LSTM_Images_IMU_forloop-Np.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from __future__ import print_function, division\n",
    "\n",
    "import shutil\n",
    "import torch\n",
    "import torchvision\n",
    "import torchvision.transforms as transforms\n",
    "from torchvision.transforms import transforms\n",
    "from torch.utils.data import DataLoader\n",
    "from torch.utils.data.dataset import Dataset\n",
    "from torchvision import transforms\n",
    "from torch.utils.data.sampler import SubsetRandomSampler\n",
    "import torch.nn.functional as F\n",
    "import torch.nn as nn\n",
    "\n",
    "import os\n",
    "import sys\n",
    "import torch\n",
    "from torch.nn import functional as F\n",
    "import numpy as np\n",
    "import glob\n",
    "import random\n",
    "from PIL import Image\n",
    "from tqdm import tqdm\n",
    "import argparse\n",
    "from bisect import bisect_right, bisect_left\n",
    "\n",
    "import os\n",
    "import torch\n",
    "import pandas as pd\n",
    "from skimage import io, transform\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "from torchvision import transforms, utils\n",
    "\n",
    "# Ignore warnings\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "\n",
    "from PIL import Image\n",
    "\n",
    "from tqdm import tqdm\n",
    "import os\n",
    "import time\n",
    "from datetime import datetime\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "plt.ion()   # interactive mode\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Some functions\n",
    "\n",
    "\n",
    "def return_idx(timestamps, data):\n",
    "    \"Return the idx of the timestamp of the camera\"\n",
    "    \n",
    "    idx_sensor = []\n",
    "    for temp in timestamps:\n",
    "        idx=int(data[data['timestamp']==temp].index[0])\n",
    "        idx_sensor.append(idx)\n",
    "    return idx_sensor\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def return_idx_single(temp, data):\n",
    "    \"Return the idx of the timestamp of the camera\"\n",
    "    \n",
    "    idx_sensor = []\n",
    "    idx = data[data['timestamp']==temp].index[0]\n",
    "\n",
    "    return idx"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Import csv data from folders\n",
    "\n",
    "imu_data = pd.read_csv('/home/francesca/Desktop/francesca/data/mav0/imu0/data.csv')\n",
    "camera_data = pd.read_csv('/home/francesca/Desktop/francesca/data/mav0/cam0/data.csv')\n",
    "vicon_data = pd.read_csv('/home/francesca/Desktop/francesca/data/mav0/vicon0/data.csv')\n",
    "\n",
    "# Drop the first row of the camera\n",
    "\n",
    "camera_data=camera_data.drop(camera_data.index[[0]])\n",
    "\n",
    "# re-assign name to columns\n",
    "\n",
    "vicon_data.columns = ['timestamp','px','py','pz','qw','qx','qy','qz']\n",
    "camera_data.columns = ['timestamp','filename']\n",
    "imu_data.columns =['timestamp','wx','wy','wz','ax','ay','az']\n",
    "\n",
    "# Get the timestamp of the camera and drop the index\n",
    "camera_timestamps = camera_data['timestamp']\n",
    "camera_timestamps = camera_timestamps.reset_index(drop=True)\n",
    "\n",
    "### Interpolate Vicon data to get data at the same timestamp of the camera and IMU\n",
    "# Vicon data\n",
    "\n",
    "vicondata = vicon_data.copy() #create a copy\n",
    "\n",
    "for i in range(len(camera_timestamps)):\n",
    "    new_row=pd.DataFrame([[camera_timestamps[i],np.nan,np.nan,np.nan,np.nan,np.nan,np.nan,np.nan]], columns = ['timestamp','px','py','pz','qw','qx','qy','qz'])\n",
    "    vicondata = pd.concat([vicondata, pd.DataFrame(new_row)], ignore_index=True)\n",
    "    \n",
    "vicondata.sort_values(by=['timestamp'],inplace=True, ascending=True)\n",
    "vicondata = vicondata.reset_index(drop = True)\n",
    "vicondata = vicondata.interpolate()\n",
    "\n",
    "# Create a new vicon dataframe with the same timestamps of the camera\n",
    "\n",
    "idx_vicon=return_idx(camera_timestamps,vicondata)\n",
    "new_vicon = vicondata.ix[idx_vicon].reset_index(drop=True)\n",
    "\n",
    "vicon_timestamp = new_vicon['timestamp'].values\n",
    "\n",
    "# Let’s read the CSV and get the filenames and labels\n",
    "\n",
    "file = pd.read_csv('/home/francesca/Desktop/francesca/data/mav0/cam0/image_data.csv')\n",
    "file.drop(file.tail(1).index,inplace=True) # drop last n rows\n",
    "file.to_csv('/home/francesca/Desktop/francesca/data/mav0/cam0/file.csv',index = None, header=True)\n",
    "\n",
    "\n",
    "img_name = file.iloc[:, 1]\n",
    "labels = file.iloc[:, 2:].as_matrix()\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "c_t = np.array(camera_timestamps)\n",
    "new_imu_values = np.zeros((len(c_t),10,7))\n",
    "\n",
    "for i in range(len(c_t)):\n",
    "    idx_imu = return_idx_single(camera_timestamps.iloc[i], imu_data)\n",
    "    new_imu_values[i][:][:] = imu_data.ix[idx_imu:idx_imu+9].reset_index(drop=True).values\n",
    "\n",
    "\n",
    "new_imu_values = new_imu_values[:1920]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create new dataframe for images and save it\n",
    "\n",
    "image_data = pd.merge(camera_data,new_vicon, on='timestamp').reset_index(drop=True)\n",
    "image_data.to_csv(r'/home/francesca/Desktop/francesca/data/mav0/cam0/image_data.csv',index = None, header=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "class NewDataset(Dataset):\n",
    "    def __init__(self, csv_file, imu_data, root_dir):       \n",
    "        \"\"\"\n",
    "        Args:\n",
    "            csv_file (string): Path to the csv file with annotations.\n",
    "            root_dir (string): Directory with all the images.\n",
    "            transform (callable, optional): Optional transform to be applied\n",
    "                on a sample.\n",
    "        \"\"\"\n",
    "        self.frame = pd.read_csv(csv_file)\n",
    "        self.imu_data = imu_data\n",
    "        # Second - 7th columns contains the angular rate and acceleration\n",
    "        self.imu_arr = self.imu_data[:,:,1:]\n",
    "\n",
    "        # Second column contains the image name\n",
    "        self.image_arr = np.asarray(self.frame.iloc[:, 1])\n",
    "        self.label_arr = np.asarray(self.frame.iloc[:, 2:])\n",
    "        self.root_dir = root_dir\n",
    "        # Transforms\n",
    "        transformation = transforms.Compose([\n",
    "            transforms.Resize(64),\n",
    "            transforms.CenterCrop((64,96)),\n",
    "            transforms.ToTensor(),\n",
    "#             transforms.Normalize([0.4543, 0.4543, 0.4543],[0.3035, 0.3035, 0.3035])\n",
    "        ])\n",
    "        self.transform = transformation\n",
    "        # Calculate len\n",
    "        self.data_len = len(self.frame.index)\n",
    "        transform_imu = transforms.Compose([\n",
    "            transforms.ToTensor(),\n",
    "        ])\n",
    "        self.transform_imu = transform_imu\n",
    "    def __len__(self):\n",
    "        return self.data_len\n",
    "    \n",
    "    def __getitem__(self, idx):\n",
    "#         print(idx)\n",
    "        single_image_name = self.image_arr[idx]\n",
    "        fp = self.root_dir\n",
    "        # Open image\n",
    "#         image = PIL.Image.open(fp + single_image_name).convert('RGB')\n",
    "        image = Image.open(fp + single_image_name).convert('RGB')\n",
    "        image = self.transform(image)\n",
    "        imu_data = self.imu_arr[idx]\n",
    "#         imu_data = self.transform_imu( imu_data)\n",
    "        # Get label of the image based on the cropped pandas column\n",
    "        labels = self.label_arr[idx]\n",
    "\n",
    "        return image, imu_data, labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = NewDataset(csv_file='/home/francesca/Desktop/francesca/data/mav0/cam0/file.csv', \n",
    "                         imu_data=new_imu_values,\n",
    "                           root_dir='/home/francesca/Desktop/francesca/data/mav0/cam0/data/')\n",
    "\n",
    "dataloader = DataLoader(dataset, batch_size=32,\n",
    "                        shuffle=False, num_workers=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 32\n",
    "validation_split = .2\n",
    "shuffle_dataset = True\n",
    "random_seed= 42\n",
    "\n",
    "# Creating data indices for training and validation splits:\n",
    "dataset_size = len(dataset)\n",
    "indices = list(range(dataset_size))\n",
    "split = int(np.floor(validation_split * dataset_size))\n",
    "if shuffle_dataset :\n",
    "    np.random.seed(random_seed)\n",
    "    np.random.shuffle(indices)\n",
    "train_indices, val_indices = indices[split:], indices[:split]\n",
    "\n",
    "# Creating PT data samplers and loaders:\n",
    "train_sampler = SubsetRandomSampler(train_indices)\n",
    "valid_sampler = SubsetRandomSampler(val_indices)\n",
    "\n",
    "train_loader = torch.utils.data.DataLoader(dataset, batch_size=batch_size, \n",
    "                                           sampler=train_sampler)\n",
    "validation_loader_1 = torch.utils.data.DataLoader(dataset, batch_size=batch_size,\n",
    "                                                sampler=valid_sampler)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 32\n",
    "validation_split = .2\n",
    "shuffle_dataset = False\n",
    "random_seed= 42\n",
    "\n",
    "# Creating data indices for training and validation splits:\n",
    "dataset_size = len(dataset)\n",
    "indices = list(range(dataset_size))\n",
    "split = int(np.floor(validation_split * dataset_size))\n",
    "if shuffle_dataset :\n",
    "    np.random.seed(random_seed)\n",
    "    np.random.shuffle(indices)\n",
    "train_indices, val_indices = indices[split:], indices[:split]\n",
    "\n",
    "# Creating PT data samplers and loaders:\n",
    "train_sampler = SubsetRandomSampler(train_indices)\n",
    "valid_sampler = SubsetRandomSampler(val_indices)\n",
    "\n",
    "train_loader_1 = torch.utils.data.DataLoader(dataset, batch_size=batch_size, \n",
    "                                           sampler=train_sampler)\n",
    "validation_loader = torch.utils.data.DataLoader(dataset, batch_size=batch_size,\n",
    "                                                sampler=valid_sampler)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def quaternion_angular_error(q1, q2):\n",
    "    \"\"\"\n",
    "    angular error between two quaternions\n",
    "    :param q1: (4, )\n",
    "    :param q2: (4, )\n",
    "    :return:\n",
    "    \"\"\"\n",
