13_rnn/07_ConvLSTM.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "view-in-github"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/machine-perception-robotics-group/MPRGDeepLearningLectureNotebook/blob/master/13_rnn/07_ConvLSTM.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rV0l0NgYXAeB"
      },
      "source": [
        "# Convolutional LSTMを用いた動画像予測\n",
        "\n",
        "---\n",
        "\n",
        "## Google ColabでGoogleドライブをマウントする\n",
        "* 以下のコードを実行するとURLが出てくるので，クリック\n",
        "* Googleアカウントに許可を求めたらcodeが出るのでコピー\n",
        "* コピーしたcodeを[空白]にペースト"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xPFFG2xhW_oD",
        "outputId": "1fa5d803-914a-427b-a854-137da8f140c2"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "G2BhrFGqFx_M"
      },
      "source": [
        "#ConvLSTM\n",
        "ConvLSTMはLSTMの内部構造に畳み込み層を導入した再帰型のネットワークです．LSTMの内部に畳み込み層を導入することで，動画像を順伝播逆伝播できるネットワークとなっています．このネットワークは，時系列データを入力として受け取り，中間層のユニットの出力が自分自身に戻る「帰還路」を持つため，過去の情報を保持しやすい構造となっています．今回はMoving Mnistを使って動画予測を行います．\n",
        "\n",
        "\n",
        "<img src=\"https://github.com/himidev/Lecture/blob/main/13_rnn/07_ConvLSTM/convLSTM.png?raw=true\" width = 70%>\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yfWe3l_76ZFQ"
      },
      "source": [
        "# モジュールのインポート"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "YSfImhoulTMI"
      },
      "outputs": [],
      "source": [
        "import gzip\n",
        "import math\n",
        "import os\n",
        "import random\n",
        "import sys\n",
        "import logging\n",
        "import numpy as np\n",
        "from PIL import Image\n",
        "from PIL import Image\n",
        "from tqdm import tqdm\n",
        "from collections import OrderedDict\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "\n",
        "import torch\n",
        "import torch.utils.data as data\n",
        "import torch.optim as optim\n",
        "import torch.nn.functional as F\n",
        "from torch.optim import lr_scheduler\n",
        "from torch import nn\n",
        "from collections import OrderedDict"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v6R8FbCFMdAC"
      },
      "source": [
        "# MovingMNIST"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "wcO1rV6vMTGj"
      },
      "outputs": [],
      "source": [
        "!wget -q \"https://drive.google.com/uc?export=download&id=1lhue3SAroqRtLgPbQknRhfMClCR45qik\" -O train-images-idx3-ubyte.gz"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_DYa9zXLXNFC"
      },
      "source": [
        "# Data loader\n",
        "Moving Mnistの読み込みを行います．このときバッチ毎にシーケンス数が同じになるように“class MovingMNIST\"内で処理を行います．MovingMNISTのリンク先[https://drive.google.com/file/d/1lhue3SAroqRtLgPbQknRhfMClCR45qik/view?usp=sharing]． "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "HxXx6PLGWB8h"
      },
      "outputs": [],
      "source": [
        "def load_mnist(root):\n",
        "    path = os.path.join(root)\n",
        "    with gzip.open(path, 'rb') as f:\n",
        "        mnist = np.frombuffer(f.read(), np.uint8, offset=16)\n",
        "        mnist = mnist.reshape(-1, 28, 28)\n",
        "    return mnist\n",
        "\n",
        "\n",
        "def load_fixed_set(root, is_train):\n",
        "    filename = 'mnist_test_seq.npy'\n",
        "    path = os.path.join(root, filename)\n",
        "    dataset = np.load(path)\n",
        "    dataset = dataset[..., np.newaxis]\n",
        "    return dataset\n",
        "\n",
        "\n",
        "class MovingMNIST(data.Dataset):\n",
        "    def __init__(self, root, is_train, n_frames_input, n_frames_output, num_objects,\n",
        "                 transform=None):\n",
        "        super(MovingMNIST, self).__init__()\n",
        "\n",
        "        self.dataset = None\n",
        "        if is_train:\n",
        "            self.mnist = load_mnist(root)\n",
        "        else:\n",
        "            if num_objects[0] != 2:\n",
        "                self.mnist = load_mnist(root)\n",
        "            else:\n",
        "                self.dataset = load_fixed_set(root, False)\n",
        "        self.length = int(1e4) if self.dataset is None else self.dataset.shape[1]\n",
        "\n",
        "        self.is_train = is_train\n",
        "        self.num_objects = num_objects\n",
        "        self.n_frames_input = n_frames_input\n",
        "        self.n_frames_output = n_frames_output\n",
        "        self.n_frames_total = self.n_frames_input + self.n_frames_output\n",
        "        self.transform = transform\n",
        "        self.image_size_ = 64\n",
        "        self.digit_size_ = 28\n",
        "        self.step_length_ = 0.1\n",
        "\n",
        "    def get_random_trajectory(self, seq_length):\n",
        "        canvas_size = self.image_size_ - self.digit_size_\n",
        "        x = random.random()\n",
        "        y = random.random()\n",
        "        theta = random.random() * 2 * np.pi\n",
        "        v_y = np.sin(theta)\n",
        "        v_x = np.cos(theta)\n",
        "\n",
        "        start_y = np.zeros(seq_length)\n",
        "        start_x = np.zeros(seq_length)\n",
        "        for i in range(seq_length):\n",
        "            y += v_y * self.step_length_\n",
        "            x += v_x * self.step_length_\n",
        "\n",
        "            if x <= 0:\n",
        "                x = 0\n",
        "                v_x = -v_x\n",
        "            if x >= 1.0:\n",
        "                x = 1.0\n",
        "                v_x = -v_x\n",
        "            if y <= 0:\n",
        "                y = 0\n",
        "                v_y = -v_y\n",
        "            if y >= 1.0:\n",
        "                y = 1.0\n",
        "                v_y = -v_y\n",
        "            start_y[i] = y\n",
        "            start_x[i] = x\n",
        "\n",
        "        start_y = (canvas_size * start_y).astype(np.int32)\n",
        "        start_x = (canvas_size * start_x).astype(np.int32)\n",
        "        return start_y, start_x\n",
        "\n",
        "    def generate_moving_mnist(self, num_digits=2):\n",
        "        data = np.zeros((self.n_frames_total, self.image_size_, self.image_size_), dtype=np.float32)\n",
        "        for n in range(num_digits):\n",
        "            start_y, start_x = self.get_random_trajectory(self.n_frames_total)\n",
        "            ind = random.randint(0, self.mnist.shape[0] - 1)\n",
        "            digit_image = self.mnist[ind]\n",
        "            for i in range(self.n_frames_total):\n",
        "                top = start_y[i]\n",
        "                left = start_x[i]\n",
        "                bottom = top + self.digit_size_\n",
        "                right = left + self.digit_size_\n",
        "                data[i, top:bottom, left:right] = np.maximum(data[i, top:bottom, left:right], digit_image)\n",
        "\n",
        "        data = data[..., np.newaxis]\n",
        "        return data\n",
        "\n",
        "    def __getitem__(self, idx):\n",
        "        length = self.n_frames_input + self.n_frames_output\n",
        "        if self.is_train or self.num_objects[0] != 2:\n",
        "            num_digits = random.choice(self.num_objects)\n",
        "            images = self.generate_moving_mnist(num_digits)\n",
        "        else:\n",
        "            images = self.dataset[:, idx, ...]\n",
        "\n",
        "\n",
        "        r = 1\n",
        "        w = int(64 / r)\n",
        "        images = images.reshape((length, w, r, w, r)).transpose(0, 2, 4, 1, 3).reshape((length, r * r, w, w))\n",
        "\n",
        "        input = images[:self.n_frames_input]\n",
        "        if self.n_frames_output > 0:\n",
        "            output = images[self.n_frames_input:length]\n",
        "        else:\n",
        "            output = []\n",
        "\n",
        "        frozen = input[-1]\n",
        "\n",
        "        output = torch.from_numpy(output / 255.0).contiguous().float()\n",
        "        input = torch.from_numpy(input / 255.0).contiguous().float()\n",
        "\n",
        "        out = [idx, output, input, frozen, np.zeros(1)]\n",
        "        return out\n",
        "\n",
        "    def __len__(self):\n",
        "        return self.length"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Kqxw5Fg253Mb"
      },
      "source": [
        "# 学習パラメータの設定やデータの読み込み\n",
        " 各パラメータを設定します．本実験では20フレームを用い，観測10フレームの動画像を入力し，その後の10フレームの予測動画像を出力します．MovingMNIST内のrootで，driveに置いてあるファイルパスを指定します．このとき，学習用と評価用にsplitしたファイルが各Folderです．\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xHySon9BVa2S"
      },
      "outputs": [],
      "source": [
        "frame_input = 10\n",
        "frame_output = 10\n",
        "batch = 8\n",
        "lr = 0.001\n",
        "epochs = 10\n",
        "\n",
        "trainFolder = MovingMNIST(is_train=True,\n",
        "                          #root='/content/drive/My Drive/Colab_Notebooks/train-images-idx3-ubyte.gz',\n",
        "                          root='./train-images-idx3-ubyte.gz',\n",
        "                          n_frames_input=frame_input,\n",
        "                          n_frames_output=frame_output,\n",
        "                          num_objects=[3])\n",
        "validFolder = MovingMNIST(is_train=False,\n",
        "                          #root='/content/drive/My Drive/Colab_Notebooks/train-images-idx3-ubyte.gz',\n",
        "                          root='./train-images-idx3-ubyte.gz',\n",
        "                          n_frames_input=frame_input,\n",
        "                          n_frames_output=frame_output,\n",
        "                          num_objects=[3])\n",
        "\n",
        "trainLoader = torch.utils.data.DataLoader(trainFolder,\n",
        "                                          batch_size=batch,\n",
        "                                          shuffle=False)\n",
        "validLoader = torch.utils.data.DataLoader(validFolder,\n",
        "                                          batch_size=batch,\n",
        "                                          shuffle=False)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8tkaZX-pJw9g"
      },
      "source": [
        "# Layer作成\n",
        "Encoder，Decoderで指定されたlayer name毎に処理を行います．"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ECQgo-DJkMSF"
      },
      "outputs": [],
      "source": [
        "def make_layers(block):\n",
        "    layers = []\n",
        "    for layer_name, v in block.items():\n",
        "        if 'pool' in layer_name:\n",
        "            layer = nn.MaxPool2d(kernel_size=v[0], stride=v[1], padding=v[2])\n",
        "            layers.append((layer_name, layer))\n",
        "        elif 'deconv' in layer_name:\n",
        "            transposeConv2d = nn.ConvTranspose2d(in_channels=v[0],\n",
        "                                                 out_channels=v[1],\n",
        "                                                 kernel_size=v[2],\n",
        "                                                 stride=v[3],\n",
        "                                                 padding=v[4])\n",
        "            layers.append((layer_name, transposeConv2d))\n",
        "            if 'relu' in layer_name:\n",
        "                layers.append(('relu_' + layer_name, nn.ReLU(inplace=True)))\n",
        "            elif 'leaky' in layer_name:\n",
        "                layers.append(('leaky_' + layer_name,\n",
        "                               nn.LeakyReLU(negative_slope=0.2, inplace=True)))\n",
        "        elif 'conv' in layer_name:\n",
        "            conv2d = nn.Conv2d(in_channels=v[0],\n",
        "                               out_channels=v[1],\n",
        "                               kernel_size=v[2],\n",
        "                               stride=v[3],\n",
        "                               padding=v[4])\n",
        "            layers.append((layer_name, conv2d))\n",
        "            if 'relu' in layer_name:\n",
        "                layers.append(('relu_' + layer_name, nn.ReLU(inplace=True)))\n",
        "            elif 'leaky' in layer_name:\n",
        "                layers.append(('leaky_' + layer_name,\n",
        "                               nn.LeakyReLU(negative_slope=0.2, inplace=True)))\n",
        "        else:\n",
        "            raise NotImplementedError\n",
        "    return nn.Sequential(OrderedDict(layers))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OWjft0JKj6jS"
      },
      "source": [
        "#Encoder\n",
        "Moving Mnistの画像特徴量をConvolutionして捉えます．このとき，各時刻で得られる画像特徴量をConvolutional LSTMで時間方向に伝播することで，動画像における特徴量を求めます．"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "hF6_l1dxj6AA"
      },
      "outputs": [],
      "source": [
        "class Encoder(nn.Module):\n",
        "    def __init__(self, subnets, rnns):\n",
        "        super().__init__()\n",
        "        assert len(subnets) == len(rnns)\n",
        "        self.blocks = len(subnets)\n",
        "\n",
        "        for index, (params, rnn) in enumerate(zip(subnets, rnns), 1):\n",
        "            setattr(self, 'stage' + str(index), make_layers(params))\n",
        "            setattr(self, 'rnn' + str(index), rnn)\n",
        "\n",
        "    def forward_by_stage(self, inputs, subnet, rnn):\n",
        "        seq_number, batch_size, input_channel, height, width = inputs.size()\n",
        "        inputs = torch.reshape(inputs, (-1, input_channel, height, width))\n",
        "        # convolution\n",
        "        # to BxS, C, H, W\n",
        "        inputs = subnet(inputs)\n",
        "        inputs = torch.reshape(inputs, (seq_number, batch_size, inputs.size(1),\n",
        "                                        inputs.size(2), inputs.size(3)))\n",
        "        # recurrent\n",
        "        outputs_stage, state_stage = rnn(inputs, None)\n",
        "        # to B, S, C, H, W\n",
        "        return outputs_stage, state_stage\n",
