.1811e-05, 1.5665e-04,\n        1.1774e-01], device='cuda:0')\ntensor([1.3545e-04, 3.1882e-04, 7.7509e-05,  ..., 2.9944e-05, 2.1924e-04,\n        1.2227e-01], device='cuda:0')\ntensor([2.2073e-04, 5.3567e-04, 1.1041e-04,  ..., 5.5806e-06, 4.6858e-05,\n        8.1695e-02], device='cuda:0')\ntensor([7.9553e-05, 2.2557e-04, 4.3695e-05,  ..., 8.2750e-06, 8.9393e-05,\n        1.0484e-01], device='cuda:0')\ntensor([1.7509e-04, 5.1372e-04, 8.9030e-05,  ..., 8.0243e-05, 4.5190e-04,\n        1.1761e-01], device='cuda:0')\ntensor([6.8566e-05, 1.9359e-04, 3.6957e-05,  ..., 1.8420e-05, 1.2900e-04,\n        1.0702e-01], device='cuda:0')\ntensor([1.5984e-04, 4.5134e-04, 9.8498e-05,  ..., 1.9905e-05, 1.7711e-04,\n        1.1039e-01], device='cuda:0')\ntensor([4.0748e-04, 1.1069e-03, 2.0331e-04,  ..., 3.8031e-05, 2.3801e-04,\n        1.2908e-01], device='cuda:0')\ntensor([2.4016e-04, 7.0380e-04, 1.6074e-04,  ..., 2.7139e-05, 1.8220e-04,\n        9.6328e-02], device='cuda:0')\ntensor([1.2529e-04, 3.5442e-04, 6.0591e-05,  ..., 2.6043e-05, 1.6797e-04,\n        9.8732e-02], device='cuda:0')\ntensor([1.8176e-04, 5.4657e-04, 1.1515e-04,  ..., 2.0287e-05, 1.5743e-04,\n        9.2721e-02], device='cuda:0')\ntensor([7.9374e-05, 2.2145e-04, 3.9823e-05,  ..., 4.4698e-05, 3.2134e-04,\n        1.1853e-01], device='cuda:0')\ntensor([2.2744e-04, 5.7447e-04, 1.1365e-04,  ..., 4.0659e-05, 2.4685e-04,\n        1.1490e-01], device='cuda:0')\ntensor([4.2132e-04, 9.9118e-04, 2.4111e-04,  ..., 1.7323e-05, 1.4792e-04,\n        1.3697e-01], device='cuda:0')\ntensor([9.8974e-05, 2.7005e-04, 4.0975e-05,  ..., 1.0000e-05, 7.4533e-05,\n        8.3170e-02], device='cuda:0')\ntensor([1.9004e-04, 4.3209e-04, 8.4081e-05,  ..., 3.4828e-05, 1.8780e-04,\n        9.2795e-02], device='cuda:0')\ntensor([3.5647e-04, 8.6178e-04, 2.2601e-04,  ..., 6.8685e-06, 6.3600e-05,\n        9.9099e-02], device='cuda:0')\ntensor([1.1490e-04, 2.9191e-04, 5.9276e-05,  ..., 9.0483e-06, 9.7135e-05,\n        1.3024e-01], device='cuda:0')\ntensor([1.5586e-04, 4.3121e-04, 9.4451e-05,  ..., 1.6007e-05, 1.3581e-04,\n        1.0392e-01], device='cuda:0')\ntensor([1.3303e-04, 4.3612e-04, 8.4677e-05,  ..., 1.1178e-05, 1.0146e-04,\n        1.1993e-01], device='cuda:0')\ntensor([1.3437e-04, 3.7778e-04, 6.3490e-05,  ..., 2.9420e-05, 2.2016e-04,\n        1.1705e-01], device='cuda:0')\ntensor([1.7554e-04, 6.2275e-04, 9.3524e-05,  ..., 3.1888e-05, 2.5260e-04,\n        1.2388e-01], device='cuda:0')\ntensor([1.2306e-04, 3.3455e-04, 6.3030e-05,  ..., 3.3523e-05, 2.4944e-04,\n        1.4130e-01], device='cuda:0')\ntensor([1.4259e-04, 4.1669e-04, 8.1027e-05,  ..., 6.4980e-05, 3.0735e-04,\n        1.0497e-01], device='cuda:0')\ntensor([1.7769e-04, 5.5260e-04, 1.1672e-04,  ..., 1.8092e-05, 1.3210e-04,\n        1.1701e-01], device='cuda:0')\ntensor([2.9934e-04, 