    "    d = abs(np.dot(q1, q2))\n",
    "    abs_q1 = np.linalg.norm(q1)\n",
    "    abs_q2 = np.linalg.norm(q2)\n",
    "    d = d / (abs_q1 * abs_q2)\n",
    "    d = min(1.0, max(-1.0, d))\n",
    "    theta = 2 * np.arccos(d) * 180 / np.pi\n",
    "    return theta"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Flatten(nn.Module):\n",
    "    def forward(self, input):\n",
    "        return input.view(input.size(0), -1)\n",
    "\n",
    "# class UnFlatten(nn.Module):\n",
    "#     def forward(self, input):\n",
    "#         return input.view(input.size(0), -1, 1, 1)\n",
    "\n",
    "\n",
    "class FlattenIMU(nn.Module):\n",
    "    def forward(self, input):\n",
    "        return input.view(32, -1)\n",
    "\n",
    "class IMUNet(torch.nn.Module):\n",
    "    \n",
    "    def __init__(self, feature_extractor, num_features=128, dropout=0.5,\n",
    "                 track_running_stats=False, pretrained=False):\n",
    "        super(IMUNet, self).__init__()\n",
    "        \n",
    "\n",
    "        self.dropout = dropout\n",
    "        self.feature_extractor = feature_extractor\n",
    "        self.feature_extractor.avgpool = torch.nn.AdaptiveAvgPool2d(1)\n",
    "        fc_in_features = self.feature_extractor.fc.in_features\n",
    "        \n",
    "        \n",
    "        self.IMU_fc = nn.Sequential(\n",
    "            Flatten(),\n",
    "            nn.Linear(7, 1024),\n",
    "            nn.ReLU(),\n",
    "            nn.Linear(1024, 32),\n",
    "            nn.ReLU(),\n",
    "            FlattenIMU()\n",
    "        )\n",
    "        \n",
    "        self.rnn = torch.nn.LSTM(\n",
    "            input_size=6, \n",
    "            hidden_size= 64*2, \n",
    "            num_layers=4,\n",
    "            batch_first=True)\n",
    "        \n",
    "        self.flat_imu = FlattenIMU()\n",
    "        self.linear = nn.Linear(128,num_features)\n",
    "\n",
    "        self.fc3 = torch.nn.Linear( 4096, num_features)\n",
    "        self.feature_extractor.fc = torch.nn.Linear(fc_in_features, num_features)\n",
    "        # Translation\n",
    "        self.fc_xyz = torch.nn.Linear(num_features, 3)\n",
    "\n",
    "        # Rotation in quaternions\n",
    "        self.fc_quat = torch.nn.Linear(num_features, 4)\n",
    "        \n",
    "    def extract_features(self, image):\n",
    "        x_features = self.feature_extractor(image)\n",
    "        x_features = F.relu(x_features)\n",
    "        if self.dropout > 0:\n",
    "            x_features = F.dropout(x_features, p=self.dropout, training=self.training)\n",
    "        return x_features\n",
    "    \n",
    "    def forward(self, data_input):\n",
    "        image = data_input[0].cuda()\n",
    "        imu_data = data_input[1].type(torch.FloatTensor).cuda()\n",
    "#         print(imu_data.size()[1])\n",
    "        #Images\n",
    "        if type(image) is list:\n",
    "            x_features = [self.extract_features(xi) for xi in image]\n",
    "        elif torch.is_tensor(image):\n",
    "            x_features = self.extract_features(image)\n",
    "            \n",
    "        imu_feat = imu_data\n",
    "        r_out, (h_n, h_c) = self.rnn(imu_feat)\n",
    "        imu_out = self.linear(r_out[:, -1, :]) # we want just the last time \n",
    "\n",
    "\n",
    "        imu_out = self.flat_imu(imu_out) # we want just the last time \n",
    "        \n",
    "        x = torch.cat((x_features, imu_out), dim=1)\n",
    "#         print(x.size)\n",
    "        x = F.relu(self.fc3(x))\n",
    "\n",
    "        if type(x) is list:\n",
    "            x_translations = [self.fc_xyz(xi) for xi in x]\n",
    "            x_rotations = [self.fc_quat(xi) for xi in x]\n",
    "            x_poses = [torch.cat((xt, xr), dim=1) for xt, xr in zip(x_translations, x_rotations)]  \n",
    "        elif torch.is_tensor(x):\n",
    "            x_translations = self.fc_xyz(x) \n",
    "            x_rotations = self.fc_quat(x)\n",
    "\n",
    "            x_poses = torch.cat((x_translations, x_rotations), dim=1)\n",
    "\n",
    "        return x_poses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torchvision.models as models\n",
    "import torch.optim as optim\n",
    "from PIL import Image\n",
    "\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "\n",
    "\n",
    "# Create pretrained feature extractor\n",
    "feature_extractor = models.resnet18(pretrained=True)\n",
    "\n",
    "# Num features for the last layer before pose regressor\n",
    "num_features = 2048\n",
    "\n",
    "# Create model\n",
    "model = IMUNet(feature_extractor, num_features=num_features, pretrained=True)\n",
    "model = model.to(device)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([32, 7])"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = IMUNet(feature_extractor, num_features=num_features, pretrained=True)\n",
    "model = model.to(device)\n",
    "\n",
    "\n",
    "out = model(next(iter(dataloader))[0:2])\n",
    "out.size()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torchvision.models as models\n",
    "import torch.optim as optim\n",
    "\n",
    "\n",
    "# Create pretrained feature extractor\n",
    "feature_extractor = models.resnet18(pretrained=True)\n",
    "\n",
    "\n",
    "# print(Net(feature_extractor))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Select primary device\n",
    "\n",
    "import os\n",
    "\n",
    "if torch.cuda.is_available():\n",
    "    device = torch.device('cuda')\n",
    "else:\n",
    "    device = torch.device('cpu')\n",
    "    \n",
    "   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "class AverageMeter():\n",
    "    def __init__(self):\n",
    "        self.reset()\n",
    "\n",
    "    def reset(self):\n",
    "        self.val = 0\n",
    "        self.avg = 0\n",
    "        self.sum = 0\n",
    "        self.count = 0\n",
    "\n",
    "    def update(self, value, n=1):\n",
    "        self.val = value\n",
    "        self.count += n\n",
    "        self.sum += value * n\n",
    "        self.avg = self.sum / self.count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train(train_loader, model, criterion, optimizer, epoch, max_epoch, log_freq=1, print_sum=True):\n",
    "    \n",
    "    position_loss=[]\n",
    "    orientation_loss = []\n",
    "    \n",
    "    model.train()\n",
    "    losses = AverageMeter()\n",
    "    epoch_time = time.time()\n",
    "    \n",
    "    gt_poses = np.empty((0, 7))\n",
    "    pred_poses = np.empty((0, 7))\n",
    "    \n",
    "    end = time.time()\n",
    "    \n",
    "    for idx, data in enumerate(train_loader):\n",
    "        if idx == 1920:\n",
    "            idx = 1919\n",
    "        batch_images= data[0].cuda()\n",
    "        imu_data= data[1].unsqueeze(0).cuda()\n",
    "    \n",
    "        batch_poses= data[2].cuda()\n",
    "        data_time = (time.time() - end)\n",
    "        batch_images = batch_images.to(device)\n",
    "        batch_poses = batch_poses.to(device)\n",
    "        batch_poses = torch.tensor(batch_poses, dtype=torch.float, device=device)\n",
    "#         imu_data = imu_data.to(device).double()\n",
    "        imu_data = torch.tensor(imu_data, dtype=torch.float, device=device)\n",
    "        out = model(data[0:2])\n",
    "\n",
    "        loss = criterion(out, batch_poses)\n",
    "              \n",
    "        # Training step\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        \n",
    "        losses.update(loss.data[0], len(batch_images) * batch_images.size(0))\n",
    "        \n",
    "        \n",
    "        # move data to cpu & numpy\n",
    "        bp = batch_poses.detach().cpu().numpy()\n",
    "        outp = out.detach().cpu().numpy()\n",
    "        gt_poses = np.vstack((gt_poses, bp))\n",
    "        pred_poses = np.vstack((pred_poses, outp))\n",
    "        \n",
    "        batch_time = (time.time() - end)\n",
    "        end = time.time()\n",
    "        \n",
    "        if log_freq != 0 and idx % log_freq == 0:\n",
    "            print('Epoch: [{}/{}]\\tBatch: [{}/{}]\\t'\n",
    "                  'Time: {batch_time:.3f}\\t'\n",
    "                  'Data Time: {data_time:.3f}\\t'\n",
    "                  'Loss: {losses.val:.3f}\\t'\n",
    "                  'Avg Loss: {losses.avg:.3f}\\t'.format(\n",
    "                   epoch, max_epoch - 1, idx, len(train_loader) - 1,\n",
    "                   batch_time=batch_time, data_time=data_time, losses=losses))\n",
    "            \n",
    "        t_loss = np.asarray([np.linalg.norm(p - t) for p, t in zip(pred_poses[:, :3], gt_poses[:, :3])])\n",
    "        q_loss = np.asarray([quaternion_angular_error(p, t) for p, t in zip(pred_poses[:, 3:], gt_poses[:, 3:])])\n",
    "    position_loss.append(t_loss)\n",
    "    orientation_loss.append(q_loss)\n",
    "\n",
    "    return losses\n",
    "    \n",
    "def validate(val_loader, model, criterion, epoch, log_freq=1, print_sum=True):\n",
    "    \n",
    "    losses = AverageMeter()\n",
    "    position_loss_testing=[]\n",
    "    orientation_loss_testing=[]\n",
    "    \n",
    "    # set model to evaluation\n",
    "    model.eval()\n",
    "    \n",
    "    with torch.no_grad():\n",
    "        epoch_time = time.time()\n",
    "        end = time.time()\n",
    "        for idx, data in enumerate(val_loader):\n",
    "\n",
    "\n",
    "            batch_images= data[0].cuda()\n",
    "            imu_data= data[1].cuda()\n",
    "        \n",
    "            batch_poses= data[2].cuda()\n",
    "            data_time = time.time() - end\n",
    "            \n",
    "            batch_images = batch_images.to(device)\n",
    "            batch_poses = batch_poses.to(device)\n",
    "            batch_poses = torch.tensor(batch_poses, dtype=torch.float, device=device)\n",
    "            imu_data = imu_data.to(device)\n",
    "            imu_data = torch.tensor(imu_data, dtype=torch.float, device=device)\n",
    "\n",
    "\n",
    "            # compute model output\n",
    "            out = model(data[:2])\n",
    "            loss = criterion(out, batch_poses)\n",
    "            \n",
    "            losses.update(loss.data[0], len(data[0]) * data[0].size(0))\n",
    "\n",
    "\n",
    "            batch_time = time.time() - end\n",
    "            end = time.time()\n",
    "            if log_freq != 0 and idx % log_freq == 0:\n",
    "                print('Val Epoch: {}\\t'\n",
    "                      'Time: {batch_time:.3f}\\t'\n",