        "\n",
        "    def forward(self, inputs):\n",
        "        inputs = inputs.transpose(0, 1)  # to S,B,1,H,W\n",
        "        hidden_states = []\n",
        "        logging.debug(inputs.size())\n",
        "        # 層数で繰り返す\n",
        "        for i in range(1, self.blocks + 1):\n",
        "            inputs, state_stage = self.forward_by_stage(\n",
        "                inputs, getattr(self, 'stage' + str(i)),\n",
        "                getattr(self, 'rnn' + str(i)))\n",
        "            hidden_states.append(state_stage)\n",
        "        return tuple(hidden_states)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hVVJa1CPkAqf"
      },
      "source": [
        "#Decoder\n",
        "畳み込んだ特徴量をDeconvolutionします．このとき，各時刻で得られる画像特徴量をDeconvolutional LSTMで時間方向に伝播しつつ，元の動画像サイズまで復元します．"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "fHU9DokFkALg"
      },
      "outputs": [],
      "source": [
        "class Decoder(nn.Module):\n",
        "    def __init__(self, subnets, rnns):\n",
        "        super().__init__()\n",
        "        assert len(subnets) == len(rnns)\n",
        "\n",
        "        self.blocks = len(subnets)\n",
        "\n",
        "        for index, (params, rnn) in enumerate(zip(subnets, rnns)):\n",
        "            setattr(self, 'rnn' + str(self.blocks - index), rnn)\n",
        "            setattr(self, 'stage' + str(self.blocks - index),\n",
        "                    make_layers(params))\n",
        "\n",
        "    def forward_by_stage(self, inputs, state, subnet, rnn):\n",
        "        # recurrent\n",
        "        # to B, C, H, W\n",
        "        inputs, state_stage = rnn(inputs, state, seq_len=10)\n",
        "        # to S, B, C, H, W\n",
        "        seq_number, batch_size, input_channel, height, width = inputs.size()\n",
        "        inputs = torch.reshape(inputs, (-1, input_channel, height, width))\n",
        "        # deconvolution\n",
        "        inputs = subnet(inputs)\n",
        "        # to B*S, C, H, W\n",
        "        inputs = torch.reshape(inputs, (seq_number, batch_size, inputs.size(1),\n",
        "                                        inputs.size(2), inputs.size(3)))\n",
        "        return inputs\n",
        "\n",
        "    def forward(self, hidden_states):\n",
        "        # Decoderの入力はEncoderの最後の内部状態のみ\n",
        "        inputs = self.forward_by_stage(None, hidden_states[-1],\n",
        "                                       getattr(self, 'stage3'),\n",
        "                                       getattr(self, 'rnn3'))\n",
        "        for i in list(range(1, self.blocks))[::-1]:\n",
        "            inputs = self.forward_by_stage(inputs, hidden_states[i - 1],\n",
        "                                           getattr(self, 'stage' + str(i)),\n",
        "                                           getattr(self, 'rnn' + str(i)))\n",
        "        inputs = inputs.transpose(0, 1)  # to B,S,1,H, W\n",
        "        return inputs\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eI2Q7j_rKnrP"
      },
      "source": [
        "# Encoder-Decoderに各入力を渡す"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "IeT5G3ITdYRX"
      },
      "outputs": [],
      "source": [
        "class activation():\n",
        "\n",
        "    def __init__(self, act_type, negative_slope=0.2, inplace=True):\n",
        "        super().__init__()\n",
        "        self._act_type = act_type\n",
        "        self.negative_slope = negative_slope\n",
        "        self.inplace = inplace\n",
        "\n",
        "    def __call__(self, input):\n",
        "        if self._act_type == 'leaky':\n",
        "            return F.leaky_relu(input, negative_slope=self.negative_slope, inplace=self.inplace)\n",
        "        elif self._act_type == 'relu':\n",
        "            return F.relu(input, inplace=self.inplace)\n",
        "        elif self._act_type == 'sigmoid':\n",
        "            return torch.sigmoid(input)\n",
        "        else:\n",
        "            raise NotImplementedError\n",
        "\n",
        "\n",
        "class ED(nn.Module):\n",
        "\n",
        "    def __init__(self, encoder, decoder):\n",
        "        super().__init__()\n",
        "        self.encoder = encoder\n",
        "        self.decoder = decoder\n",
        "\n",
        "    def forward(self, input):\n",
        "        state = self.encoder(input)\n",
        "        output = self.decoder(state)\n",
        "        return output"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8wenPUf1XRSb"
      },
      "source": [
        "# Convolution LSTM\n",
        "動画像における特徴量を求めます．Encoder-Decoderへの引数としてencoder_paramsとdecoder_paramsを定義します．Conv.LSTM層では，それぞれConvolutionまたはDeconvolution処理を行います．"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "J584vKaXbp37"
      },
      "outputs": [],
      "source": [
        "class CLSTM_cell(nn.Module):\n",
        "    \"\"\"ConvLSTMCell\n",
        "    \"\"\"\n",
        "    def __init__(self, shape, input_channels, filter_size, num_features):\n",
        "        super(CLSTM_cell, self).__init__()\n",
        "\n",
        "        self.shape = shape  # H, W\n",
        "        self.input_channels = input_channels\n",
        "        self.filter_size = filter_size\n",
        "        self.num_features = num_features\n",
        "        self.padding = (filter_size - 1) // 2\n",
        "        self.conv = nn.Sequential(\n",
        "            nn.Conv2d(self.input_channels + self.num_features,\n",
        "                      4 * self.num_features, self.filter_size, 1,\n",
        "                      self.padding),\n",
        "            nn.GroupNorm(4 * self.num_features // 32, 4 * self.num_features))\n",
        "        \n",
        "    # LSTMの計算を畳み込み演算に変えるのみでできます．\n",
        "    # ただ，pytorchにはconv.lstmは実装されてないため自身で構築する必要があります．\n",
        "    def forward(self, inputs=None, hidden_state=None, seq_len=10):\n",
        "        # LSTMでいう初期内部状態\n",
        "        if hidden_state is None:\n",
        "            hx = torch.zeros(inputs.size(1), self.num_features, self.shape[0],\n",
        "                             self.shape[1]).cuda()\n",
        "            cx = torch.zeros(inputs.size(1), self.num_features, self.shape[0],\n",
        "                             self.shape[1]).cuda()\n",
        "        else:\n",
        "            hx, cx = hidden_state\n",
        "        output_inner = []\n",
        "        # sequentialに内部状態を更新\n",
        "        # LSTMと同じ計算\n",
        "        for index in range(seq_len):\n",
        "            if inputs is None:\n",
        "                x = torch.zeros(hx.size(0), self.input_channels, self.shape[0],\n",
        "                                self.shape[1]).cuda()\n",
        "            else:\n",
        "                x = inputs[index, ...]\n",
        "\n",
        "            combined = torch.cat((x, hx), 1)\n",
        "            gates = self.conv(combined)\n",
        "            ingate, forgetgate, cellgate, outgate = torch.split(\n",
        "                gates, self.num_features, dim=1)\n",
        "            ingate = torch.sigmoid(ingate)\n",
        "            forgetgate = torch.sigmoid(forgetgate)\n",
        "            cellgate = torch.tanh(cellgate)\n",
        "            outgate = torch.sigmoid(outgate)\n",
        "\n",
        "            cy = (forgetgate * cx) + (ingate * cellgate)\n",
        "            hy = outgate * torch.tanh(cy)\n",
        "            output_inner.append(hy)\n",
        "            hx = hy\n",
        "            cx = cy\n",
        "        return torch.stack(output_inner), (hy, cy)\n",
        "\n",
        "\n",
        "from collections import OrderedDict\n",
        "encoder_params = [\n",
        "    [\n",
        "        OrderedDict({'conv1_leaky_1': [1, 16, 3, 1, 1]}),\n",
        "        OrderedDict({'conv2_leaky_1': [64, 64, 3, 2, 1]}),\n",
        "        OrderedDict({'conv3_leaky_1': [96, 96, 3, 2, 1]}),\n",
        "    ],\n",
        "\n",
        "    [\n",
        "        CLSTM_cell(shape=(64,64), input_channels=16, filter_size=5, num_features=64),\n",
        "        CLSTM_cell(shape=(32,32), input_channels=64, filter_size=5, num_features=96),\n",
        "        CLSTM_cell(shape=(16,16), input_channels=96, filter_size=5, num_features=96)\n",
        "    ]\n",
        "]\n",
        "\n",
        "decoder_params = [\n",
        "    [\n",
        "        OrderedDict({'deconv1_leaky_1': [96, 96, 4, 2, 1]}),\n",
        "        OrderedDict({'deconv2_leaky_1': [96, 96, 4, 2, 1]}),\n",
        "        OrderedDict({\n",
        "            'conv3_leaky_1': [64, 16, 3, 1, 1],\n",
        "            'conv4_leaky_1': [16, 1, 1, 1, 0]\n",
        "        }),\n",
        "    ],\n",
        "\n",
        "    [\n",
        "        CLSTM_cell(shape=(16,16), input_channels=96, filter_size=5, num_features=96),\n",
        "        CLSTM_cell(shape=(32,32), input_channels=96, filter_size=5, num_features=96),\n",
        "        CLSTM_cell(shape=(64,64), input_channels=96, filter_size=5, num_features=64),\n",
        "    ]\n",
        "]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "T5Y50zBJXVCo"
      },
      "source": [
        "# モデル定義\n",
        "EncoderとDecoderに各パラメータを引数として入れます．損失関数にはMSELossを用います．最適化手法にはAdamを使用します．"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "SQdkgeqLUE3z"
      },
      "outputs": [],
      "source": [
        "encoder = Encoder(encoder_params[0], encoder_params[1]).cuda()\n",
        "decoder = Decoder(decoder_params[0], decoder_params[1]).cuda()\n",
        "net = ED(encoder, decoder)\n",
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
        "\n",
        "if torch.cuda.device_count() > 1:\n",
        "    net = nn.DataParallel(net)\n",
        "net.to(device)\n",
        "lossfunction = nn.MSELoss().cuda()\n",
        "optimizer = optim.Adam(net.parameters(), lr=lr)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XUumnymHb6lq"
      },
      "source": [
        "#学習\n",
        "学習と検証用データを回します．動画特徴量を捉えるため，エポック数を少々増やさないとうまく学習されないのと，学習に3日前後使うので，注意してください．"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "--Gm99Q9bhNe"
      },
      "outputs": [],
      "source": [
        "train_losses = []\n",
        "valid_losses = []\n",
        "avg_train_losses = []\n",
        "avg_valid_losses = []\n",
        "for epoch in range(0, epochs + 1):\n",
        "    ######################\n",
        "    # 学習\n",
        "    ######################\n",
        "    t = tqdm(trainLoader, leave=False, total=len(trainLoader))\n",
        "    train_loss = 0\n",
        "    test_loss = 0\n",
        "\n",
        "    for i, (idx, targetVar, inputVar, _, _) in enumerate(t):\n",
        "        inputs = inputVar.to(device)  # B,S,C,H,W\n",
        "        label = targetVar.to(device)  # B,S,C,H,W\n",
        "        optimizer.zero_grad()\n",
        "        net.train()\n",
        "        pred = net(inputs)  # B,S,C,H,W\n",
        "        loss = lossfunction(pred, label)\n",
        "        loss_aver = loss.item() / batch\n",
        "        train_loss += loss_aver\n",
        "        loss.backward()\n",
        "        torch.nn.utils.clip_grad_value_(net.parameters(), clip_value=10.0)\n",
        "        optimizer.step()\n",
        "\n",
        "        t.set_postfix({\n",
        "            'trainloss': '{:.6f}'.format(loss_aver),\n",
        "            'epoch': '{:02d}'.format(epoch)\n",
        "        })\n",
        "    # モデルを保存    \n",
        "    # torch.save(net.state_dict(), 'pre-train.pth')\n",
        "    \n",
        "    ######################\n",
        "    # 検証\n",
        "    ######################\n",
        "    with torch.no_grad():\n",
        "        net.eval()\n",
        "        t = tqdm(validLoader, leave=False, total=len(validLoader))\n",
        "        for i, (idx, targetVar, inputVar, _, _) in enumerate(t):\n",
        "            if i == 3000:\n",
        "                break\n",
        "            inputs = inputVar.to(device)\n",
        "            label = targetVar.to(device)\n",
        "            pred = net(inputs)\n",
        "            loss = lossfunction(pred, label)\n",
        "            loss_aver = loss.item() / batch\n",
        "            test_loss += loss_aver\n",
        "            valid_losses.append(loss_aver)\n",
        "            t.set_postfix({\n",
        "                'validloss': '{:.6f}'.format(loss_aver),\n",
        "                'epoch': '{:02d}'.format(epoch)\n",
        "            })\n",
        "    print('%03d-epoch: train_loss %.5f : val_loss %.5f' %(epoch, train_loss/len(t), test_loss/len(t)))\n",
        "    torch.cuda.empty_cache()\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_fDK9A2cLac1"
      },
      "source": [
        "# Pre-trained model\n",
        "学習時間が長いので，学習済みモデルをインポートします．学習から始める場合は，学習のモデルを保存のコメントアウトを外します．"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Kf9BBqjgLdo9",
        "outputId": "91e23e5c-fc72-4e13-8277-f5dfc67f6239"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<All keys matched successfully>"
            ]
          },
          "execution_count": 12,
          "metadata": {
            "tags": []
          },
          "output_type": "execute_result"
        }
      ],
      "source": [
        "!wget -q \"https://drive.google.com/uc?export=download&id=1geXhoW6ij1JodjcOOqHTgYVjdFGbSraI\" -O pre-train.pth\n",
        "net.load_state_dict(torch.load(\"./pre-train.pth\"))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PV0szBcJVQc_"
      },
      "source": [
        "# 評価\n",
        "予測した動画像と真値の動画像を比較します．損失計算はMSEで計算します．"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tlsS_E9FU4w6",