7.7052e-04, 1.6880e-04,  ..., 2.9699e-05, 2.0034e-04,\n        1.4160e-01], device='cuda:0')\ntensor([2.1328e-04, 5.3285e-04, 1.1792e-04,  ..., 4.6452e-05, 3.0814e-04,\n        1.3623e-01], device='cuda:0')\n2019-05-19 16:49:36,598 Main INFO [Using bin1/2019-05-19-16-25-25/1best1.pth] done in 0.1984107494354248 s\n","name":"stdout"},{"output_type":"stream","text":"try\ntensor([1.6519e-04, 4.3715e-04, 7.8689e-05,  ..., 3.2086e-05, 1.5128e-04,\n        1.1245e-01], device='cuda:0')\ntensor([2.4015e-04, 5.3722e-04, 1.1349e-04,  ..., 3.5127e-05, 1.8444e-04,\n        1.0318e-01], device='cuda:0')\ntensor([3.5212e-04, 7.2613e-04, 1.9520e-04,  ..., 2.0357e-05, 1.0906e-04,\n        1.0010e-01], device='cuda:0')\ntensor([1.2023e-04, 3.4939e-04, 5.6232e-05,  ..., 1.7963e-05, 1.0899e-04,\n        1.0362e-01], device='cuda:0')\ntensor([1.9896e-04, 4.8297e-04, 1.0644e-04,  ..., 4.5298e-05, 2.2285e-04,\n        1.1393e-01], device='cuda:0')\ntensor([1.9802e-04, 5.1493e-04, 1.0400e-04,  ..., 2.5358e-05, 1.4228e-04,\n        1.0459e-01], device='cuda:0')\ntensor([3.0758e-04, 7.7911e-04, 1.4876e-04,  ..., 9.6787e-05, 4.3341e-04,\n        1.3379e-01], device='cuda:0')\ntensor([2.7259e-04, 6.9817e-04, 1.3811e-04,  ..., 8.9808e-05, 4.5641e-04,\n        1.3972e-01], device='cuda:0')\ntensor([2.4778e-04, 6.7353e-04, 1.3230e-04,  ..., 4.8435e-05, 2.3615e-04,\n        1.1443e-01], device='cuda:0')\ntensor([3.2216e-04, 8.1791e-04, 1.6063e-04,  ..., 2.9946e-05, 1.9680e-04,\n        1.1219e-01], device='cuda:0')\ntensor([2.7832e-04, 6.3775e-04, 1.3588e-04,  ..., 2.9302e-05, 1.6717e-04,\n        1.2169e-01], device='cuda:0')\ntensor([1.4318e-04, 4.2719e-04, 6.0611e-05,  ..., 3.1577e-05, 2.3566e-04,\n        1.1342e-01], device='cuda:0')\ntensor([1.0767e-04, 2.7695e-04, 5.6545e-05,  ..., 1.3651e-05, 7.4428e-05,\n        9.1414e-02], device='cuda:0')\ntensor([9.4173e-05, 2.6651e-04, 4.8042e-05,  ..., 2.5900e-05, 1.4009e-04,\n        1.0618e-01], device='cuda:0')\ntensor([4.3630e-04, 9.4694e-04, 2.2116e-04,  ..., 2.7059e-05, 1.5528e-04,\n        1.0363e-01], device='cuda:0')\ntensor([1.5672e-04, 4.1595e-04, 6.7182e-05,  ..., 5.6772e-05, 3.0481e-04,\n        1.1861e-01], device='cuda:0')\ntensor([1.4546e-04, 4.5645e-04, 6.8681e-05,  ..., 1.7112e-05, 8.6252e-05,\n        9.2774e-02], device='cuda:0')\ntensor([7.4523e-05, 2.4042e-04, 3.5833e-05,  ..., 3.8293e-05, 1.7858e-04,\n        1.1097e-01], device='cuda:0')\ntensor([2.7207e-04, 7.3190e-04, 1.5348e-04,  ..., 2.4352e-05, 1.4904e-04,\n        1.1199e-01], device='cuda:0')\ntensor([2.0545e-04, 5.4001e-04, 9.6849e-05,  ..., 7.6823e-05, 3.8265e-04,\n        1.3564e-01], device='cuda:0')\ntensor([1.7399e-04, 3.8839e-04, 7.8929e-05,  ..., 7.3163e-05, 3.3739e-04,\n        1.3509e-01], device='cuda:0')\ntensor([2.4521e-04, 5.1557e-04, 9.5018e-05,  ..., 