    "                      'Data Time: {data_time:.3f}\\t'\n",
    "                      'Loss: {losses.val:.3f}\\t'\n",
    "                      'Avg Loss: {losses.avg:.3f}'.format(\n",
    "                       epoch, batch_time=batch_time, data_time=data_time, losses=losses))\n",
    "\n",
    "    return losses\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torchvision.models as models\n",
    "import torch.optim as optim\n",
    "\n",
    "\n",
    "# Create pretrained feature extractor\n",
    "feature_extractor = models.resnet18(pretrained=True)\n",
    "\n",
    "# Num features for tahe last layer before pose regressor\n",
    "num_features = 2048\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "imu = next(iter(train_loader))[1]\n",
    "# Create model\n",
    "model = IMUNet(feature_extractor, num_features=num_features, pretrained=True)\n",
    "model = model.to(device)\n",
    "\n",
    "# Criterion\n",
    "criterion = NetCriterion(learn_beta=True)\n",
    "criterion = criterion.to(device)\n",
    "\n",
    "# Add all params for optimization\n",
    "param_list = [{'params': model.parameters()}]\n",
    "if criterion.learn_beta:\n",
    "    # Add sx and sq from loss function to optimizer params\n",
    "    param_list.append({'params': criterion.parameters()})\n",
    "\n",
    "# Create optimizer\n",
    "# optimizer = optim.Adam(params=param_list, lr=1e-5, weight_decay=0.0005)\n",
    "optimizer = optim.Adam(params=param_list, lr=0.01, weight_decay=0.0001)\n",
    "\n",
    "\n",
    "# Epochs to train\n",
    "start_epoch = 0\n",
    "\n",
    "# n_epochs = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training ...\n",
      "Epoch: [0/619]\tBatch: [0/47]\tTime: 0.221\tData Time: 0.164\tLoss: 5.847\tAvg Loss: 5.847\t\n",
      "Val Epoch: 0\tTime: 0.169\tData Time: 0.150\tLoss: 5.040\tAvg Loss: 5.040\n",
      "Epoch: [1/619]\tBatch: [0/47]\tTime: 0.207\tData Time: 0.152\tLoss: 5.352\tAvg Loss: 5.352\t\n",
      "Epoch: [2/619]\tBatch: [0/47]\tTime: 0.199\tData Time: 0.161\tLoss: 5.499\tAvg Loss: 5.499\t\n",
      "Epoch: [3/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 5.715\tAvg Loss: 5.715\t\n",
      "Epoch: [4/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 5.272\tAvg Loss: 5.272\t\n",
      "Epoch: [5/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 5.652\tAvg Loss: 5.652\t\n",
      "Epoch: [6/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.149\tLoss: 5.224\tAvg Loss: 5.224\t\n",
      "Epoch: [7/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 5.846\tAvg Loss: 5.846\t\n",
      "Epoch: [8/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.148\tLoss: 5.593\tAvg Loss: 5.593\t\n",
      "Epoch: [9/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 5.031\tAvg Loss: 5.031\t\n",
      "Epoch: [10/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.149\tLoss: 5.101\tAvg Loss: 5.101\t\n",
      "Val Epoch: 10\tTime: 0.162\tData Time: 0.152\tLoss: 4.602\tAvg Loss: 4.602\n",
      "Epoch: [11/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.148\tLoss: 4.975\tAvg Loss: 4.975\t\n",
      "Epoch: [12/619]\tBatch: [0/47]\tTime: 0.203\tData Time: 0.148\tLoss: 5.498\tAvg Loss: 5.498\t\n",
      "Epoch: [13/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 5.280\tAvg Loss: 5.280\t\n",
      "Epoch: [14/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 5.144\tAvg Loss: 5.144\t\n",
      "Epoch: [15/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.144\tLoss: 4.991\tAvg Loss: 4.991\t\n",
      "Epoch: [16/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 4.935\tAvg Loss: 4.935\t\n",
      "Epoch: [17/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 4.678\tAvg Loss: 4.678\t\n",
      "Epoch: [18/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.144\tLoss: 5.034\tAvg Loss: 5.034\t\n",
      "Epoch: [19/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.144\tLoss: 4.427\tAvg Loss: 4.427\t\n",
      "Epoch: [20/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.150\tLoss: 4.284\tAvg Loss: 4.284\t\n",
      "Val Epoch: 20\tTime: 0.177\tData Time: 0.150\tLoss: 4.077\tAvg Loss: 4.077\n",
      "Epoch: [21/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.146\tLoss: 4.695\tAvg Loss: 4.695\t\n",
      "Epoch: [22/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.147\tLoss: 4.567\tAvg Loss: 4.567\t\n",
      "Epoch: [23/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.143\tLoss: 4.511\tAvg Loss: 4.511\t\n",
      "Epoch: [24/619]\tBatch: [0/47]\tTime: 0.193\tData Time: 0.148\tLoss: 4.602\tAvg Loss: 4.602\t\n",
      "Epoch: [25/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.147\tLoss: 4.395\tAvg Loss: 4.395\t\n",
      "Epoch: [26/619]\tBatch: [0/47]\tTime: 0.200\tData Time: 0.149\tLoss: 4.118\tAvg Loss: 4.118\t\n",
      "Epoch: [27/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 4.460\tAvg Loss: 4.460\t\n",
      "Epoch: [28/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.153\tLoss: 4.368\tAvg Loss: 4.368\t\n",
      "Epoch: [29/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.148\tLoss: 4.571\tAvg Loss: 4.571\t\n",
      "Epoch: [30/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.150\tLoss: 4.002\tAvg Loss: 4.002\t\n",
      "Val Epoch: 30\tTime: 0.175\tData Time: 0.150\tLoss: 3.631\tAvg Loss: 3.631\n",
      "Epoch: [31/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.145\tLoss: 4.486\tAvg Loss: 4.486\t\n",
      "Epoch: [32/619]\tBatch: [0/47]\tTime: 0.211\tData Time: 0.146\tLoss: 4.242\tAvg Loss: 4.242\t\n",
      "Epoch: [33/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.150\tLoss: 4.119\tAvg Loss: 4.119\t\n",
      "Epoch: [34/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.146\tLoss: 4.325\tAvg Loss: 4.325\t\n",
      "Epoch: [35/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 4.133\tAvg Loss: 4.133\t\n",
      "Epoch: [36/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.144\tLoss: 4.417\tAvg Loss: 4.417\t\n",
      "Epoch: [37/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.146\tLoss: 4.131\tAvg Loss: 4.131\t\n",
      "Epoch: [38/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.151\tLoss: 3.964\tAvg Loss: 3.964\t\n",
      "Epoch: [39/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.146\tLoss: 4.233\tAvg Loss: 4.233\t\n",
      "Epoch: [40/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.146\tLoss: 4.006\tAvg Loss: 4.006\t\n",
      "Val Epoch: 40\tTime: 0.165\tData Time: 0.152\tLoss: 3.383\tAvg Loss: 3.383\n",
      "Epoch: [41/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 4.080\tAvg Loss: 4.080\t\n",
      "Epoch: [42/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.152\tLoss: 3.703\tAvg Loss: 3.703\t\n",
      "Epoch: [43/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.141\tLoss: 4.177\tAvg Loss: 4.177\t\n",
      "Epoch: [44/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 4.010\tAvg Loss: 4.010\t\n",
      "Epoch: [45/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 3.972\tAvg Loss: 3.972\t\n",
      "Epoch: [46/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.148\tLoss: 3.841\tAvg Loss: 3.841\t\n",
      "Epoch: [47/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.143\tLoss: 3.917\tAvg Loss: 3.917\t\n",
      "Epoch: [48/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.145\tLoss: 3.905\tAvg Loss: 3.905\t\n",
      "Epoch: [49/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.145\tLoss: 3.718\tAvg Loss: 3.718\t\n",
      "Epoch: [50/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.145\tLoss: 3.792\tAvg Loss: 3.792\t\n",
      "Val Epoch: 50\tTime: 0.164\tData Time: 0.151\tLoss: 3.187\tAvg Loss: 3.187\n",
      "Epoch: [51/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 3.894\tAvg Loss: 3.894\t\n",
      "Epoch: [52/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 3.902\tAvg Loss: 3.902\t\n",
      "Epoch: [53/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 3.860\tAvg Loss: 3.860\t\n",
      "Epoch: [54/619]\tBatch: [0/47]\tTime: 0.204\tData Time: 0.144\tLoss: 3.715\tAvg Loss: 3.715\t\n",
      "Epoch: [55/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.145\tLoss: 3.804\tAvg Loss: 3.804\t\n",
      "Epoch: [56/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.147\tLoss: 3.592\tAvg Loss: 3.592\t\n",
      "Epoch: [57/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.148\tLoss: 3.576\tAvg Loss: 3.576\t\n",
      "Epoch: [58/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.144\tLoss: 3.957\tAvg Loss: 3.957\t\n",
      "Epoch: [59/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 3.656\tAvg Loss: 3.656\t\n",
      "Epoch: [60/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 3.414\tAvg Loss: 3.414\t\n",
      "Val Epoch: 60\tTime: 0.158\tData Time: 0.148\tLoss: 2.947\tAvg Loss: 2.947\n",
      "Epoch: [61/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 3.740\tAvg Loss: 3.740\t\n",
      "Epoch: [62/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.146\tLoss: 3.789\tAvg Loss: 3.789\t\n",
      "Epoch: [63/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.145\tLoss: 3.722\tAvg Loss: 3.722\t\n",
      "Epoch: [64/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.150\tLoss: 3.563\tAvg Loss: 3.563\t\n",
      "Epoch: [65/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 3.553\tAvg Loss: 3.553\t\n",
      "Epoch: [66/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 3.516\tAvg Loss: 3.516\t\n",
      "Epoch: [67/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 3.678\tAvg Loss: 3.678\t\n",
      "Epoch: [68/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.142\tLoss: 3.723\tAvg Loss: 3.723\t\n",
      "Epoch: [69/619]\tBatch: [0/47]\tTime: 0.204\tData Time: 0.148\tLoss: 3.560\tAvg Loss: 3.560\t\n",
      "Epoch: [70/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.148\tLoss: 3.638\tAvg Loss: 3.638\t\n",
      "Val Epoch: 70\tTime: 0.159\tData Time: 0.150\tLoss: 2.707\tAvg Loss: 2.707\n",
      "Epoch: [71/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.141\tLoss: 3.756\tAvg Loss: 3.756\t\n",
      "Epoch: [72/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 3.196\tAvg Loss: 3.196\t\n",
      "Epoch: [73/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.144\tLoss: 3.332\tAvg Loss: 3.332\t\n",
      "Epoch: [74/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 3.189\tAvg Loss: 3.189\t\n",
      "Epoch: [75/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 3.172\tAvg Loss: 3.172\t\n",