        "outputId": "856ad2fa-1dd6-48f6-b08a-8df7e4cf5d6c"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\u001b[1;30;43mStreaming output truncated to the last 5000 lines.\u001b[0m\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 20%|██        | 252/1250 [01:20<05:14,  3.18it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 20%|██        | 253/1250 [01:21<05:14,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 20%|██        | 254/1250 [01:21<05:13,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 20%|██        | 255/1250 [01:21<05:13,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 20%|██        | 256/1250 [01:22<05:13,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██        | 257/1250 [01:22<05:12,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██        | 258/1250 [01:22<05:13,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██        | 259/1250 [01:23<05:12,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██        | 260/1250 [01:23<05:12,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██        | 261/1250 [01:23<05:12,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██        | 262/1250 [01:23<05:11,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██        | 263/1250 [01:24<05:11,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██        | 264/1250 [01:24<05:11,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██        | 265/1250 [01:24<05:10,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██▏       | 266/1250 [01:25<05:10,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██▏       | 267/1250 [01:25<05:09,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 21%|██▏       | 268/1250 [01:25<05:09,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 269/1250 [01:26<05:09,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 270/1250 [01:26<05:09,  3.17it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 271/1250 [01:26<05:09,  3.16it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 272/1250 [01:27<05:09,  3.16it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 273/1250 [01:27<05:09,  3.16it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 274/1250 [01:27<05:09,  3.16it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 275/1250 [01:28<05:08,  3.16it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 276/1250 [01:28<05:08,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 277/1250 [01:28<05:08,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 278/1250 [01:29<05:08,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 279/1250 [01:29<05:08,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 280/1250 [01:29<05:08,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 22%|██▏       | 281/1250 [01:29<05:07,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 282/1250 [01:30<05:07,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 283/1250 [01:30<05:06,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 284/1250 [01:30<05:06,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 285/1250 [01:31<05:07,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 286/1250 [01:31<05:07,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 287/1250 [01:31<05:05,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 288/1250 [01:32<05:05,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 289/1250 [01:32<05:05,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 290/1250 [01:32<05:04,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 291/1250 [01:33<05:04,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 292/1250 [01:33<05:04,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 23%|██▎       | 293/1250 [01:33<05:04,  3.15it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▎       | 294/1250 [01:34<05:04,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▎       | 295/1250 [01:34<05:03,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▎       | 296/1250 [01:34<05:06,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▍       | 297/1250 [01:35<05:05,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▍       | 298/1250 [01:35<05:04,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▍       | 299/1250 [01:35<05:03,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▍       | 300/1250 [01:36<05:04,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▍       | 301/1250 [01:36<05:03,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▍       | 302/1250 [01:36<05:03,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▍       | 303/1250 [01:36<05:02,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▍       | 304/1250 [01:37<05:03,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▍       | 305/1250 [01:37<05:03,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 24%|██▍       | 306/1250 [01:37<05:03,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▍       | 307/1250 [01:38<05:02,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▍       | 308/1250 [01:38<05:01,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▍       | 309/1250 [01:38<05:00,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▍       | 310/1250 [01:39<05:01,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▍       | 311/1250 [01:39<05:02,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▍       | 312/1250 [01:39<05:01,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▌       | 313/1250 [01:40<05:00,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▌       | 314/1250 [01:40<04:59,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▌       | 315/1250 [01:40<04:59,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▌       | 316/1250 [01:41<04:59,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▌       | 317/1250 [01:41<05:00,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 25%|██▌       | 318/1250 [01:41<05:00,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▌       | 319/1250 [01:42<04:59,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▌       | 320/1250 [01:42<04:58,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▌       | 321/1250 [01:42<04:57,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▌       | 322/1250 [01:43<04:57,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▌       | 323/1250 [01:43<04:57,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▌       | 324/1250 [01:43<04:58,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▌       | 325/1250 [01:44<04:57,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▌       | 326/1250 [01:44<04:56,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▌       | 327/1250 [01:44<04:56,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▌       | 328/1250 [01:45<04:55,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▋       | 329/1250 [01:45<04:55,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▋       | 330/1250 [01:45<04:56,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 26%|██▋       | 331/1250 [01:45<04:55,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 332/1250 [01:46<04:56,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 333/1250 [01:46<04:57,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 334/1250 [01:46<04:56,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 335/1250 [01:47<04:56,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 336/1250 [01:47<04:56,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 337/1250 [01:47<04:55,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 338/1250 [01:48<04:55,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 339/1250 [01:48<04:55,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 340/1250 [01:48<04:55,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 341/1250 [01:49<04:54,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 342/1250 [01:49<04:55,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 27%|██▋       | 343/1250 [01:49<04:54,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 344/1250 [01:50<04:53,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 345/1250 [01:50<04:54,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 346/1250 [01:50<04:53,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 347/1250 [01:51<04:52,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 348/1250 [01:51<04:52,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 349/1250 [01:51<04:51,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 350/1250 [01:52<04:51,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 351/1250 [01:52<04:51,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 352/1250 [01:52<04:51,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 353/1250 [01:53<04:50,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 354/1250 [01:53<04:50,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 355/1250 [01:53<04:50,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 28%|██▊       | 356/1250 [01:54<04:50,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▊       | 357/1250 [01:54<04:49,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▊       | 358/1250 [01:54<04:49,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▊       | 359/1250 [01:55<04:49,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▉       | 360/1250 [01:55<04:48,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▉       | 361/1250 [01:55<04:48,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▉       | 362/1250 [01:56<04:47,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▉       | 363/1250 [01:56<04:47,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▉       | 364/1250 [01:56<04:47,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▉       | 365/1250 [01:57<04:47,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▉       | 366/1250 [01:57<04:46,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▉       | 367/1250 [01:57<04:46,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 29%|██▉       | 368/1250 [01:57<04:46,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|██▉       | 369/1250 [01:58<04:46,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|██▉       | 370/1250 [01:58<04:45,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|██▉       | 371/1250 [01:58<04:45,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|██▉       | 372/1250 [01:59<04:44,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|██▉       | 373/1250 [01:59<04:44,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|██▉       | 374/1250 [01:59<04:43,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|███       | 375/1250 [02:00<04:43,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|███       | 376/1250 [02:00<04:43,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|███       | 377/1250 [02:00<04:42,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|███       | 378/1250 [02:01<04:42,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|███       | 379/1250 [02:01<04:42,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|███       | 380/1250 [02:01<04:41,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 30%|███       | 381/1250 [02:02<04:41,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███       | 382/1250 [02:02<04:41,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███       | 383/1250 [02:02<04:41,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███       | 384/1250 [02:03<04:41,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███       | 385/1250 [02:03<04:40,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███       | 386/1250 [02:03<04:40,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███       | 387/1250 [02:04<04:39,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███       | 388/1250 [02:04<04:40,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███       | 389/1250 [02:04<04:39,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███       | 390/1250 [02:05<04:38,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███▏      | 391/1250 [02:05<04:39,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███▏      | 392/1250 [02:05<04:38,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 31%|███▏      | 393/1250 [02:06<04:37,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 394/1250 [02:06<04:37,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 395/1250 [02:06<04:37,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 396/1250 [02:07<04:36,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 397/1250 [02:07<04:35,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 398/1250 [02:07<04:34,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 399/1250 [02:08<04:34,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 400/1250 [02:08<04:32,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 401/1250 [02:08<04:32,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 402/1250 [02:09<04:32,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 403/1250 [02:09<04:32,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 404/1250 [02:09<04:33,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 405/1250 [02:09<04:33,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 32%|███▏      | 406/1250 [02:10<04:32,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 407/1250 [02:10<04:33,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 408/1250 [02:10<04:32,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 409/1250 [02:11<04:31,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 410/1250 [02:11<04:30,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 411/1250 [02:11<04:29,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 412/1250 [02:12<04:28,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 413/1250 [02:12<04:28,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 414/1250 [02:12<04:28,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 415/1250 [02:13<04:28,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 416/1250 [02:13<04:28,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 417/1250 [02:13<04:29,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 33%|███▎      | 418/1250 [02:14<04:28,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▎      | 419/1250 [02:14<04:27,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▎      | 420/1250 [02:14<04:25,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▎      | 421/1250 [02:15<04:26,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▍      | 422/1250 [02:15<04:26,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▍      | 423/1250 [02:15<04:26,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▍      | 424/1250 [02:16<04:26,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▍      | 425/1250 [02:16<04:26,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▍      | 426/1250 [02:16<04:26,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▍      | 427/1250 [02:17<04:25,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▍      | 428/1250 [02:17<04:24,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▍      | 429/1250 [02:17<04:23,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▍      | 430/1250 [02:18<04:22,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 34%|███▍      | 431/1250 [02:18<04:22,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▍      | 432/1250 [02:18<04:23,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▍      | 433/1250 [02:18<04:23,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▍      | 434/1250 [02:19<04:23,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▍      | 435/1250 [02:19<04:24,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▍      | 436/1250 [02:19<04:24,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▍      | 437/1250 [02:20<04:22,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▌      | 438/1250 [02:20<04:24,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▌      | 439/1250 [02:20<04:22,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▌      | 440/1250 [02:21<04:22,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▌      | 441/1250 [02:21<04:22,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▌      | 442/1250 [02:21<04:21,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 35%|███▌      | 443/1250 [02:22<04:20,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▌      | 444/1250 [02:22<04:19,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▌      | 445/1250 [02:22<04:18,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▌      | 446/1250 [02:23<04:17,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▌      | 447/1250 [02:23<04:17,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▌      | 448/1250 [02:23<04:16,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▌      | 449/1250 [02:24<04:15,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▌      | 450/1250 [02:24<04:14,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▌      | 451/1250 [02:24<04:15,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▌      | 452/1250 [02:25<04:14,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▌      | 453/1250 [02:25<04:13,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▋      | 454/1250 [02:25<04:14,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▋      | 455/1250 [02:26<04:14,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 36%|███▋      | 456/1250 [02:26<04:14,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 457/1250 [02:26<04:13,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 458/1250 [02:27<04:13,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 459/1250 [02:27<04:13,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 460/1250 [02:27<04:14,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 461/1250 [02:27<04:14,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 462/1250 [02:28<04:13,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 463/1250 [02:28<04:14,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 464/1250 [02:28<04:13,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 465/1250 [02:29<04:13,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 466/1250 [02:29<04:12,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 467/1250 [02:29<04:11,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 37%|███▋      | 468/1250 [02:30<04:10,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 469/1250 [02:30<04:09,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 470/1250 [02:30<04:11,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 471/1250 [02:31<04:10,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 472/1250 [02:31<04:09,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 473/1250 [02:31<04:08,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 474/1250 [02:32<04:08,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 475/1250 [02:32<04:09,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 476/1250 [02:32<04:08,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 477/1250 [02:33<04:09,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 478/1250 [02:33<04:08,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 479/1250 [02:33<04:06,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 480/1250 [02:34<04:06,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 38%|███▊      | 481/1250 [02:34<04:05,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▊      | 482/1250 [02:34<04:05,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▊      | 483/1250 [02:35<04:04,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▊      | 484/1250 [02:35<04:04,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▉      | 485/1250 [02:35<04:03,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▉      | 486/1250 [02:36<04:03,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▉      | 487/1250 [02:36<04:02,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▉      | 488/1250 [02:36<04:02,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▉      | 489/1250 [02:36<04:03,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▉      | 490/1250 [02:37<04:04,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▉      | 491/1250 [02:37<04:02,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▉      | 492/1250 [02:37<04:01,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 39%|███▉      | 493/1250 [02:38<04:01,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|███▉      | 494/1250 [02:38<04:01,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|███▉      | 495/1250 [02:38<04:01,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|███▉      | 496/1250 [02:39<04:00,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|███▉      | 497/1250 [02:39<04:00,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|███▉      | 498/1250 [02:39<04:00,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|███▉      | 499/1250 [02:40<04:01,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|████      | 500/1250 [02:40<03:59,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|████      | 501/1250 [02:40<03:59,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|████      | 502/1250 [02:41<03:58,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|████      | 503/1250 [02:41<03:59,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|████      | 504/1250 [02:41<04:00,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|████      | 505/1250 [02:42<04:00,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 40%|████      | 506/1250 [02:42<03:59,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████      | 507/1250 [02:42<03:58,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████      | 508/1250 [02:43<03:57,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████      | 509/1250 [02:43<03:56,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████      | 510/1250 [02:43<03:57,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████      | 511/1250 [02:44<03:56,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████      | 512/1250 [02:44<03:56,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████      | 513/1250 [02:44<03:55,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████      | 514/1250 [02:44<03:55,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████      | 515/1250 [02:45<03:55,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████▏     | 516/1250 [02:45<03:54,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████▏     | 517/1250 [02:45<03:54,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 41%|████▏     | 518/1250 [02:46<03:54,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 519/1250 [02:46<03:53,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 520/1250 [02:46<03:54,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 521/1250 [02:47<03:53,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 522/1250 [02:47<03:52,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 523/1250 [02:47<03:52,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 524/1250 [02:48<03:51,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 525/1250 [02:48<03:51,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 526/1250 [02:48<03:50,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 527/1250 [02:49<03:50,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 528/1250 [02:49<03:50,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 529/1250 [02:49<03:49,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 530/1250 [02:50<03:49,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 42%|████▏     | 531/1250 [02:50<03:48,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 532/1250 [02:50<03:48,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 533/1250 [02:51<03:48,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 534/1250 [02:51<03:47,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 535/1250 [02:51<03:48,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 536/1250 [02:51<03:49,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 537/1250 [02:52<03:50,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 538/1250 [02:52<03:49,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 539/1250 [02:52<03:49,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 540/1250 [02:53<03:49,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 541/1250 [02:53<03:47,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 542/1250 [02:53<03:46,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 43%|████▎     | 543/1250 [02:54<03:46,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▎     | 544/1250 [02:54<03:46,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▎     | 545/1250 [02:54<03:46,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▎     | 546/1250 [02:55<03:46,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▍     | 547/1250 [02:55<03:45,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▍     | 548/1250 [02:55<03:44,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▍     | 549/1250 [02:56<03:44,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▍     | 550/1250 [02:56<03:44,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▍     | 551/1250 [02:56<03:45,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▍     | 552/1250 [02:57<03:45,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▍     | 553/1250 [02:57<03:44,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▍     | 554/1250 [02:57<03:43,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▍     | 555/1250 [02:58<03:42,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 44%|████▍     | 556/1250 [02:58<03:42,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▍     | 557/1250 [02:58<03:42,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▍     | 558/1250 [02:59<03:43,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▍     | 559/1250 [02:59<03:42,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▍     | 560/1250 [02:59<03:43,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▍     | 561/1250 [03:00<03:43,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▍     | 562/1250 [03:00<03:42,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▌     | 563/1250 [03:00<03:41,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▌     | 564/1250 [03:01<03:41,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▌     | 565/1250 [03:01<03:39,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▌     | 566/1250 [03:01<03:39,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▌     | 567/1250 [03:01<03:39,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 45%|████▌     | 568/1250 [03:02<03:39,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▌     | 569/1250 [03:02<03:39,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▌     | 570/1250 [03:02<03:39,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▌     | 571/1250 [03:03<03:38,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▌     | 572/1250 [03:03<03:38,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▌     | 573/1250 [03:03<03:37,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▌     | 574/1250 [03:04<03:36,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▌     | 575/1250 [03:04<03:37,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▌     | 576/1250 [03:04<03:37,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▌     | 577/1250 [03:05<03:37,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▌     | 578/1250 [03:05<03:37,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▋     | 579/1250 [03:05<03:37,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▋     | 580/1250 [03:06<03:37,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 46%|████▋     | 581/1250 [03:06<03:36,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 582/1250 [03:06<03:36,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 583/1250 [03:07<03:35,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 584/1250 [03:07<03:35,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 585/1250 [03:07<03:34,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 586/1250 [03:08<03:34,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 587/1250 [03:08<03:33,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 588/1250 [03:08<03:32,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 589/1250 [03:09<03:32,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 590/1250 [03:09<03:32,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 591/1250 [03:09<03:32,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 592/1250 [03:10<03:32,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 47%|████▋     | 593/1250 [03:10<03:33,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 594/1250 [03:10<03:31,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 595/1250 [03:11<03:32,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 596/1250 [03:11<03:32,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 597/1250 [03:11<03:31,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 598/1250 [03:11<03:31,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 599/1250 [03:12<03:31,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 600/1250 [03:12<03:30,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 601/1250 [03:12<03:30,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 602/1250 [03:13<03:30,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 603/1250 [03:13<03:29,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 604/1250 [03:13<03:28,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 605/1250 [03:14<03:29,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 48%|████▊     | 606/1250 [03:14<03:29,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▊     | 607/1250 [03:14<03:28,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▊     | 608/1250 [03:15<03:28,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▊     | 609/1250 [03:15<03:27,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▉     | 610/1250 [03:15<03:27,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▉     | 611/1250 [03:16<03:27,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▉     | 612/1250 [03:16<03:27,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▉     | 613/1250 [03:16<03:26,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▉     | 614/1250 [03:17<03:26,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▉     | 615/1250 [03:17<03:25,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▉     | 616/1250 [03:17<03:25,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▉     | 617/1250 [03:18<03:25,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 49%|████▉     | 618/1250 [03:18<03:24,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|████▉     | 619/1250 [03:18<03:24,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|████▉     | 620/1250 [03:19<03:24,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|████▉     | 621/1250 [03:19<03:24,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|████▉     | 622/1250 [03:19<03:23,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|████▉     | 623/1250 [03:20<03:23,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|████▉     | 624/1250 [03:20<03:22,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|█████     | 625/1250 [03:20<03:22,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|█████     | 626/1250 [03:21<03:22,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|█████     | 627/1250 [03:21<03:22,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|█████     | 628/1250 [03:21<03:22,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|█████     | 629/1250 [03:22<03:21,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|█████     | 630/1250 [03:22<03:20,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 50%|█████     | 631/1250 [03:22<03:20,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████     | 632/1250 [03:23<03:20,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████     | 633/1250 [03:23<03:20,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████     | 634/1250 [03:23<03:20,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████     | 635/1250 [03:23<03:19,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████     | 636/1250 [03:24<03:18,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████     | 637/1250 [03:24<03:18,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████     | 638/1250 [03:24<03:18,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████     | 639/1250 [03:25<03:18,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████     | 640/1250 [03:25<03:17,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████▏    | 641/1250 [03:25<03:17,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████▏    | 642/1250 [03:26<03:17,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 51%|█████▏    | 643/1250 [03:26<03:16,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 644/1250 [03:26<03:16,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 645/1250 [03:27<03:15,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 646/1250 [03:27<03:15,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 647/1250 [03:27<03:15,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 648/1250 [03:28<03:14,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 649/1250 [03:28<03:14,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 650/1250 [03:28<03:13,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 651/1250 [03:29<03:12,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 652/1250 [03:29<03:11,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 653/1250 [03:29<03:11,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 654/1250 [03:30<03:11,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 655/1250 [03:30<03:11,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 52%|█████▏    | 656/1250 [03:30<03:10,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 657/1250 [03:31<03:09,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 658/1250 [03:31<03:09,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 659/1250 [03:31<03:08,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 660/1250 [03:32<03:08,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 661/1250 [03:32<03:08,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 662/1250 [03:32<03:07,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 663/1250 [03:32<03:07,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 664/1250 [03:33<03:07,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 665/1250 [03:33<03:07,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 666/1250 [03:33<03:08,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 667/1250 [03:34<03:07,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 53%|█████▎    | 668/1250 [03:34<03:07,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▎    | 669/1250 [03:34<03:06,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▎    | 670/1250 [03:35<03:05,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▎    | 671/1250 [03:35<03:05,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▍    | 672/1250 [03:35<03:05,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▍    | 673/1250 [03:36<03:06,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▍    | 674/1250 [03:36<03:05,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▍    | 675/1250 [03:36<03:04,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▍    | 676/1250 [03:37<03:03,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▍    | 677/1250 [03:37<03:03,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▍    | 678/1250 [03:37<03:02,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▍    | 679/1250 [03:38<03:02,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▍    | 680/1250 [03:38<03:01,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 54%|█████▍    | 681/1250 [03:38<03:02,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▍    | 682/1250 [03:39<03:02,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▍    | 683/1250 [03:39<03:02,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▍    | 684/1250 [03:39<03:02,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▍    | 685/1250 [03:40<03:02,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▍    | 686/1250 [03:40<03:02,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▍    | 687/1250 [03:40<03:01,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▌    | 688/1250 [03:41<03:00,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▌    | 689/1250 [03:41<03:00,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▌    | 690/1250 [03:41<02:59,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▌    | 691/1250 [03:41<02:59,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▌    | 692/1250 [03:42<02:59,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 55%|█████▌    | 693/1250 [03:42<02:59,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▌    | 694/1250 [03:42<02:59,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▌    | 695/1250 [03:43<02:58,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▌    | 696/1250 [03:43<02:57,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▌    | 697/1250 [03:43<02:56,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▌    | 698/1250 [03:44<02:57,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▌    | 699/1250 [03:44<02:57,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▌    | 700/1250 [03:44<02:57,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▌    | 701/1250 [03:45<02:57,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▌    | 702/1250 [03:45<02:57,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▌    | 703/1250 [03:45<02:56,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▋    | 704/1250 [03:46<02:56,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▋    | 705/1250 [03:46<02:57,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 56%|█████▋    | 706/1250 [03:46<02:56,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 707/1250 [03:47<02:56,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 708/1250 [03:47<02:55,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 709/1250 [03:47<02:55,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 710/1250 [03:48<02:54,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 711/1250 [03:48<02:54,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 712/1250 [03:48<02:54,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 713/1250 [03:49<02:53,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 714/1250 [03:49<02:52,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 715/1250 [03:49<02:51,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 716/1250 [03:50<02:51,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 717/1250 [03:50<02:51,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 57%|█████▋    | 718/1250 [03:50<02:51,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 719/1250 [03:51<02:50,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 720/1250 [03:51<02:50,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 721/1250 [03:51<02:49,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 722/1250 [03:51<02:49,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 723/1250 [03:52<02:48,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 724/1250 [03:52<02:48,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 725/1250 [03:52<02:47,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 726/1250 [03:53<02:46,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 727/1250 [03:53<02:46,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 728/1250 [03:53<02:46,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 729/1250 [03:54<02:45,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 730/1250 [03:54<02:45,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 58%|█████▊    | 731/1250 [03:54<02:46,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▊    | 732/1250 [03:55<02:46,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▊    | 733/1250 [03:55<02:45,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▊    | 734/1250 [03:55<02:45,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▉    | 735/1250 [03:56<02:44,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▉    | 736/1250 [03:56<02:44,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▉    | 737/1250 [03:56<02:44,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▉    | 738/1250 [03:57<02:45,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▉    | 739/1250 [03:57<02:44,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▉    | 740/1250 [03:57<02:44,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▉    | 741/1250 [03:58<02:43,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▉    | 742/1250 [03:58<02:42,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 59%|█████▉    | 743/1250 [03:58<02:42,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|█████▉    | 744/1250 [03:59<02:42,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|█████▉    | 745/1250 [03:59<02:42,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|█████▉    | 746/1250 [03:59<02:42,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|█████▉    | 747/1250 [04:00<02:42,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|█████▉    | 748/1250 [04:00<02:42,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|█████▉    | 749/1250 [04:00<02:42,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|██████    | 750/1250 [04:00<02:41,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|██████    | 751/1250 [04:01<02:40,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|██████    | 752/1250 [04:01<02:39,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|██████    | 753/1250 [04:01<02:39,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|██████    | 754/1250 [04:02<02:38,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|██████    | 755/1250 [04:02<02:38,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 60%|██████    | 756/1250 [04:02<02:37,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████    | 757/1250 [04:03<02:37,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████    | 758/1250 [04:03<02:37,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████    | 759/1250 [04:03<02:37,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████    | 760/1250 [04:04<02:37,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████    | 761/1250 [04:04<02:36,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████    | 762/1250 [04:04<02:35,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████    | 763/1250 [04:05<02:35,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████    | 764/1250 [04:05<02:36,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████    | 765/1250 [04:05<02:36,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████▏   | 766/1250 [04:06<02:36,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████▏   | 767/1250 [04:06<02:35,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 61%|██████▏   | 768/1250 [04:06<02:35,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 769/1250 [04:07<02:34,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 770/1250 [04:07<02:33,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 771/1250 [04:07<02:33,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 772/1250 [04:08<02:33,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 773/1250 [04:08<02:33,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 774/1250 [04:08<02:33,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 775/1250 [04:09<02:33,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 776/1250 [04:09<02:33,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 777/1250 [04:09<02:33,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 778/1250 [04:09<02:32,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 779/1250 [04:10<02:32,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 780/1250 [04:10<02:32,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 62%|██████▏   | 781/1250 [04:10<02:31,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 782/1250 [04:11<02:32,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 783/1250 [04:11<02:31,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 784/1250 [04:11<02:31,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 785/1250 [04:12<02:30,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 786/1250 [04:12<02:30,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 787/1250 [04:12<02:29,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 788/1250 [04:13<02:28,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 789/1250 [04:13<02:27,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 790/1250 [04:13<02:28,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 791/1250 [04:14<02:27,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 792/1250 [04:14<02:27,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 63%|██████▎   | 793/1250 [04:14<02:27,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▎   | 794/1250 [04:15<02:27,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▎   | 795/1250 [04:15<02:27,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▎   | 796/1250 [04:15<02:26,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▍   | 797/1250 [04:16<02:26,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▍   | 798/1250 [04:16<02:26,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▍   | 799/1250 [04:16<02:26,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▍   | 800/1250 [04:17<02:25,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▍   | 801/1250 [04:17<02:25,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▍   | 802/1250 [04:17<02:25,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▍   | 803/1250 [04:18<02:24,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▍   | 804/1250 [04:18<02:24,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▍   | 805/1250 [04:18<02:23,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 64%|██████▍   | 806/1250 [04:19<02:22,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▍   | 807/1250 [04:19<02:22,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▍   | 808/1250 [04:19<02:22,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▍   | 809/1250 [04:19<02:22,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▍   | 810/1250 [04:20<02:21,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▍   | 811/1250 [04:20<02:20,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▍   | 812/1250 [04:20<02:20,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▌   | 813/1250 [04:21<02:19,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▌   | 814/1250 [04:21<02:19,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▌   | 815/1250 [04:21<02:19,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▌   | 816/1250 [04:22<02:18,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▌   | 817/1250 [04:22<02:18,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 65%|██████▌   | 818/1250 [04:22<02:17,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▌   | 819/1250 [04:23<02:17,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▌   | 820/1250 [04:23<02:16,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▌   | 821/1250 [04:23<02:16,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▌   | 822/1250 [04:24<02:16,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▌   | 823/1250 [04:24<02:15,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▌   | 824/1250 [04:24<02:15,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▌   | 825/1250 [04:25<02:16,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▌   | 826/1250 [04:25<02:16,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▌   | 827/1250 [04:25<02:15,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▌   | 828/1250 [04:26<02:15,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▋   | 829/1250 [04:26<02:14,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▋   | 830/1250 [04:26<02:14,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 66%|██████▋   | 831/1250 [04:27<02:14,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 832/1250 [04:27<02:14,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 833/1250 [04:27<02:13,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 834/1250 [04:27<02:13,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 835/1250 [04:28<02:12,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 836/1250 [04:28<02:12,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 837/1250 [04:28<02:12,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 838/1250 [04:29<02:12,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 839/1250 [04:29<02:12,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 840/1250 [04:29<02:12,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 841/1250 [04:30<02:12,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 842/1250 [04:30<02:12,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 67%|██████▋   | 843/1250 [04:30<02:11,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 844/1250 [04:31<02:11,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 845/1250 [04:31<02:11,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 846/1250 [04:31<02:11,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 847/1250 [04:32<02:11,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 848/1250 [04:32<02:10,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 849/1250 [04:32<02:10,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 850/1250 [04:33<02:09,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 851/1250 [04:33<02:09,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 852/1250 [04:33<02:08,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 853/1250 [04:34<02:07,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 854/1250 [04:34<02:06,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 855/1250 [04:34<02:06,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 68%|██████▊   | 856/1250 [04:35<02:06,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▊   | 857/1250 [04:35<02:06,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▊   | 858/1250 [04:35<02:05,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▊   | 859/1250 [04:36<02:05,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▉   | 860/1250 [04:36<02:04,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▉   | 861/1250 [04:36<02:04,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▉   | 862/1250 [04:37<02:04,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▉   | 863/1250 [04:37<02:03,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▉   | 864/1250 [04:37<02:03,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▉   | 865/1250 [04:37<02:02,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▉   | 866/1250 [04:38<02:03,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▉   | 867/1250 [04:38<02:03,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 69%|██████▉   | 868/1250 [04:38<02:02,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|██████▉   | 869/1250 [04:39<02:01,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|██████▉   | 870/1250 [04:39<02:01,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|██████▉   | 871/1250 [04:39<02:01,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|██████▉   | 872/1250 [04:40<02:01,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|██████▉   | 873/1250 [04:40<02:01,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|██████▉   | 874/1250 [04:40<02:00,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|███████   | 875/1250 [04:41<02:00,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|███████   | 876/1250 [04:41<01:59,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|███████   | 877/1250 [04:41<01:59,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|███████   | 878/1250 [04:42<01:59,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|███████   | 879/1250 [04:42<01:59,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|███████   | 880/1250 [04:42<01:58,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 70%|███████   | 881/1250 [04:43<01:58,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████   | 882/1250 [04:43<01:57,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████   | 883/1250 [04:43<01:57,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████   | 884/1250 [04:44<01:57,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████   | 885/1250 [04:44<01:57,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████   | 886/1250 [04:44<01:56,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████   | 887/1250 [04:45<01:56,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████   | 888/1250 [04:45<01:55,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████   | 889/1250 [04:45<01:55,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████   | 890/1250 [04:45<01:55,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████▏  | 891/1250 [04:46<01:56,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████▏  | 892/1250 [04:46<01:55,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 71%|███████▏  | 893/1250 [04:46<01:55,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 894/1250 [04:47<01:54,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 895/1250 [04:47<01:54,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 896/1250 [04:47<01:53,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 897/1250 [04:48<01:53,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 898/1250 [04:48<01:52,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 899/1250 [04:48<01:52,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 900/1250 [04:49<01:52,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 901/1250 [04:49<01:52,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 902/1250 [04:49<01:52,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 903/1250 [04:50<01:52,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 904/1250 [04:50<01:51,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 905/1250 [04:50<01:51,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 72%|███████▏  | 906/1250 [04:51<01:51,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 907/1250 [04:51<01:51,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 908/1250 [04:51<01:50,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 909/1250 [04:52<01:50,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 910/1250 [04:52<01:50,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 911/1250 [04:52<01:49,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 912/1250 [04:53<01:49,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 913/1250 [04:53<01:49,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 914/1250 [04:53<01:49,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 915/1250 [04:54<01:48,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 916/1250 [04:54<01:48,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 917/1250 [04:54<01:47,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 73%|███████▎  | 918/1250 [04:55<01:47,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▎  | 919/1250 [04:55<01:47,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▎  | 920/1250 [04:55<01:46,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▎  | 921/1250 [04:56<01:46,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▍  | 922/1250 [04:56<01:46,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▍  | 923/1250 [04:56<01:46,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▍  | 924/1250 [04:56<01:45,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▍  | 925/1250 [04:57<01:45,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▍  | 926/1250 [04:57<01:44,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▍  | 927/1250 [04:57<01:45,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▍  | 928/1250 [04:58<01:44,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▍  | 929/1250 [04:58<01:44,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▍  | 930/1250 [04:58<01:43,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 74%|███████▍  | 931/1250 [04:59<01:42,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▍  | 932/1250 [04:59<01:42,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▍  | 933/1250 [04:59<01:41,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▍  | 934/1250 [05:00<01:41,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▍  | 935/1250 [05:00<01:40,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▍  | 936/1250 [05:00<01:41,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▍  | 937/1250 [05:01<01:40,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▌  | 938/1250 [05:01<01:40,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▌  | 939/1250 [05:01<01:40,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▌  | 940/1250 [05:02<01:39,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▌  | 941/1250 [05:02<01:38,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▌  | 942/1250 [05:02<01:38,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 75%|███████▌  | 943/1250 [05:03<01:38,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▌  | 944/1250 [05:03<01:38,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▌  | 945/1250 [05:03<01:38,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▌  | 946/1250 [05:04<01:38,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▌  | 947/1250 [05:04<01:38,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▌  | 948/1250 [05:04<01:37,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▌  | 949/1250 [05:05<01:37,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▌  | 950/1250 [05:05<01:37,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▌  | 951/1250 [05:05<01:36,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▌  | 952/1250 [05:06<01:36,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▌  | 953/1250 [05:06<01:36,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▋  | 954/1250 [05:06<01:35,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▋  | 955/1250 [05:06<01:35,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 76%|███████▋  | 956/1250 [05:07<01:35,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 957/1250 [05:07<01:34,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 958/1250 [05:07<01:34,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 959/1250 [05:08<01:33,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 960/1250 [05:08<01:33,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 961/1250 [05:08<01:32,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 962/1250 [05:09<01:32,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 963/1250 [05:09<01:32,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 964/1250 [05:09<01:32,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 965/1250 [05:10<01:31,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 966/1250 [05:10<01:31,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 967/1250 [05:10<01:31,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 77%|███████▋  | 968/1250 [05:11<01:31,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 969/1250 [05:11<01:31,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 970/1250 [05:11<01:30,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 971/1250 [05:12<01:30,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 972/1250 [05:12<01:29,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 973/1250 [05:12<01:29,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 974/1250 [05:13<01:28,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 975/1250 [05:13<01:28,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 976/1250 [05:13<01:27,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 977/1250 [05:14<01:27,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 978/1250 [05:14<01:27,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 979/1250 [05:14<01:27,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 980/1250 [05:15<01:26,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 78%|███████▊  | 981/1250 [05:15<01:26,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▊  | 982/1250 [05:15<01:25,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▊  | 983/1250 [05:16<01:25,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▊  | 984/1250 [05:16<01:25,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▉  | 985/1250 [05:16<01:25,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▉  | 986/1250 [05:16<01:25,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▉  | 987/1250 [05:17<01:24,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▉  | 988/1250 [05:17<01:24,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▉  | 989/1250 [05:17<01:24,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▉  | 990/1250 [05:18<01:24,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▉  | 991/1250 [05:18<01:24,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▉  | 992/1250 [05:18<01:23,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 79%|███████▉  | 993/1250 [05:19<01:23,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|███████▉  | 994/1250 [05:19<01:23,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|███████▉  | 995/1250 [05:19<01:22,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|███████▉  | 996/1250 [05:20<01:22,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|███████▉  | 997/1250 [05:20<01:21,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|███████▉  | 998/1250 [05:20<01:21,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|███████▉  | 999/1250 [05:21<01:21,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|████████  | 1000/1250 [05:21<01:21,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|████████  | 1001/1250 [05:21<01:20,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|████████  | 1002/1250 [05:22<01:20,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|████████  | 1003/1250 [05:22<01:20,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|████████  | 1004/1250 [05:22<01:19,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|████████  | 1005/1250 [05:23<01:19,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 80%|████████  | 1006/1250 [05:23<01:19,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████  | 1007/1250 [05:23<01:18,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████  | 1008/1250 [05:24<01:18,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████  | 1009/1250 [05:24<01:18,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████  | 1010/1250 [05:24<01:17,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████  | 1011/1250 [05:25<01:17,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████  | 1012/1250 [05:25<01:17,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████  | 1013/1250 [05:25<01:16,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████  | 1014/1250 [05:26<01:16,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████  | 1015/1250 [05:26<01:16,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████▏ | 1016/1250 [05:26<01:15,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████▏ | 1017/1250 [05:27<01:15,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 81%|████████▏ | 1018/1250 [05:27<01:15,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1019/1250 [05:27<01:15,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1020/1250 [05:28<01:14,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1021/1250 [05:28<01:14,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1022/1250 [05:28<01:13,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1023/1250 [05:28<01:13,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1024/1250 [05:29<01:12,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1025/1250 [05:29<01:12,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1026/1250 [05:29<01:11,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1027/1250 [05:30<01:11,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1028/1250 [05:30<01:11,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1029/1250 [05:30<01:11,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1030/1250 [05:31<01:10,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 82%|████████▏ | 1031/1250 [05:31<01:10,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1032/1250 [05:31<01:09,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1033/1250 [05:32<01:09,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1034/1250 [05:32<01:09,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1035/1250 [05:32<01:09,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1036/1250 [05:33<01:08,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1037/1250 [05:33<01:08,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1038/1250 [05:33<01:07,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1039/1250 [05:34<01:07,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1040/1250 [05:34<01:07,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1041/1250 [05:34<01:06,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1042/1250 [05:35<01:06,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 83%|████████▎ | 1043/1250 [05:35<01:06,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▎ | 1044/1250 [05:35<01:05,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▎ | 1045/1250 [05:36<01:05,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▎ | 1046/1250 [05:36<01:05,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▍ | 1047/1250 [05:36<01:04,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▍ | 1048/1250 [05:36<01:04,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▍ | 1049/1250 [05:37<01:04,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▍ | 1050/1250 [05:37<01:04,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▍ | 1051/1250 [05:37<01:03,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▍ | 1052/1250 [05:38<01:03,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▍ | 1053/1250 [05:38<01:02,  3.14it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▍ | 1054/1250 [05:38<01:02,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▍ | 1055/1250 [05:39<01:02,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 84%|████████▍ | 1056/1250 [05:39<01:02,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▍ | 1057/1250 [05:39<01:02,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▍ | 1058/1250 [05:40<01:02,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▍ | 1059/1250 [05:40<01:01,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▍ | 1060/1250 [05:40<01:01,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▍ | 1061/1250 [05:41<01:00,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▍ | 1062/1250 [05:41<01:00,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▌ | 1063/1250 [05:41<00:59,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▌ | 1064/1250 [05:42<00:59,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▌ | 1065/1250 [05:42<00:59,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▌ | 1066/1250 [05:42<00:58,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▌ | 1067/1250 [05:43<00:58,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 85%|████████▌ | 1068/1250 [05:43<00:58,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▌ | 1069/1250 [05:43<00:58,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▌ | 1070/1250 [05:44<00:58,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▌ | 1071/1250 [05:44<00:57,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▌ | 1072/1250 [05:44<00:57,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▌ | 1073/1250 [05:45<00:57,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▌ | 1074/1250 [05:45<00:57,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▌ | 1075/1250 [05:45<00:56,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▌ | 1076/1250 [05:46<00:56,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▌ | 1077/1250 [05:46<00:55,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▌ | 1078/1250 [05:46<00:55,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▋ | 1079/1250 [05:46<00:55,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▋ | 1080/1250 [05:47<00:55,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 86%|████████▋ | 1081/1250 [05:47<00:54,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1082/1250 [05:47<00:54,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1083/1250 [05:48<00:54,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1084/1250 [05:48<00:53,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1085/1250 [05:48<00:53,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1086/1250 [05:49<00:53,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1087/1250 [05:49<00:52,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1088/1250 [05:49<00:52,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1089/1250 [05:50<00:52,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1090/1250 [05:50<00:51,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1091/1250 [05:50<00:51,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1092/1250 [05:51<00:51,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 87%|████████▋ | 1093/1250 [05:51<00:50,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1094/1250 [05:51<00:50,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1095/1250 [05:52<00:50,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1096/1250 [05:52<00:49,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1097/1250 [05:52<00:49,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1098/1250 [05:53<00:48,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1099/1250 [05:53<00:48,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1100/1250 [05:53<00:48,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1101/1250 [05:54<00:47,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1102/1250 [05:54<00:47,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1103/1250 [05:54<00:47,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1104/1250 [05:55<00:46,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1105/1250 [05:55<00:46,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 88%|████████▊ | 1106/1250 [05:55<00:46,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▊ | 1107/1250 [05:56<00:46,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▊ | 1108/1250 [05:56<00:45,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▊ | 1109/1250 [05:56<00:45,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▉ | 1110/1250 [05:56<00:45,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▉ | 1111/1250 [05:57<00:44,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▉ | 1112/1250 [05:57<00:44,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▉ | 1113/1250 [05:57<00:44,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▉ | 1114/1250 [05:58<00:44,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▉ | 1115/1250 [05:58<00:43,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▉ | 1116/1250 [05:58<00:43,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▉ | 1117/1250 [05:59<00:43,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 89%|████████▉ | 1118/1250 [05:59<00:42,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|████████▉ | 1119/1250 [05:59<00:42,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|████████▉ | 1120/1250 [06:00<00:42,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|████████▉ | 1121/1250 [06:00<00:41,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|████████▉ | 1122/1250 [06:00<00:41,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|████████▉ | 1123/1250 [06:01<00:41,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|████████▉ | 1124/1250 [06:01<00:40,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|█████████ | 1125/1250 [06:01<00:40,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|█████████ | 1126/1250 [06:02<00:40,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|█████████ | 1127/1250 [06:02<00:39,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|█████████ | 1128/1250 [06:02<00:39,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|█████████ | 1129/1250 [06:03<00:39,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|█████████ | 1130/1250 [06:03<00:39,  3.07it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 90%|█████████ | 1131/1250 [06:03<00:38,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████ | 1132/1250 [06:04<00:38,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████ | 1133/1250 [06:04<00:37,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████ | 1134/1250 [06:04<00:37,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████ | 1135/1250 [06:05<00:36,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████ | 1136/1250 [06:05<00:36,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████ | 1137/1250 [06:05<00:36,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████ | 1138/1250 [06:06<00:36,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████ | 1139/1250 [06:06<00:35,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████ | 1140/1250 [06:06<00:35,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████▏| 1141/1250 [06:07<00:34,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████▏| 1142/1250 [06:07<00:34,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 91%|█████████▏| 1143/1250 [06:07<00:34,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1144/1250 [06:07<00:34,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1145/1250 [06:08<00:33,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1146/1250 [06:08<00:33,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1147/1250 [06:08<00:33,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1148/1250 [06:09<00:32,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1149/1250 [06:09<00:32,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1150/1250 [06:09<00:32,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1151/1250 [06:10<00:31,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1152/1250 [06:10<00:31,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1153/1250 [06:10<00:31,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1154/1250 [06:11<00:30,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1155/1250 [06:11<00:30,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 92%|█████████▏| 1156/1250 [06:11<00:30,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1157/1250 [06:12<00:29,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1158/1250 [06:12<00:29,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1159/1250 [06:12<00:29,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1160/1250 [06:13<00:28,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1161/1250 [06:13<00:28,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1162/1250 [06:13<00:28,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1163/1250 [06:14<00:27,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1164/1250 [06:14<00:27,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1165/1250 [06:14<00:27,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1166/1250 [06:15<00:26,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1167/1250 [06:15<00:26,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 93%|█████████▎| 1168/1250 [06:15<00:26,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▎| 1169/1250 [06:15<00:25,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▎| 1170/1250 [06:16<00:25,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▎| 1171/1250 [06:16<00:25,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▍| 1172/1250 [06:16<00:24,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▍| 1173/1250 [06:17<00:24,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▍| 1174/1250 [06:17<00:24,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▍| 1175/1250 [06:17<00:24,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▍| 1176/1250 [06:18<00:23,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▍| 1177/1250 [06:18<00:23,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▍| 1178/1250 [06:18<00:23,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▍| 1179/1250 [06:19<00:22,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▍| 1180/1250 [06:19<00:22,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 94%|█████████▍| 1181/1250 [06:19<00:22,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▍| 1182/1250 [06:20<00:21,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▍| 1183/1250 [06:20<00:21,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▍| 1184/1250 [06:20<00:21,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▍| 1185/1250 [06:21<00:21,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▍| 1186/1250 [06:21<00:20,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▍| 1187/1250 [06:21<00:20,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▌| 1188/1250 [06:22<00:20,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▌| 1189/1250 [06:22<00:19,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▌| 1190/1250 [06:22<00:19,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▌| 1191/1250 [06:23<00:19,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▌| 1192/1250 [06:23<00:18,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 95%|█████████▌| 1193/1250 [06:23<00:18,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▌| 1194/1250 [06:24<00:17,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▌| 1195/1250 [06:24<00:17,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▌| 1196/1250 [06:24<00:17,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▌| 1197/1250 [06:25<00:17,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▌| 1198/1250 [06:25<00:16,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▌| 1199/1250 [06:25<00:16,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▌| 1200/1250 [06:25<00:16,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▌| 1201/1250 [06:26<00:15,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▌| 1202/1250 [06:26<00:15,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▌| 1203/1250 [06:26<00:15,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▋| 1204/1250 [06:27<00:14,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▋| 1205/1250 [06:27<00:14,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 96%|█████████▋| 1206/1250 [06:27<00:14,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1207/1250 [06:28<00:13,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1208/1250 [06:28<00:13,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1209/1250 [06:28<00:13,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1210/1250 [06:29<00:12,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1211/1250 [06:29<00:12,  3.13it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1212/1250 [06:29<00:12,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1213/1250 [06:30<00:11,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1214/1250 [06:30<00:11,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1215/1250 [06:30<00:11,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1216/1250 [06:31<00:10,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1217/1250 [06:31<00:10,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 97%|█████████▋| 1218/1250 [06:31<00:10,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1219/1250 [06:32<00:10,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1220/1250 [06:32<00:09,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1221/1250 [06:32<00:09,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1222/1250 [06:33<00:09,  3.08it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1223/1250 [06:33<00:08,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1224/1250 [06:33<00:08,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1225/1250 [06:34<00:08,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1226/1250 [06:34<00:07,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1227/1250 [06:34<00:07,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1228/1250 [06:35<00:07,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1229/1250 [06:35<00:06,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1230/1250 [06:35<00:06,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 98%|█████████▊| 1231/1250 [06:35<00:06,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▊| 1232/1250 [06:36<00:05,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▊| 1233/1250 [06:36<00:05,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▊| 1234/1250 [06:36<00:05,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▉| 1235/1250 [06:37<00:04,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▉| 1236/1250 [06:37<00:04,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▉| 1237/1250 [06:37<00:04,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▉| 1238/1250 [06:38<00:03,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▉| 1239/1250 [06:38<00:03,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▉| 1240/1250 [06:38<00:03,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▉| 1241/1250 [06:39<00:02,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▉| 1242/1250 [06:39<00:02,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            " 99%|█████████▉| 1243/1250 [06:39<00:02,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "100%|█████████▉| 1244/1250 [06:40<00:01,  3.09it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "100%|█████████▉| 1245/1250 [06:40<00:01,  3.10it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "100%|█████████▉| 1246/1250 [06:40<00:01,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "100%|█████████▉| 1247/1250 [06:41<00:00,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "100%|█████████▉| 1248/1250 [06:41<00:00,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "100%|█████████▉| 1249/1250 [06:41<00:00,  3.12it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "100%|██████████| 1250/1250 [06:42<00:00,  3.11it/s]\u001b[A\u001b[A\u001b[A\u001b[A\u001b[A\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "                                                   \u001b[A\u001b[A\u001b[A\u001b[A\u001b[A"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "loss:0.002377\n"
          ]
        }
      ],
      "source": [
        "net.eval()\n",
        "t = tqdm(validLoader, leave=False, total=len(validLoader))\n",
        "test_loss = 0\n",
        "for i, (idx, targetVar, inputVar, _, _) in enumerate(t):\n",
        "    if i == 3000:\n",
        "        break\n",
        "    inputs = inputVar.to(device)\n",
        "    label = targetVar.to(device)\n",
        "    pred = net(inputs)\n",
        "    loss = lossfunction(pred, label)\n",
        "    loss_aver = loss.item() / batch\n",
        "    test_loss += loss_aver\n",
        "print('loss:{:.6f}'.format(test_loss/len(t)))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cK36mAplESG2"
      },
      "source": [
        "# 可視化"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 453
        },
        "id": "aXIdYSL3kU95",
        "outputId": "e0c84b40-8446-4b26-cc26-facf282434af"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "\r  0%|          | 0/1250 [00:00<?, ?it/s]"
          ]
        },
        {
          "ename": "SystemExit",
          "evalue": "ignored",
          "output_type": "error",
          "traceback": [
            "An exception has occurred, use %tb to see the full traceback.\n",
            "\u001b[0;31mSystemExit\u001b[0m\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/IPython/core/interactiveshell.py:2890: UserWarning: To exit: use 'exit', 'quit', or Ctrl-D.\n",
            "  warn(\"To exit: use 'exit', 'quit', or Ctrl-D.\", stacklevel=1)\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1440x432 with 20 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light",
            "tags": []
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "net.eval()\n",
        "t = tqdm(validLoader, leave=False, total=len(validLoader))\n",
        "for i, (idx, targetVar, inputVar, _, _) in enumerate(t):\n",
        "    if i == 3000:\n",
        "        break\n",
        "    inputs = inputVar.to(device)\n",
        "    label = targetVar.to(device)\n",
        "    pred = net(inputs)\n",
        "    plt.figure(figsize=[20, 6])\n",
        "    for i in range(10):\n",
        "        plt.subplot(2, 10, i+1)\n",
        "        plt.imshow(label[0, i, 0].detach().cpu().numpy())\n",
        "        plt.axis('off')\n",
        "        plt.title('gt # %2d' %i)\n",
        "        plt.subplot(2, 10, i+11)\n",
        "        plt.imshow(pred[0, i, 0].detach().cpu().numpy())\n",
        "        plt.axis('off')\n",
        "        plt.title('pred # %2d' %i)\n",
        "    sys.exit()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yfgrm5jrF1BJ"
      },
      "source": [
        "可視化すると以下のようになります．\n",
        "\n",
        "\n",
        "<img src=\"https://github.com/himidev/Lecture/blob/main/13_rnn/07_ConvLSTM/pred1.png?raw=true\" width = 100%>\n",
        "\n",
        "<img src=\"https://github.com/himidev/Lecture/blob/main/13_rnn/07_ConvLSTM/pred2.png?raw=true\" width = 100%>\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JS3IALMa_k4-"
      },
      "source": [
        "# 課題\n",
        "各自のサーバでコードを動かしてみましょう．"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "tOK7tr6a_rsl"
      },
      "outputs": [],
      "source": []
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "authorship_tag": "ABX9TyPVji6ZqRiXYVXJgzFiHxWN",
      "collapsed_sections": [],
      "include_colab_link": true,
      "name": "07_ConvLSTM",
      "provenance": [],
      "toc_visible": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}