1.9461e-05, 1.3991e-04,\n        9.4245e-02], device='cuda:0')\ntensor([1.3470e-04, 4.2153e-04, 6.6600e-05,  ..., 3.7709e-05, 2.0418e-04,\n        1.1425e-01], device='cuda:0')\ntensor([2.3276e-04, 6.4141e-04, 1.1791e-04,  ..., 5.0598e-05, 2.7964e-04,\n        1.3925e-01], device='cuda:0')\ntensor([2.5760e-04, 6.1090e-04, 1.2695e-04,  ..., 1.8779e-05, 9.7200e-05,\n        1.0188e-01], device='cuda:0')\ntensor([1.6967e-04, 4.6359e-04, 8.3923e-05,  ..., 4.3102e-05, 2.1557e-04,\n        1.1242e-01], device='cuda:0')\ntensor([1.6251e-04, 3.7982e-04, 8.1038e-05,  ..., 1.6960e-05, 1.5082e-04,\n        1.1820e-01], device='cuda:0')\ntensor([2.0764e-04, 5.8361e-04, 1.1645e-04,  ..., 1.7546e-05, 1.0796e-04,\n        1.0232e-01], device='cuda:0')\ntensor([1.8276e-04, 4.8731e-04, 1.0906e-04,  ..., 5.5657e-05, 2.9265e-04,\n        1.1536e-01], device='cuda:0')\ntensor([2.2130e-04, 4.5676e-04, 1.1174e-04,  ..., 3.4336e-05, 2.0012e-04,\n        1.1248e-01], device='cuda:0')\ntensor([2.6682e-04, 6.5517e-04, 1.3548e-04,  ..., 1.5716e-05, 1.0981e-04,\n        1.1240e-01], device='cuda:0')\ntensor([1.2192e-04, 3.3780e-04, 7.2273e-05,  ..., 5.6965e-05, 3.0509e-04,\n        1.2270e-01], device='cuda:0')\ntensor([1.7859e-04, 4.5636e-04, 8.0737e-05,  ..., 2.5682e-05, 1.5615e-04,\n        1.0898e-01], device='cuda:0')\ntensor([1.7749e-04, 3.8988e-04, 9.4303e-05,  ..., 3.8602e-05, 2.3349e-04,\n        1.1538e-01], device='cuda:0')\ntensor([2.0305e-04, 4.9518e-04, 9.6946e-05,  ..., 1.1246e-05, 6.1559e-05,\n        7.8139e-02], device='cuda:0')\ntensor([9.1930e-05, 2.2774e-04, 4.8986e-05,  ..., 1.1834e-05, 1.0223e-04,\n        9.4193e-02], device='cuda:0')\ntensor([1.8232e-04, 4.4390e-04, 8.8379e-05,  ..., 5.2598e-05, 2.6750e-04,\n        1.2272e-01], device='cuda:0')\ntensor([1.2102e-04, 2.9033e-04, 5.5653e-05,  ..., 1.7205e-05, 1.0356e-04,\n        9.6564e-02], device='cuda:0')\ntensor([1.6430e-04, 4.5911e-04, 8.4708e-05,  ..., 2.7624e-05, 1.9717e-04,\n        1.1130e-01], device='cuda:0')\ntensor([4.0464e-04, 8.5505e-04, 1.7743e-04,  ..., 4.7614e-05, 2.5090e-04,\n        1.2775e-01], device='cuda:0')\ntensor([2.5007e-04, 6.8682e-04, 1.5470e-04,  ..., 4.5358e-05, 2.2125e-04,\n        1.0560e-01], device='cuda:0')\ntensor([1.2239e-04, 3.2570e-04, 4.9433e-05,  ..., 3.2421e-05, 1.7488e-04,\n        1.0396e-01], device='cuda:0')\ntensor([2.2674e-04, 6.5253e-04, 1.2240e-04,  ..., 3.2249e-05, 1.8056e-04,\n        1.1062e-01], device='cuda:0')\ntensor([1.3750e-04, 3.3539e-04, 5.9078e-05,  ..., 5.3064e-05, 2.9062e-04,\n        1.1925e-01], device='cuda:0')\ntensor([2.8532e-04, 6.6629e-04, 1.3861e-04,  ..., 5.4476e-05, 2.4315e-04,\n        1.1924e-01], device='cuda:0')\ntensor([5.6516e-04, 1.1534e-03, 3.0356e-04,  ..., 2.7101e-05, 1.6755e-04,\n        1.1602e-01], device='cuda:0')\ntensor([1.3499e-04, 