      "Epoch: [76/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.145\tLoss: 3.221\tAvg Loss: 3.221\t\n",
      "Epoch: [77/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.143\tLoss: 3.333\tAvg Loss: 3.333\t\n",
      "Epoch: [78/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.144\tLoss: 3.422\tAvg Loss: 3.422\t\n",
      "Epoch: [79/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.147\tLoss: 2.887\tAvg Loss: 2.887\t\n",
      "Epoch: [80/619]\tBatch: [0/47]\tTime: 0.203\tData Time: 0.147\tLoss: 3.181\tAvg Loss: 3.181\t\n",
      "Val Epoch: 80\tTime: 0.161\tData Time: 0.151\tLoss: 2.411\tAvg Loss: 2.411\n",
      "Epoch: [81/619]\tBatch: [0/47]\tTime: 0.193\tData Time: 0.145\tLoss: 3.352\tAvg Loss: 3.352\t\n",
      "Epoch: [82/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 2.909\tAvg Loss: 2.909\t\n",
      "Epoch: [83/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.143\tLoss: 3.022\tAvg Loss: 3.022\t\n",
      "Epoch: [84/619]\tBatch: [0/47]\tTime: 0.200\tData Time: 0.148\tLoss: 2.954\tAvg Loss: 2.954\t\n",
      "Epoch: [85/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 2.789\tAvg Loss: 2.789\t\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: [86/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.146\tLoss: 3.068\tAvg Loss: 3.068\t\n",
      "Epoch: [87/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.148\tLoss: 2.895\tAvg Loss: 2.895\t\n",
      "Epoch: [88/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.149\tLoss: 2.689\tAvg Loss: 2.689\t\n",
      "Epoch: [89/619]\tBatch: [0/47]\tTime: 0.197\tData Time: 0.148\tLoss: 2.767\tAvg Loss: 2.767\t\n",
      "Epoch: [90/619]\tBatch: [0/47]\tTime: 0.193\tData Time: 0.146\tLoss: 2.709\tAvg Loss: 2.709\t\n",
      "Val Epoch: 90\tTime: 0.159\tData Time: 0.150\tLoss: 2.038\tAvg Loss: 2.038\n",
      "Epoch: [91/619]\tBatch: [0/47]\tTime: 0.199\tData Time: 0.152\tLoss: 2.571\tAvg Loss: 2.571\t\n",
      "Epoch: [92/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.147\tLoss: 2.599\tAvg Loss: 2.599\t\n",
      "Epoch: [93/619]\tBatch: [0/47]\tTime: 0.193\tData Time: 0.149\tLoss: 2.310\tAvg Loss: 2.310\t\n",
      "Epoch: [94/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 2.220\tAvg Loss: 2.220\t\n",
      "Epoch: [95/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.150\tLoss: 2.610\tAvg Loss: 2.610\t\n",
      "Epoch: [96/619]\tBatch: [0/47]\tTime: 0.202\tData Time: 0.156\tLoss: 2.760\tAvg Loss: 2.760\t\n",
      "Epoch: [97/619]\tBatch: [0/47]\tTime: 0.198\tData Time: 0.148\tLoss: 2.411\tAvg Loss: 2.411\t\n",
      "Epoch: [98/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.146\tLoss: 2.590\tAvg Loss: 2.590\t\n",
      "Epoch: [99/619]\tBatch: [0/47]\tTime: 0.193\tData Time: 0.146\tLoss: 2.109\tAvg Loss: 2.109\t\n",
      "Epoch: [100/619]\tBatch: [0/47]\tTime: 0.197\tData Time: 0.150\tLoss: 2.391\tAvg Loss: 2.391\t\n",
      "Val Epoch: 100\tTime: 0.160\tData Time: 0.151\tLoss: 1.485\tAvg Loss: 1.485\n",
      "Epoch: [101/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 1.859\tAvg Loss: 1.859\t\n",
      "Epoch: [102/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 2.172\tAvg Loss: 2.172\t\n",
      "Epoch: [103/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.148\tLoss: 2.300\tAvg Loss: 2.300\t\n",
      "Epoch: [104/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.148\tLoss: 2.190\tAvg Loss: 2.190\t\n",
      "Epoch: [105/619]\tBatch: [0/47]\tTime: 0.203\tData Time: 0.147\tLoss: 2.030\tAvg Loss: 2.030\t\n",
      "Epoch: [106/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 2.054\tAvg Loss: 2.054\t\n",
      "Epoch: [107/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.144\tLoss: 2.240\tAvg Loss: 2.240\t\n",
      "Epoch: [108/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.145\tLoss: 1.811\tAvg Loss: 1.811\t\n",
      "Epoch: [109/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 1.910\tAvg Loss: 1.910\t\n",
      "Epoch: [110/619]\tBatch: [0/47]\tTime: 0.193\tData Time: 0.147\tLoss: 1.595\tAvg Loss: 1.595\t\n",
      "Val Epoch: 110\tTime: 0.173\tData Time: 0.151\tLoss: 1.117\tAvg Loss: 1.117\n",
      "Epoch: [111/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.620\tAvg Loss: 1.620\t\n",
      "Epoch: [112/619]\tBatch: [0/47]\tTime: 0.200\tData Time: 0.148\tLoss: 1.634\tAvg Loss: 1.634\t\n",
      "Epoch: [113/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.816\tAvg Loss: 1.816\t\n",
      "Epoch: [114/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.149\tLoss: 1.337\tAvg Loss: 1.337\t\n",
      "Epoch: [115/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.150\tLoss: 1.401\tAvg Loss: 1.401\t\n",
      "Epoch: [116/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.147\tLoss: 1.240\tAvg Loss: 1.240\t\n",
      "Epoch: [117/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.347\tAvg Loss: 1.347\t\n",
      "Epoch: [118/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 1.411\tAvg Loss: 1.411\t\n",
      "Epoch: [119/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.134\tAvg Loss: 1.134\t\n",
      "Epoch: [120/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.145\tLoss: 1.253\tAvg Loss: 1.253\t\n",
      "Val Epoch: 120\tTime: 0.161\tData Time: 0.152\tLoss: 0.413\tAvg Loss: 0.413\n",
      "Epoch: [121/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 1.207\tAvg Loss: 1.207\t\n",
      "Epoch: [122/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.149\tLoss: 0.920\tAvg Loss: 0.920\t\n",
      "Epoch: [123/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.142\tLoss: 0.945\tAvg Loss: 0.945\t\n",
      "Epoch: [124/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.149\tLoss: 0.713\tAvg Loss: 0.713\t\n",
      "Epoch: [125/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 0.591\tAvg Loss: 0.591\t\n",
      "Epoch: [126/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.147\tLoss: 0.978\tAvg Loss: 0.978\t\n",
      "Epoch: [127/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.145\tLoss: 0.618\tAvg Loss: 0.618\t\n",
      "Epoch: [128/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.149\tLoss: 0.565\tAvg Loss: 0.565\t\n",
      "Epoch: [129/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.152\tLoss: 0.571\tAvg Loss: 0.571\t\n",
      "Epoch: [130/619]\tBatch: [0/47]\tTime: 0.205\tData Time: 0.147\tLoss: 0.471\tAvg Loss: 0.471\t\n",
      "Val Epoch: 130\tTime: 0.160\tData Time: 0.150\tLoss: 0.092\tAvg Loss: 0.092\n",
      "Epoch: [131/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 0.807\tAvg Loss: 0.807\t\n",
      "Epoch: [132/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.149\tLoss: 0.662\tAvg Loss: 0.662\t\n",
      "Epoch: [133/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.146\tLoss: 0.307\tAvg Loss: 0.307\t\n",
      "Epoch: [134/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 0.418\tAvg Loss: 0.418\t\n",
      "Epoch: [135/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.148\tLoss: 0.541\tAvg Loss: 0.541\t\n",
      "Epoch: [136/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.148\tLoss: 0.729\tAvg Loss: 0.729\t\n",
      "Epoch: [137/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 0.253\tAvg Loss: 0.253\t\n",
      "Epoch: [138/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.147\tLoss: 0.662\tAvg Loss: 0.662\t\n",
      "Epoch: [139/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.155\tLoss: 0.450\tAvg Loss: 0.450\t\n",
      "Epoch: [140/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 0.431\tAvg Loss: 0.431\t\n",
      "Val Epoch: 140\tTime: 0.161\tData Time: 0.152\tLoss: -0.004\tAvg Loss: -0.004\n",
      "Epoch: [141/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.147\tLoss: 0.267\tAvg Loss: 0.267\t\n",
      "Epoch: [142/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.147\tLoss: 0.968\tAvg Loss: 0.968\t\n",
      "Epoch: [143/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.147\tLoss: 2.034\tAvg Loss: 2.034\t\n",
      "Epoch: [144/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 1.865\tAvg Loss: 1.865\t\n",
      "Epoch: [145/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.152\tLoss: 1.403\tAvg Loss: 1.403\t\n",
      "Epoch: [146/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.140\tAvg Loss: 1.140\t\n",
      "Epoch: [147/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.144\tLoss: 0.968\tAvg Loss: 0.968\t\n",
      "Epoch: [148/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 1.011\tAvg Loss: 1.011\t\n",
      "Epoch: [149/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.141\tLoss: 1.085\tAvg Loss: 1.085\t\n",
      "Epoch: [150/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.151\tLoss: 0.962\tAvg Loss: 0.962\t\n",
      "Val Epoch: 150\tTime: 0.162\tData Time: 0.152\tLoss: 0.514\tAvg Loss: 0.514\n",
      "Epoch: [151/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.145\tLoss: 0.828\tAvg Loss: 0.828\t\n",
      "Epoch: [152/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.143\tLoss: 1.007\tAvg Loss: 1.007\t\n",
      "Epoch: [153/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 2.236\tAvg Loss: 2.236\t\n",
      "Epoch: [154/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.152\tLoss: 0.613\tAvg Loss: 0.613\t\n",
      "Epoch: [155/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 0.560\tAvg Loss: 0.560\t\n",
      "Epoch: [156/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.145\tLoss: 0.831\tAvg Loss: 0.831\t\n",
      "Epoch: [157/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 0.642\tAvg Loss: 0.642\t\n",
      "Epoch: [158/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 1.102\tAvg Loss: 1.102\t\n",
      "Epoch: [159/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 0.408\tAvg Loss: 0.408\t\n",
      "Epoch: [160/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.142\tLoss: 0.900\tAvg Loss: 0.900\t\n",
      "Val Epoch: 160\tTime: 0.157\tData Time: 0.148\tLoss: 0.495\tAvg Loss: 0.495\n",
      "Epoch: [161/619]\tBatch: [0/47]\tTime: 0.200\tData Time: 0.147\tLoss: 0.849\tAvg Loss: 0.849\t\n",
      "Epoch: [162/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.146\tLoss: 0.779\tAvg Loss: 0.779\t\n",
      "Epoch: [163/619]\tBatch: [0/47]\tTime: 0.198\tData Time: 0.147\tLoss: 0.948\tAvg Loss: 0.948\t\n",
      "Epoch: [164/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.146\tLoss: 0.771\tAvg Loss: 0.771\t\n",
      "Epoch: [165/619]\tBatch: [0/47]\tTime: 0.197\tData Time: 0.149\tLoss: 0.841\tAvg Loss: 0.841\t\n",
      "Epoch: [166/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 0.838\tAvg Loss: 0.838\t\n",
      "Epoch: [167/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.144\tLoss: 0.924\tAvg Loss: 0.924\t\n",
      "Epoch: [168/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.146\tLoss: 0.583\tAvg Loss: 0.583\t\n",
      "Epoch: [169/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.144\tLoss: 0.777\tAvg Loss: 0.777\t\n",
      "Epoch: [170/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.143\tLoss: 0.626\tAvg Loss: 0.626\t\n",
      "Val Epoch: 170\tTime: 0.160\tData Time: 0.150\tLoss: 0.310\tAvg Loss: 0.310\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: [171/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.149\tLoss: 0.654\tAvg Loss: 0.654\t\n",
      "Epoch: [172/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 0.616\tAvg Loss: 0.616\t\n",
      "Epoch: [173/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.145\tLoss: 0.675\tAvg Loss: 0.675\t\n",
      "Epoch: [174/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.149\tLoss: 0.527\tAvg Loss: 0.527\t\n",
      "Epoch: [175/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 0.673\tAvg Loss: 0.673\t\n",
      "Epoch: [176/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 0.681\tAvg Loss: 0.681\t\n",
      "Epoch: [177/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.149\tLoss: 0.489\tAvg Loss: 0.489\t\n",
      "Epoch: [178/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.146\tLoss: 0.679\tAvg Loss: 0.679\t\n",
      "Epoch: [179/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.150\tLoss: 0.632\tAvg Loss: 0.632\t\n",
      "Epoch: [180/619]\tBatch: [0/47]\tTime: 0.203\tData Time: 0.146\tLoss: 0.697\tAvg Loss: 0.697\t\n",
      "Val Epoch: 180\tTime: 0.160\tData Time: 0.151\tLoss: 0.403\tAvg Loss: 0.403\n",
      "Epoch: [181/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.143\tLoss: 0.670\tAvg Loss: 0.670\t\n",
      "Epoch: [182/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 0.714\tAvg Loss: 0.714\t\n",
      "Epoch: [183/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.142\tLoss: 0.730\tAvg Loss: 0.730\t\n",
      "Epoch: [184/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 0.615\tAvg Loss: 0.615\t\n",
      "Epoch: [185/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.146\tLoss: 0.488\tAvg Loss: 0.488\t\n",
      "Epoch: [186/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 0.769\tAvg Loss: 0.769\t\n",
      "Epoch: [187/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 0.662\tAvg Loss: 0.662\t\n",
      "Epoch: [188/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 0.704\tAvg Loss: 0.704\t\n",
      "Epoch: [189/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 0.580\tAvg Loss: 0.580\t\n",
      "Epoch: [190/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 0.574\tAvg Loss: 0.574\t\n",
      "Val Epoch: 190\tTime: 0.160\tData Time: 0.150\tLoss: 0.350\tAvg Loss: 0.350\n",
      "Epoch: [191/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.147\tLoss: 0.481\tAvg Loss: 0.481\t\n",
      "Epoch: [192/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.146\tLoss: 0.541\tAvg Loss: 0.541\t\n",
      "Epoch: [193/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 0.363\tAvg Loss: 0.363\t\n",
      "Epoch: [194/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.149\tLoss: 0.616\tAvg Loss: 0.616\t\n",
      "Epoch: [195/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.149\tLoss: 0.271\tAvg Loss: 0.271\t\n",
      "Epoch: [196/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 0.530\tAvg Loss: 0.530\t\n",
      "Epoch: [197/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.147\tLoss: 0.538\tAvg Loss: 0.538\t\n",
      "Epoch: [198/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.148\tLoss: 0.391\tAvg Loss: 0.391\t\n",
      "Epoch: [199/619]\tBatch: [0/47]\tTime: 0.199\tData Time: 0.152\tLoss: 0.578\tAvg Loss: 0.578\t\n",
      "Epoch: [200/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.149\tLoss: 0.407\tAvg Loss: 0.407\t\n",
      "Val Epoch: 200\tTime: 0.169\tData Time: 0.150\tLoss: 0.126\tAvg Loss: 0.126\n",
      "Epoch: [201/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.147\tLoss: 0.506\tAvg Loss: 0.506\t\n",
      "Epoch: [202/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 0.367\tAvg Loss: 0.367\t\n",
      "Epoch: [203/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 0.463\tAvg Loss: 0.463\t\n",
      "Epoch: [204/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.150\tLoss: 0.577\tAvg Loss: 0.577\t\n",
      "Epoch: [205/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 0.444\tAvg Loss: 0.444\t\n",
      "Epoch: [206/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 0.432\tAvg Loss: 0.432\t\n",
      "Epoch: [207/619]\tBatch: [0/47]\tTime: 0.193\tData Time: 0.150\tLoss: 0.354\tAvg Loss: 0.354\t\n",
      "Epoch: [208/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.152\tLoss: 0.356\tAvg Loss: 0.356\t\n",
      "Epoch: [209/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.149\tLoss: 0.335\tAvg Loss: 0.335\t\n",
      "Epoch: [210/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.150\tLoss: 0.812\tAvg Loss: 0.812\t\n",
      "Val Epoch: 210\tTime: 0.162\tData Time: 0.152\tLoss: 0.132\tAvg Loss: 0.132\n",
      "Epoch: [211/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.149\tLoss: 0.358\tAvg Loss: 0.358\t\n",
      "Epoch: [212/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 49.010\tAvg Loss: 49.010\t\n",
      "Epoch: [213/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 0.411\tAvg Loss: 0.411\t\n",
      "Epoch: [214/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.146\tLoss: 0.586\tAvg Loss: 0.586\t\n",
      "Epoch: [215/619]\tBatch: [0/47]\tTime: 0.201\tData Time: 0.147\tLoss: 0.376\tAvg Loss: 0.376\t\n",
      "Epoch: [216/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 0.324\tAvg Loss: 0.324\t\n",
      "Epoch: [217/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.148\tLoss: 0.325\tAvg Loss: 0.325\t\n",
      "Epoch: [218/619]\tBatch: [0/47]\tTime: 0.200\tData Time: 0.144\tLoss: 0.295\tAvg Loss: 0.295\t\n",
      "Epoch: [219/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.147\tLoss: 1.495\tAvg Loss: 1.495\t\n",
      "Epoch: [220/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 0.390\tAvg Loss: 0.390\t\n",
      "Val Epoch: 220\tTime: 0.159\tData Time: 0.149\tLoss: 0.085\tAvg Loss: 0.085\n",
      "Epoch: [221/619]\tBatch: [0/47]\tTime: 0.197\tData Time: 0.149\tLoss: 0.248\tAvg Loss: 0.248\t\n",
      "Epoch: [222/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.145\tLoss: 0.333\tAvg Loss: 0.333\t\n",
      "Epoch: [223/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 0.281\tAvg Loss: 0.281\t\n",
      "Epoch: [224/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 0.380\tAvg Loss: 0.380\t\n",
      "Epoch: [225/619]\tBatch: [0/47]\tTime: 0.203\tData Time: 0.152\tLoss: 0.319\tAvg Loss: 0.319\t\n",
      "Epoch: [226/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.147\tLoss: 0.354\tAvg Loss: 0.354\t\n",
      "Epoch: [227/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.146\tLoss: 1.207\tAvg Loss: 1.207\t\n",
      "Epoch: [228/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 0.894\tAvg Loss: 0.894\t\n",
      "Epoch: [229/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.143\tLoss: 0.723\tAvg Loss: 0.723\t\n",
      "Epoch: [230/619]\tBatch: [0/47]\tTime: 0.212\tData Time: 0.149\tLoss: 0.804\tAvg Loss: 0.804\t\n",
      "Val Epoch: 230\tTime: 0.159\tData Time: 0.149\tLoss: 0.633\tAvg Loss: 0.633\n",
      "Epoch: [231/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.148\tLoss: 0.658\tAvg Loss: 0.658\t\n",
      "Epoch: [232/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.194\tAvg Loss: 1.194\t\n",
      "Epoch: [233/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 0.630\tAvg Loss: 0.630\t\n",
      "Epoch: [234/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.149\tLoss: 0.592\tAvg Loss: 0.592\t\n",
      "Epoch: [235/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.149\tLoss: 0.743\tAvg Loss: 0.743\t\n",
      "Epoch: [236/619]\tBatch: [0/47]\tTime: 0.193\tData Time: 0.144\tLoss: 0.818\tAvg Loss: 0.818\t\n",
      "Epoch: [237/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 0.846\tAvg Loss: 0.846\t\n",
      "Epoch: [238/619]\tBatch: [0/47]\tTime: 0.200\tData Time: 0.145\tLoss: 0.844\tAvg Loss: 0.844\t\n",
      "Epoch: [239/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.048\tAvg Loss: 1.048\t\n",
      "Epoch: [240/619]\tBatch: [0/47]\tTime: 0.202\tData Time: 0.148\tLoss: 0.850\tAvg Loss: 0.850\t\n",
      "Val Epoch: 240\tTime: 0.159\tData Time: 0.149\tLoss: 0.748\tAvg Loss: 0.748\n",
      "Epoch: [241/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.145\tLoss: 0.898\tAvg Loss: 0.898\t\n",
      "Epoch: [242/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 0.905\tAvg Loss: 0.905\t\n",
      "Epoch: [243/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 0.874\tAvg Loss: 0.874\t\n",
      "Epoch: [244/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.146\tLoss: 0.845\tAvg Loss: 0.845\t\n",
      "Epoch: [245/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 2.538\tAvg Loss: 2.538\t\n",
      "Epoch: [246/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 0.798\tAvg Loss: 0.798\t\n",
      "Epoch: [247/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.152\tLoss: 0.769\tAvg Loss: 0.769\t\n",
      "Epoch: [248/619]\tBatch: [0/47]\tTime: 0.206\tData Time: 0.145\tLoss: 0.697\tAvg Loss: 0.697\t\n",
      "Epoch: [249/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.145\tLoss: 0.787\tAvg Loss: 0.787\t\n",
      "Epoch: [250/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 0.703\tAvg Loss: 0.703\t\n",
      "Val Epoch: 250\tTime: 0.159\tData Time: 0.150\tLoss: 0.522\tAvg Loss: 0.522\n",
      "Epoch: [251/619]\tBatch: [0/47]\tTime: 0.198\tData Time: 0.148\tLoss: 0.713\tAvg Loss: 0.713\t\n",
      "Epoch: [252/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 0.617\tAvg Loss: 0.617\t\n",
      "Epoch: [253/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.146\tLoss: 0.676\tAvg Loss: 0.676\t\n",
      "Epoch: [254/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 7.773\tAvg Loss: 7.773\t\n",
      "Epoch: [255/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.148\tLoss: 0.539\tAvg Loss: 0.539\t\n",