3.5500e-04, 4.7007e-05,  ..., 1.2825e-05, 7.4889e-05,\n        8.7142e-02], device='cuda:0')\ntensor([1.5262e-04, 3.2247e-04, 5.8465e-05,  ..., 3.9815e-05, 1.8979e-04,\n        1.1209e-01], device='cuda:0')\ntensor([5.4618e-04, 1.1615e-03, 3.1273e-04,  ..., 1.1286e-05, 7.7968e-05,\n        9.4935e-02], device='cuda:0')\ntensor([1.9031e-04, 4.2533e-04, 8.2979e-05,  ..., 1.4556e-05, 1.0986e-04,\n        1.0390e-01], device='cuda:0')\ntensor([2.2293e-04, 5.5215e-04, 1.1827e-04,  ..., 2.7792e-05, 1.7650e-04,\n        1.1379e-01], device='cuda:0')\ntensor([1.1749e-04, 3.6548e-04, 6.5611e-05,  ..., 1.1251e-05, 8.3300e-05,\n        1.0314e-01], device='cuda:0')\ntensor([1.2845e-04, 3.0125e-04, 5.5575e-05,  ..., 2.3751e-05, 1.5141e-04,\n        1.1039e-01], device='cuda:0')\ntensor([1.8839e-04, 5.2368e-04, 8.5561e-05,  ..., 4.3198e-05, 2.7689e-04,\n        1.1946e-01], device='cuda:0')\ntensor([1.8851e-04, 4.7860e-04, 8.0816e-05,  ..., 5.0494e-05, 2.8371e-04,\n        1.2051e-01], device='cuda:0')\ntensor([1.7881e-04, 4.8272e-04, 8.2949e-05,  ..., 8.0261e-05, 3.0749e-04,\n        1.1003e-01], device='cuda:0')\ntensor([1.3941e-04, 3.7394e-04, 8.1125e-05,  ..., 3.4296e-05, 1.7890e-04,\n        1.1938e-01], device='cuda:0')\ntensor([3.5054e-04, 8.1431e-04, 1.7419e-04,  ..., 4.0588e-05, 2.1048e-04,\n        1.1374e-01], device='cuda:0')\ntensor([2.1847e-04, 5.2473e-04, 9.8091e-05,  ..., 4.3905e-05, 2.5454e-04,\n        1.2283e-01], device='cuda:0')\n2019-05-19 16:49:36,787 Main INFO [Using bin1/2019-05-19-16-25-25/1best0.pth] done in 0.18769049644470215 s\ntry\ntensor([2.9465e-04, 1.3667e-03, 2.3224e-04,  ..., 1.7817e-05, 2.2510e-04,\n        1.4228e-01], device='cuda:0')\ntensor([6.5579e-05, 5.8624e-04, 6.4293e-05,  ..., 1.8043e-06, 6.4573e-05,\n        2.9756e-01], device='cuda:0')\ntensor([3.0046e-04, 1.1335e-03, 3.1733e-04,  ..., 6.9492e-07, 2.5570e-05,\n        1.9981e-02], device='cuda:0')\ntensor([1.0394e-04, 7.7769e-04, 8.8542e-05,  ..., 8.4002e-06, 1.9749e-04,\n        3.3166e-02], device='cuda:0')\ntensor([0.0003, 0.0007, 0.0002,  ..., 0.0001, 0.0013, 0.0546], device='cuda:0')\ntensor([1.7566e-04, 8.8524e-04, 1.7530e-04,  ..., 1.5727e-06, 4.1307e-05,\n        8.8085e-02], device='cuda:0')\ntensor([1.4017e-04, 4.7417e-04, 1.0219e-04,  ..., 1.9546e-05, 3.8961e-04,\n        2.5616e-01], device='cuda:0')\ntensor([3.3339e-04, 9.4354e-04, 4.0567e-04,  ..., 4.1636e-05, 8.2054e-04,\n        6.6367e-02], device='cuda:0')\ntensor([2.5083e-04, 7.7014e-04, 1.9185e-04,  ..., 4.6057e-06, 9.1608e-05,\n        2.0370e-02], device='cuda:0')\ntensor([3.3916e-05, 1.6992e-04, 1.8140e-05,  ..., 2.1272e-05, 3.6297e-04,\n        9.2996e-02], device='cuda:0')\ntensor([4.2618e-04, 1.2329e-03, 4.0827e-04,  ..., 9.5598e-06, 1.8138e-04,\n        1.0695e-01], device='cuda:0')\ntensor([5.9341e-05, 