      "Epoch: [256/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.150\tLoss: 0.633\tAvg Loss: 0.633\t\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: [257/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 0.546\tAvg Loss: 0.546\t\n",
      "Epoch: [258/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.143\tLoss: 0.620\tAvg Loss: 0.620\t\n",
      "Epoch: [259/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 0.552\tAvg Loss: 0.552\t\n",
      "Epoch: [260/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.149\tLoss: 0.464\tAvg Loss: 0.464\t\n",
      "Val Epoch: 260\tTime: 0.168\tData Time: 0.150\tLoss: 0.359\tAvg Loss: 0.359\n",
      "Epoch: [261/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 0.544\tAvg Loss: 0.544\t\n",
      "Epoch: [262/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.143\tLoss: 0.609\tAvg Loss: 0.609\t\n",
      "Epoch: [263/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.141\tLoss: 0.576\tAvg Loss: 0.576\t\n",
      "Epoch: [264/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 0.402\tAvg Loss: 0.402\t\n",
      "Epoch: [265/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.143\tLoss: 0.398\tAvg Loss: 0.398\t\n",
      "Epoch: [266/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.146\tLoss: 0.458\tAvg Loss: 0.458\t\n",
      "Epoch: [267/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.149\tLoss: 0.369\tAvg Loss: 0.369\t\n",
      "Epoch: [268/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.144\tLoss: 0.465\tAvg Loss: 0.465\t\n",
      "Epoch: [269/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.145\tLoss: 0.306\tAvg Loss: 0.306\t\n",
      "Epoch: [270/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 0.304\tAvg Loss: 0.304\t\n",
      "Val Epoch: 270\tTime: 0.159\tData Time: 0.150\tLoss: 0.302\tAvg Loss: 0.302\n",
      "Epoch: [271/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.143\tLoss: 0.386\tAvg Loss: 0.386\t\n",
      "Epoch: [272/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.144\tLoss: 0.330\tAvg Loss: 0.330\t\n",
      "Epoch: [273/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 0.377\tAvg Loss: 0.377\t\n",
      "Epoch: [274/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 0.239\tAvg Loss: 0.239\t\n",
      "Epoch: [275/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.152\tLoss: 0.270\tAvg Loss: 0.270\t\n",
      "Epoch: [276/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.146\tLoss: 0.290\tAvg Loss: 0.290\t\n",
      "Epoch: [277/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 0.289\tAvg Loss: 0.289\t\n",
      "Epoch: [278/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.144\tLoss: 0.214\tAvg Loss: 0.214\t\n",
      "Epoch: [279/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 0.174\tAvg Loss: 0.174\t\n",
      "Epoch: [280/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.149\tLoss: 330.202\tAvg Loss: 330.202\t\n",
      "Val Epoch: 280\tTime: 0.159\tData Time: 0.149\tLoss: 5274.771\tAvg Loss: 5274.771\n",
      "Epoch: [281/619]\tBatch: [0/47]\tTime: 0.193\tData Time: 0.149\tLoss: 4690.003\tAvg Loss: 4690.003\t\n",
      "Epoch: [282/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.148\tLoss: 0.452\tAvg Loss: 0.452\t\n",
      "Epoch: [283/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.145\tLoss: 0.454\tAvg Loss: 0.454\t\n",
      "Epoch: [284/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.149\tLoss: 7489.208\tAvg Loss: 7489.208\t\n",
      "Epoch: [285/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.148\tLoss: 0.453\tAvg Loss: 0.453\t\n",
      "Epoch: [286/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 0.481\tAvg Loss: 0.481\t\n",
      "Epoch: [287/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.145\tLoss: 0.538\tAvg Loss: 0.538\t\n",
      "Epoch: [288/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.146\tLoss: 0.544\tAvg Loss: 0.544\t\n",
      "Epoch: [289/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.147\tLoss: 0.502\tAvg Loss: 0.502\t\n",
      "Epoch: [290/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 0.511\tAvg Loss: 0.511\t\n",
      "Val Epoch: 290\tTime: 0.157\tData Time: 0.148\tLoss: 0.454\tAvg Loss: 0.454\n",
      "Epoch: [291/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.148\tLoss: 0.546\tAvg Loss: 0.546\t\n",
      "Epoch: [292/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 0.501\tAvg Loss: 0.501\t\n",
      "Epoch: [293/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 0.525\tAvg Loss: 0.525\t\n",
      "Epoch: [294/619]\tBatch: [0/47]\tTime: 0.207\tData Time: 0.152\tLoss: 26.986\tAvg Loss: 26.986\t\n",
      "Epoch: [295/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.149\tLoss: 0.543\tAvg Loss: 0.543\t\n",
      "Epoch: [296/619]\tBatch: [0/47]\tTime: 0.203\tData Time: 0.148\tLoss: 0.554\tAvg Loss: 0.554\t\n",
      "Epoch: [297/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.143\tLoss: 0.522\tAvg Loss: 0.522\t\n",
      "Epoch: [298/619]\tBatch: [0/47]\tTime: 0.202\tData Time: 0.149\tLoss: 0.567\tAvg Loss: 0.567\t\n",
      "Epoch: [299/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 0.522\tAvg Loss: 0.522\t\n",
      "Epoch: [300/619]\tBatch: [0/47]\tTime: 0.209\tData Time: 0.147\tLoss: 0.539\tAvg Loss: 0.539\t\n",
      "Val Epoch: 300\tTime: 0.163\tData Time: 0.154\tLoss: 0.487\tAvg Loss: 0.487\n",
      "Epoch: [301/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 0.560\tAvg Loss: 0.560\t\n",
      "Epoch: [302/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.145\tLoss: 0.607\tAvg Loss: 0.607\t\n",
      "Epoch: [303/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 0.569\tAvg Loss: 0.569\t\n",
      "Epoch: [304/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 0.538\tAvg Loss: 0.538\t\n",
      "Epoch: [305/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.147\tLoss: 0.576\tAvg Loss: 0.576\t\n",
      "Epoch: [306/619]\tBatch: [0/47]\tTime: 0.202\tData Time: 0.146\tLoss: 0.593\tAvg Loss: 0.593\t\n",
      "Epoch: [307/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 0.571\tAvg Loss: 0.571\t\n",
      "Epoch: [308/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 0.533\tAvg Loss: 0.533\t\n",
      "Epoch: [309/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 0.557\tAvg Loss: 0.557\t\n",
      "Epoch: [310/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 0.587\tAvg Loss: 0.587\t\n",
      "Val Epoch: 310\tTime: 0.160\tData Time: 0.150\tLoss: 0.464\tAvg Loss: 0.464\n",
      "Epoch: [311/619]\tBatch: [0/47]\tTime: 0.199\tData Time: 0.150\tLoss: 0.516\tAvg Loss: 0.516\t\n",
      "Epoch: [312/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.142\tLoss: 0.586\tAvg Loss: 0.586\t\n",
      "Epoch: [313/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.150\tLoss: 0.551\tAvg Loss: 0.551\t\n",
      "Epoch: [314/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 0.555\tAvg Loss: 0.555\t\n",
      "Epoch: [315/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.149\tLoss: 0.530\tAvg Loss: 0.530\t\n",
      "Epoch: [316/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.148\tLoss: 0.528\tAvg Loss: 0.528\t\n",
      "Epoch: [317/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.148\tLoss: 0.588\tAvg Loss: 0.588\t\n",
      "Epoch: [318/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.147\tLoss: 0.595\tAvg Loss: 0.595\t\n",
      "Epoch: [319/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.147\tLoss: 0.562\tAvg Loss: 0.562\t\n",
      "Epoch: [320/619]\tBatch: [0/47]\tTime: 0.203\tData Time: 0.147\tLoss: 0.566\tAvg Loss: 0.566\t\n",
      "Val Epoch: 320\tTime: 0.159\tData Time: 0.150\tLoss: 0.464\tAvg Loss: 0.464\n",
      "Epoch: [321/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.143\tLoss: 0.568\tAvg Loss: 0.568\t\n",
      "Epoch: [322/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.149\tLoss: 5.538\tAvg Loss: 5.538\t\n",
      "Epoch: [323/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.150\tLoss: 0.589\tAvg Loss: 0.589\t\n",
      "Epoch: [324/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.149\tLoss: 0.520\tAvg Loss: 0.520\t\n",
      "Epoch: [325/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 0.524\tAvg Loss: 0.524\t\n",
      "Epoch: [326/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 0.579\tAvg Loss: 0.579\t\n",
      "Epoch: [327/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.145\tLoss: 0.617\tAvg Loss: 0.617\t\n",
      "Epoch: [328/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.144\tLoss: 1.197\tAvg Loss: 1.197\t\n",
      "Epoch: [329/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 0.552\tAvg Loss: 0.552\t\n",
      "Epoch: [330/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.149\tLoss: 0.689\tAvg Loss: 0.689\t\n",
      "Val Epoch: 330\tTime: 0.157\tData Time: 0.147\tLoss: 0.694\tAvg Loss: 0.694\n",
      "Epoch: [331/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.150\tLoss: 0.801\tAvg Loss: 0.801\t\n",
      "Epoch: [332/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.145\tLoss: 0.960\tAvg Loss: 0.960\t\n",
      "Epoch: [333/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.147\tLoss: 0.887\tAvg Loss: 0.887\t\n",
      "Epoch: [334/619]\tBatch: [0/47]\tTime: 0.204\tData Time: 0.149\tLoss: 0.972\tAvg Loss: 0.972\t\n",
      "Epoch: [335/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 1.028\tAvg Loss: 1.028\t\n",
      "Epoch: [336/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.150\tLoss: 13.734\tAvg Loss: 13.734\t\n",
      "Epoch: [337/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.144\tLoss: 236.660\tAvg Loss: 236.660\t\n",
      "Epoch: [338/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.242\tAvg Loss: 1.242\t\n",
      "Epoch: [339/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.146\tLoss: 1.385\tAvg Loss: 1.385\t\n",
      "Epoch: [340/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.148\tLoss: 1.424\tAvg Loss: 1.424\t\n",
      "Val Epoch: 340\tTime: 0.158\tData Time: 0.148\tLoss: 1.290\tAvg Loss: 1.290\n",
      "Epoch: [341/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.144\tLoss: 1.435\tAvg Loss: 1.435\t\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: [342/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.149\tLoss: 1.369\tAvg Loss: 1.369\t\n",
      "Epoch: [343/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.144\tLoss: 1.500\tAvg Loss: 1.500\t\n",
      "Epoch: [344/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 1.479\tAvg Loss: 1.479\t\n",
      "Epoch: [345/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.151\tLoss: 1.532\tAvg Loss: 1.532\t\n",
      "Epoch: [346/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.144\tLoss: 1.610\tAvg Loss: 1.610\t\n",
      "Epoch: [347/619]\tBatch: [0/47]\tTime: 0.205\tData Time: 0.156\tLoss: 1.514\tAvg Loss: 1.514\t\n",
      "Epoch: [348/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 1.595\tAvg Loss: 1.595\t\n",
      "Epoch: [349/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.147\tLoss: 1.862\tAvg Loss: 1.862\t\n",
      "Epoch: [350/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.143\tLoss: 1.796\tAvg Loss: 1.796\t\n",
      "Val Epoch: 350\tTime: 0.161\tData Time: 0.152\tLoss: 1.695\tAvg Loss: 1.695\n",
      "Epoch: [351/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 1.916\tAvg Loss: 1.916\t\n",
      "Epoch: [352/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.146\tLoss: 1.761\tAvg Loss: 1.761\t\n",
      "Epoch: [353/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.150\tLoss: 139.589\tAvg Loss: 139.589\t\n",
      "Epoch: [354/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.144\tLoss: 1.758\tAvg Loss: 1.758\t\n",
      "Epoch: [355/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 1.760\tAvg Loss: 1.760\t\n",
      "Epoch: [356/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.146\tLoss: 1.689\tAvg Loss: 1.689\t\n",
      "Epoch: [357/619]\tBatch: [0/47]\tTime: 0.193\tData Time: 0.147\tLoss: 837.906\tAvg Loss: 837.906\t\n",
      "Epoch: [358/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.824\tAvg Loss: 1.824\t\n",
      "Epoch: [359/619]\tBatch: [0/47]\tTime: 0.198\tData Time: 0.143\tLoss: 201.126\tAvg Loss: 201.126\t\n",
      "Epoch: [360/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.143\tLoss: 1.844\tAvg Loss: 1.844\t\n",
      "Val Epoch: 360\tTime: 0.168\tData Time: 0.149\tLoss: 1.695\tAvg Loss: 1.695\n",
      "Epoch: [361/619]\tBatch: [0/47]\tTime: 0.194\tData Time: 0.145\tLoss: 1.855\tAvg Loss: 1.855\t\n",
      "Epoch: [362/619]\tBatch: [0/47]\tTime: 0.179\tData Time: 0.140\tLoss: 1.783\tAvg Loss: 1.783\t\n",
      "Epoch: [363/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 1.731\tAvg Loss: 1.731\t\n",
      "Epoch: [364/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.145\tLoss: 1.743\tAvg Loss: 1.743\t\n",
      "Epoch: [365/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.145\tLoss: 1.732\tAvg Loss: 1.732\t\n",
      "Epoch: [366/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 1.699\tAvg Loss: 1.699\t\n",
      "Epoch: [367/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 1.651\tAvg Loss: 1.651\t\n",
      "Epoch: [368/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.710\tAvg Loss: 1.710\t\n",
      "Epoch: [369/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.622\tAvg Loss: 1.622\t\n",
      "Epoch: [370/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.150\tLoss: 1.576\tAvg Loss: 1.576\t\n",
      "Val Epoch: 370\tTime: 0.157\tData Time: 0.148\tLoss: 1.502\tAvg Loss: 1.502\n",
      "Epoch: [371/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.141\tLoss: 1.619\tAvg Loss: 1.619\t\n",
      "Epoch: [372/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.690\tAvg Loss: 1.690\t\n",
      "Epoch: [373/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 1.536\tAvg Loss: 1.536\t\n",
      "Epoch: [374/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.529\tAvg Loss: 1.529\t\n",
      "Epoch: [375/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.562\tAvg Loss: 1.562\t\n",
      "Epoch: [376/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.147\tLoss: 1.562\tAvg Loss: 1.562\t\n",
      "Epoch: [377/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.143\tLoss: 1.543\tAvg Loss: 1.543\t\n",
      "Epoch: [378/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.148\tLoss: 1.363\tAvg Loss: 1.363\t\n",
      "Epoch: [379/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.148\tLoss: 1.455\tAvg Loss: 1.455\t\n",
      "Epoch: [380/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.146\tLoss: 1.425\tAvg Loss: 1.425\t\n",
      "Val Epoch: 380\tTime: 0.158\tData Time: 0.149\tLoss: 1.327\tAvg Loss: 1.327\n",
      "Epoch: [381/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.143\tLoss: 1.375\tAvg Loss: 1.375\t\n",
      "Epoch: [382/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.149\tLoss: 1.362\tAvg Loss: 1.362\t\n",
      "Epoch: [383/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 1.290\tAvg Loss: 1.290\t\n",
      "Epoch: [384/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.299\tAvg Loss: 1.299\t\n",
      "Epoch: [385/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.204\tAvg Loss: 1.204\t\n",
      "Epoch: [386/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.146\tLoss: 1.222\tAvg Loss: 1.222\t\n",
      "Epoch: [387/619]\tBatch: [0/47]\tTime: 0.178\tData Time: 0.143\tLoss: 53.003\tAvg Loss: 53.003\t\n",
      "Epoch: [388/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.150\tAvg Loss: 1.150\t\n",
      "Epoch: [389/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.150\tLoss: 1.141\tAvg Loss: 1.141\t\n",
      "Epoch: [390/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.146\tLoss: 1.074\tAvg Loss: 1.074\t\n",
      "Val Epoch: 390\tTime: 0.159\tData Time: 0.150\tLoss: 0.987\tAvg Loss: 0.987\n",
      "Epoch: [391/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.056\tAvg Loss: 1.056\t\n",
      "Epoch: [392/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.144\tLoss: 1.095\tAvg Loss: 1.095\t\n",
      "Epoch: [393/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 0.967\tAvg Loss: 0.967\t\n",
      "Epoch: [394/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.145\tLoss: 1.007\tAvg Loss: 1.007\t\n",
      "Epoch: [395/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 0.921\tAvg Loss: 0.921\t\n",
      "Epoch: [396/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 0.932\tAvg Loss: 0.932\t\n",
      "Epoch: [397/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 0.874\tAvg Loss: 0.874\t\n",
      "Epoch: [398/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.144\tLoss: 0.787\tAvg Loss: 0.787\t\n",
      "Epoch: [399/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.147\tLoss: 269.693\tAvg Loss: 269.693\t\n",
      "Epoch: [400/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.145\tLoss: 0.769\tAvg Loss: 0.769\t\n",
      "Val Epoch: 400\tTime: 0.159\tData Time: 0.149\tLoss: 0.709\tAvg Loss: 0.709\n",
      "Epoch: [401/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 0.755\tAvg Loss: 0.755\t\n",
      "Epoch: [402/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 0.777\tAvg Loss: 0.777\t\n",
      "Epoch: [403/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.143\tLoss: 0.795\tAvg Loss: 0.795\t\n",
      "Epoch: [404/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 0.777\tAvg Loss: 0.777\t\n",
      "Epoch: [405/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 0.792\tAvg Loss: 0.792\t\n",
      "Epoch: [406/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.147\tLoss: 1.862\tAvg Loss: 1.862\t\n",
      "Epoch: [407/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.148\tLoss: 0.847\tAvg Loss: 0.847\t\n",
      "Epoch: [408/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.145\tLoss: 429.385\tAvg Loss: 429.385\t\n",
      "Epoch: [409/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 1.053\tAvg Loss: 1.053\t\n",
      "Epoch: [410/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.145\tLoss: 1.086\tAvg Loss: 1.086\t\n",
      "Val Epoch: 410\tTime: 0.159\tData Time: 0.150\tLoss: 1.092\tAvg Loss: 1.092\n",
      "Epoch: [411/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 1.115\tAvg Loss: 1.115\t\n",
      "Epoch: [412/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.146\tLoss: 1.138\tAvg Loss: 1.138\t\n",
      "Epoch: [413/619]\tBatch: [0/47]\tTime: 0.178\tData Time: 0.142\tLoss: 1.120\tAvg Loss: 1.120\t\n",
      "Epoch: [414/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.118\tAvg Loss: 1.118\t\n",
      "Epoch: [415/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 1.114\tAvg Loss: 1.114\t\n",
      "Epoch: [416/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.145\tLoss: 1.088\tAvg Loss: 1.088\t\n",
      "Epoch: [417/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.147\tLoss: 1.123\tAvg Loss: 1.123\t\n",
      "Epoch: [418/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.149\tLoss: 1.093\tAvg Loss: 1.093\t\n",
      "Epoch: [419/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.147\tLoss: 1.190\tAvg Loss: 1.190\t\n",
      "Epoch: [420/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.149\tLoss: 1.115\tAvg Loss: 1.115\t\n",
      "Val Epoch: 420\tTime: 0.158\tData Time: 0.149\tLoss: 1.053\tAvg Loss: 1.053\n",
      "Epoch: [421/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 1.074\tAvg Loss: 1.074\t\n",
      "Epoch: [422/619]\tBatch: [0/47]\tTime: 0.179\tData Time: 0.144\tLoss: 1.075\tAvg Loss: 1.075\t\n",
      "Epoch: [423/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.144\tLoss: 1.108\tAvg Loss: 1.108\t\n",
      "Epoch: [424/619]\tBatch: [0/47]\tTime: 0.178\tData Time: 0.144\tLoss: 1.115\tAvg Loss: 1.115\t\n",
      "Epoch: [425/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.143\tLoss: 1.069\tAvg Loss: 1.069\t\n",
      "Epoch: [426/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.149\tLoss: 1.094\tAvg Loss: 1.094\t\n",
      "Epoch: [427/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.148\tLoss: 1.078\tAvg Loss: 1.078\t\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: [428/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.145\tLoss: 1.109\tAvg Loss: 1.109\t\n",
      "Epoch: [429/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.124\tAvg Loss: 1.124\t\n",
      "Epoch: [430/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.144\tLoss: 1.098\tAvg Loss: 1.098\t\n",
      "Val Epoch: 430\tTime: 0.158\tData Time: 0.149\tLoss: 1.074\tAvg Loss: 1.074\n",
      "Epoch: [431/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.147\tLoss: 1.100\tAvg Loss: 1.100\t\n",
      "Epoch: [432/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.145\tLoss: 1.094\tAvg Loss: 1.094\t\n",
      "Epoch: [433/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.145\tLoss: 1.098\tAvg Loss: 1.098\t\n",
      "Epoch: [434/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 3.191\tAvg Loss: 3.191\t\n",
      "Epoch: [435/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.144\tLoss: 1.069\tAvg Loss: 1.069\t\n",