3.7739e-04, 4.0433e-05,  ..., 3.1512e-05, 6.5450e-04,\n        7.5206e-02], device='cuda:0')\ntensor([2.3491e-05, 9.7792e-05, 1.1311e-05,  ..., 3.5959e-06, 7.2879e-05,\n        3.9195e-02], device='cuda:0')\ntensor([3.0269e-05, 9.7369e-05, 1.5635e-05,  ..., 2.1093e-05, 2.6846e-04,\n        6.0795e-02], device='cuda:0')\ntensor([2.9270e-04, 7.2327e-04, 2.2392e-04,  ..., 3.6774e-05, 4.7130e-04,\n        5.9387e-02], device='cuda:0')\ntensor([1.0062e-04, 4.0396e-04, 7.2468e-05,  ..., 1.0871e-05, 2.4887e-04,\n        1.0341e-01], device='cuda:0')\ntensor([2.3391e-05, 1.3233e-04, 2.1068e-05,  ..., 3.9594e-06, 9.5903e-05,\n        1.5857e-01], device='cuda:0')\ntensor([3.4937e-05, 7.9389e-05, 1.5591e-05,  ..., 5.3740e-06, 7.8479e-05,\n        9.0914e-02], device='cuda:0')\ntensor([1.9867e-04, 1.3133e-03, 1.9977e-04,  ..., 2.2848e-06, 6.5746e-05,\n        1.3642e-02], device='cuda:0')\ntensor([3.8284e-05, 1.0898e-04, 2.5468e-05,  ..., 6.3947e-05, 7.9764e-04,\n        1.2654e-01], device='cuda:0')\ntensor([1.3873e-04, 9.3802e-04, 1.6678e-04,  ..., 2.7497e-05, 4.6714e-04,\n        1.3662e-01], device='cuda:0')\ntensor([2.4920e-04, 8.8052e-04, 1.7971e-04,  ..., 8.3159e-06, 2.5483e-04,\n        1.6300e-02], device='cuda:0')\ntensor([2.1528e-04, 1.0758e-03, 1.7343e-04,  ..., 1.3123e-06, 4.1932e-05,\n        2.6005e-02], device='cuda:0')\ntensor([4.1869e-05, 1.5919e-04, 3.1447e-05,  ..., 4.1329e-05, 5.9436e-04,\n        7.4010e-02], device='cuda:0')\ntensor([1.7581e-04, 7.2274e-04, 1.6086e-04,  ..., 2.7464e-06, 6.1626e-05,\n        3.2533e-02], device='cuda:0')\ntensor([5.4733e-04, 2.0119e-03, 5.7967e-04,  ..., 2.8432e-05, 4.3853e-04,\n        6.5093e-02], device='cuda:0')\ntensor([5.2515e-05, 1.0506e-04, 2.3848e-05,  ..., 4.8465e-07, 2.0914e-05,\n        2.8583e-02], device='cuda:0')\ntensor([1.5905e-04, 7.3170e-04, 1.6787e-04,  ..., 8.1118e-06, 1.6646e-04,\n        5.5315e-02], device='cuda:0')\ntensor([0.0002, 0.0006, 0.0002,  ..., 0.0001, 0.0011, 0.0452], device='cuda:0')\n","name":"stdout"},{"output_type":"stream","text":"tensor([1.9163e-04, 8.3461e-04, 2.5252e-04,  ..., 3.3172e-05, 4.5772e-04,\n        5.0903e-02], device='cuda:0')\ntensor([1.8435e-04, 1.3276e-03, 1.9494e-04,  ..., 3.0256e-07, 1.3414e-05,\n        1.2778e-02], device='cuda:0')\ntensor([0.0001, 0.0010, 0.0002,  ..., 0.0001, 0.0018, 0.0492], device='cuda:0')\ntensor([7.6155e-04, 1.4865e-03, 7.0449e-04,  ..., 1.3889e-05, 2.1724e-04,\n        1.3569e-01], device='cuda:0')\ntensor([5.1897e-05, 1.2862e-04, 2.4960e-05,  ..., 1.3672e-05, 2.2059e-04,\n        1.3430e-01], device='cuda:0')\ntensor([2.3301e-04, 6.4679e-04, 1.9782e-04,  ..., 1.3718e-06, 3.7137e-05,\n        6.2916e-02], device='cuda:0')\ntensor([5.4898e-05, 4.6627e-04, 6.1194e-05,  ..., 2.7543e-06, 9.4776e-05,\n        3.4110e-02], device='cuda:0')\ntensor([0.0002, 0.0008, 0.0001,  ..., 0.0001, 0.0018, 0.0502], device='cuda:0')\ntensor([3.1601e-05, 8.6750e-05, 1.8574e-05,  ..., 1.9647e-05, 2.7204e-04,\n        8.4344e-02], device='cuda:0')\ntensor([1.7159e-04, 4.3773e-04, 1.2030e-04,  ..., 7.9918e-06, 1.5422e-04,\n        1.0968e-01], device='cuda:0')\ntensor([3.7486e-04, 2.2232e-03, 3.7247e-04,  ..., 2.2581e-05, 3.6080e-04,\n        4.5923e-02], device='cuda:0')\ntensor([1.0556e-04, 6.6948e-04, 8.5545e-05,  ..., 5.9517e-06, 1.0747e-04,\n        2.6401e-02], device='cuda:0')\ntensor([1.0759e-04, 3.5435e-04, 9.9875e-05,  ..., 1.0658e-05, 1.4850e-04,\n        8.8453e-02], device='cuda:0')\ntensor([1.0037e-04, 2.3489e-04, 4.9824e-05,  ..., 4.8054e-06, 1.1734e-04,\n        1.3122e-02], device='cuda:0')\ntensor([2.0878e-05, 6.4770e-05, 1.0654e-05,  ..., 2.6015e-05, 4.5730e-04,\n        2.7345e-02], device='cuda:0')\ntensor([1.0733e-04, 3.9972e-04, 6.0871e-05,  ..., 8.0500e-06, 2.0520e-04,\n        3.7493e-02], device='cuda:0')\ntensor([3.8342e-04, 1.0646e-03, 2.7351e-04,  ..., 1.0198e-06, 4.1329e-05,\n        1.3008e-01], device='cuda:0')\ntensor([2.3832e-05, 8.4503e-05, 1.3110e-05,  ..., 5.8777e-06, 8.6142e-05,\n        5.6387e-02], device='cuda:0')\ntensor([2.6621e-04, 6.4860e-04, 1.9819e-04,  ..., 1.3518e-05, 1.5087e-04,\n        5.0587e-02], device='cuda:0')\ntensor([3.3909e-04, 8.8181e-04, 3.6163e-04,  ..., 9.7882e-07, 4.3385e-05,\n        4.4669e-02], device='cuda:0')\ntensor([2.9390e-05, 1.2119e-04, 2.5093e-05,  ..., 1.4403e-06, 5.4296e-05,\n        1.5058e-01], device='cuda:0')\ntensor([1.0117e-04, 3.1147e-04, 1.0262e-04,  ..., 2.3168e-06, 6.3199e-05,\n        2.6488e-02], device='cuda:0')\ntensor([1.6088e-04, 5.2592e-04, 1.5606e-04,  ..., 1.0101e-05, 1.7817e-04,\n        1.3266e-01], device='cuda:0')\ntensor([1.0352e-04, 7.8761e-04, 1.0541e-04,  ..., 3.3072e-05, 5.7547e-04,\n        9.4363e-02], device='cuda:0')\ntensor([1.2393e-04, 8.4125e-04, 1.1639e-04,  ..., 9.5060e-06, 1.6207e-04,\n        2.2449e-01], device='cuda:0')\ntensor([4.1574e-05, 1.1940e-04, 2.1313e-05,  ..., 1.2198e-05, 2.4882e-04,\n        3.1241e-01], device='cuda:0')\ntensor([4.3559e-05, 1.6912e-04, 3.7826e-05,  ..., 7.5398e-05, 6.0877e-04,\n        7.9569e-02], device='cuda:0')\ntensor([4.1130e-04, 1.8767e-03, 6.2460e-04,  ..., 7.4297e-06, 1.5377e-04,\n        9.7215e-02], device='cuda:0')\ntensor([1.4197e-04, 4.8370e-04, 1.0457e-04,  ..., 4.9994e-06, 1.0010e-04,\n        1.5642e-01], device='cuda:0')\ntensor([3.3039e-04, 6.0090e-04, 2.1256e-04,  ..., 3.4940e-05, 5.6029e-04,\n        1.2419e-01], device='cuda:0')\n2019-05-19 16:49:36,981 Main INFO [Using bin1/2019-05-19-16-25-25/1best4.pth] done in 0.19382596015930176 s\n2019-05-19 16:49:37,120 Main INFO [Using bin2/2019-05-19-16-25-25/2best1.pth] done in 0.13104844093322754 s\n2019-05-19 16:49:37,253 Main INFO [Using bin2/2019-05-19-16-25-25/2best2.pth] done in 0.12940645217895508 s\n2019-05-19 16:49:37,372 Main INFO [Using bin2/2019-05-19-16-25-25/2best3.pth] done in 0.11762404441833496 s\n2019-05-19 16:49:37,487 Main INFO [Using bin2/2019-05-19-16-25-25/2best4.pth] done in 0.11082792282104492 s\n2019-05-19 16:49:37,619 Main INFO [Using bin2/2019-05-19-16-25-25/2best0.pth] done in 0.13113999366760254 s\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"preds1 = (test_preds1 > best_threshold1).astype(int)\npreds2 = (test_preds2 > best_threshold2).astype(int)","execution_count":179,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"prediction = []\nfor i in range(preds1.shape[0]):\n    pred1 = [i for i in np.argwhere(preds1[i] == 1.0).reshape(-1).tolist() if i != (num_classes_c - 1)]\n    pred2 = [(i + num_classes_c - 1) for i in np.argwhere(preds2[i] == 1.0).reshape(-1).tolist() if i != (num_classes_c + num_classes_t - 2)]\n    pred_str = \" \".join(list(map(str, pred1 + pred2)))\n    prediction.append(pred_str)\n#print(test_preds1[test_preds1 != 0])\nsample.attribute_ids = prediction\nsample.to_csv(\"submission.csv\", index=False)\nsample.head()","execution_count":180,"outputs":[{"output_type":"stream","text":"[2.16447306e-04 8.33356660e-04 1.56778563e-04 ... 4.32497574e-05\n 4.06892644e-04 1.23885095e-01]\n","name":"stdout"},{"output_type":"execute_result","execution_count":180,"data":{"text/plain":"                 id                                      attribute_ids\n0  10023b2cc4ed5f68          13 121 147 189 369 671 813 1039 1059 1092\n1  100fbe75ed8fd887                   13 121 147 671 780 813 1059 1092\n2  101b627524a04f19                          121 147 189 813 1059 1092\n3  10234480c41284c6  13 51 147 189 480 483 501 671 737 738 776 813 ...\n4  1023b0e2636dcea8                51 79 147 189 671 780 813 1059 1092","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>id</th>\n      <th>attribute_ids</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>10023b2cc4ed5f68</td>\n      <td>13 121 147 189 369 671 813 1039 1059 1092</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>100fbe75ed8fd887</td>\n      <td>13 121 147 671 780 813 1059 1092</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>101b627524a04f19</td>\n      <td>121 147 189 813 1059 1092</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>10234480c41284c6</td>\n      <td>13 51 147 189 480 483 501 671 737 738 776 813 ...</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>1023b0e2636dcea8</td>\n      <td>51 79 147 189 671 780 813 1059 1092</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]}],"metadata":{"colab":{"name":"iMet.ipynb","provenance":[],"version":"0.3.2"},"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.7.3"}},"nbformat":4,"nbformat_minor":1}