      "Epoch: [436/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.144\tLoss: 1.105\tAvg Loss: 1.105\t\n",
      "Epoch: [437/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.145\tLoss: 1.103\tAvg Loss: 1.103\t\n",
      "Epoch: [438/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.080\tAvg Loss: 1.080\t\n",
      "Epoch: [439/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.091\tAvg Loss: 1.091\t\n",
      "Epoch: [440/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.144\tLoss: 1.088\tAvg Loss: 1.088\t\n",
      "Val Epoch: 440\tTime: 0.155\tData Time: 0.146\tLoss: 1.055\tAvg Loss: 1.055\n",
      "Epoch: [441/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 1.127\tAvg Loss: 1.127\t\n",
      "Epoch: [442/619]\tBatch: [0/47]\tTime: 0.177\tData Time: 0.143\tLoss: 1.078\tAvg Loss: 1.078\t\n",
      "Epoch: [443/619]\tBatch: [0/47]\tTime: 0.179\tData Time: 0.142\tLoss: 1.094\tAvg Loss: 1.094\t\n",
      "Epoch: [444/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.144\tLoss: 1.065\tAvg Loss: 1.065\t\n",
      "Epoch: [445/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.144\tLoss: 1.088\tAvg Loss: 1.088\t\n",
      "Epoch: [446/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 1.078\tAvg Loss: 1.078\t\n",
      "Epoch: [447/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.040\tAvg Loss: 1.040\t\n",
      "Epoch: [448/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.146\tLoss: 1.042\tAvg Loss: 1.042\t\n",
      "Epoch: [449/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.142\tLoss: 1.037\tAvg Loss: 1.037\t\n",
      "Epoch: [450/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.038\tAvg Loss: 1.038\t\n",
      "Val Epoch: 450\tTime: 0.154\tData Time: 0.145\tLoss: 1.014\tAvg Loss: 1.014\n",
      "Epoch: [451/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.143\tLoss: 1.048\tAvg Loss: 1.048\t\n",
      "Epoch: [452/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.148\tLoss: 1.030\tAvg Loss: 1.030\t\n",
      "Epoch: [453/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.022\tAvg Loss: 1.022\t\n",
      "Epoch: [454/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.046\tAvg Loss: 1.046\t\n",
      "Epoch: [455/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.054\tAvg Loss: 1.054\t\n",
      "Epoch: [456/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.142\tLoss: 1.014\tAvg Loss: 1.014\t\n",
      "Epoch: [457/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.148\tLoss: 1.022\tAvg Loss: 1.022\t\n",
      "Epoch: [458/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.146\tLoss: 1.025\tAvg Loss: 1.025\t\n",
      "Epoch: [459/619]\tBatch: [0/47]\tTime: 0.179\tData Time: 0.146\tLoss: 1.016\tAvg Loss: 1.016\t\n",
      "Epoch: [460/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.144\tLoss: 1.018\tAvg Loss: 1.018\t\n",
      "Val Epoch: 460\tTime: 0.156\tData Time: 0.147\tLoss: 1.009\tAvg Loss: 1.009\n",
      "Epoch: [461/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.143\tLoss: 1.052\tAvg Loss: 1.052\t\n",
      "Epoch: [462/619]\tBatch: [0/47]\tTime: 0.178\tData Time: 0.143\tLoss: 1.015\tAvg Loss: 1.015\t\n",
      "Epoch: [463/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.021\tAvg Loss: 1.021\t\n",
      "Epoch: [464/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.011\tAvg Loss: 1.011\t\n",
      "Epoch: [465/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.149\tLoss: 1.010\tAvg Loss: 1.010\t\n",
      "Epoch: [466/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.146\tLoss: 1.000\tAvg Loss: 1.000\t\n",
      "Epoch: [467/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.073\tAvg Loss: 1.073\t\n",
      "Epoch: [468/619]\tBatch: [0/47]\tTime: 0.179\tData Time: 0.143\tLoss: 1.031\tAvg Loss: 1.031\t\n",
      "Epoch: [469/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.146\tLoss: 1.040\tAvg Loss: 1.040\t\n",
      "Epoch: [470/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.013\tAvg Loss: 1.013\t\n",
      "Val Epoch: 470\tTime: 0.158\tData Time: 0.149\tLoss: 1.006\tAvg Loss: 1.006\n",
      "Epoch: [471/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 1.036\tAvg Loss: 1.036\t\n",
      "Epoch: [472/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.028\tAvg Loss: 1.028\t\n",
      "Epoch: [473/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.144\tLoss: 1.049\tAvg Loss: 1.049\t\n",
      "Epoch: [474/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.149\tLoss: 1.060\tAvg Loss: 1.060\t\n",
      "Epoch: [475/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 1.053\tAvg Loss: 1.053\t\n",
      "Epoch: [476/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 1.035\tAvg Loss: 1.035\t\n",
      "Epoch: [477/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.144\tLoss: 1.772\tAvg Loss: 1.772\t\n",
      "Epoch: [478/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.148\tLoss: 1.045\tAvg Loss: 1.045\t\n",
      "Epoch: [479/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.147\tLoss: 1.081\tAvg Loss: 1.081\t\n",
      "Epoch: [480/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.144\tLoss: 1.106\tAvg Loss: 1.106\t\n",
      "Val Epoch: 480\tTime: 0.157\tData Time: 0.148\tLoss: 1.034\tAvg Loss: 1.034\n",
      "Epoch: [481/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.144\tLoss: 1.076\tAvg Loss: 1.076\t\n",
      "Epoch: [482/619]\tBatch: [0/47]\tTime: 0.179\tData Time: 0.141\tLoss: 1.064\tAvg Loss: 1.064\t\n",
      "Epoch: [483/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 1.045\tAvg Loss: 1.045\t\n",
      "Epoch: [484/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.146\tLoss: 1.026\tAvg Loss: 1.026\t\n",
      "Epoch: [485/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.051\tAvg Loss: 1.051\t\n",
      "Epoch: [486/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.022\tAvg Loss: 1.022\t\n",
      "Epoch: [487/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.142\tLoss: 58.126\tAvg Loss: 58.126\t\n",
      "Epoch: [488/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.142\tLoss: 1.043\tAvg Loss: 1.043\t\n",
      "Epoch: [489/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.146\tLoss: 1.036\tAvg Loss: 1.036\t\n",
      "Epoch: [490/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.142\tLoss: 1.037\tAvg Loss: 1.037\t\n",
      "Val Epoch: 490\tTime: 0.158\tData Time: 0.148\tLoss: 0.993\tAvg Loss: 0.993\n",
      "Epoch: [491/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 1.007\tAvg Loss: 1.007\t\n",
      "Epoch: [492/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.144\tLoss: 1.011\tAvg Loss: 1.011\t\n",
      "Epoch: [493/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.008\tAvg Loss: 1.008\t\n",
      "Epoch: [494/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.150\tLoss: 1.001\tAvg Loss: 1.001\t\n",
      "Epoch: [495/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.044\tAvg Loss: 1.044\t\n",
      "Epoch: [496/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.045\tAvg Loss: 1.045\t\n",
      "Epoch: [497/619]\tBatch: [0/47]\tTime: 0.179\tData Time: 0.141\tLoss: 1.031\tAvg Loss: 1.031\t\n",
      "Epoch: [498/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.147\tLoss: 1.008\tAvg Loss: 1.008\t\n",
      "Epoch: [499/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.150\tLoss: 1.020\tAvg Loss: 1.020\t\n",
      "Epoch: [500/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.144\tLoss: 1.014\tAvg Loss: 1.014\t\n",
      "Val Epoch: 500\tTime: 0.157\tData Time: 0.148\tLoss: 0.994\tAvg Loss: 0.994\n",
      "Epoch: [501/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.016\tAvg Loss: 1.016\t\n",
      "Epoch: [502/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.147\tLoss: 1.033\tAvg Loss: 1.033\t\n",
      "Epoch: [503/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.042\tAvg Loss: 1.042\t\n",
      "Epoch: [504/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.017\tAvg Loss: 1.017\t\n",
      "Epoch: [505/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.144\tLoss: 1.038\tAvg Loss: 1.038\t\n",
      "Epoch: [506/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.142\tLoss: 1.023\tAvg Loss: 1.023\t\n",
      "Epoch: [507/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.021\tAvg Loss: 1.021\t\n",
      "Epoch: [508/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.143\tLoss: 1.008\tAvg Loss: 1.008\t\n",
      "Epoch: [509/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.034\tAvg Loss: 1.034\t\n",
      "Epoch: [510/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.145\tLoss: 1.028\tAvg Loss: 1.028\t\n",
      "Val Epoch: 510\tTime: 0.160\tData Time: 0.151\tLoss: 1.003\tAvg Loss: 1.003\n",
      "Epoch: [511/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.142\tLoss: 1.043\tAvg Loss: 1.043\t\n",
      "Epoch: [512/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.041\tAvg Loss: 1.041\t\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: [513/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.145\tLoss: 1.034\tAvg Loss: 1.034\t\n",
      "Epoch: [514/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 1.016\tAvg Loss: 1.016\t\n",
      "Epoch: [515/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.031\tAvg Loss: 1.031\t\n",
      "Epoch: [516/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.011\tAvg Loss: 1.011\t\n",
      "Epoch: [517/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.149\tLoss: 1.026\tAvg Loss: 1.026\t\n",
      "Epoch: [518/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.146\tLoss: 1.011\tAvg Loss: 1.011\t\n",
      "Epoch: [519/619]\tBatch: [0/47]\tTime: 0.189\tData Time: 0.149\tLoss: 1.009\tAvg Loss: 1.009\t\n",
      "Epoch: [520/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 1.005\tAvg Loss: 1.005\t\n",
      "Val Epoch: 520\tTime: 0.158\tData Time: 0.149\tLoss: 0.993\tAvg Loss: 0.993\n",
      "Epoch: [521/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.146\tLoss: 1.022\tAvg Loss: 1.022\t\n",
      "Epoch: [522/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.020\tAvg Loss: 1.020\t\n",
      "Epoch: [523/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 1.017\tAvg Loss: 1.017\t\n",
      "Epoch: [524/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.147\tLoss: 0.992\tAvg Loss: 0.992\t\n",
      "Epoch: [525/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.150\tLoss: 1.031\tAvg Loss: 1.031\t\n",
      "Epoch: [526/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.464\tAvg Loss: 1.464\t\n",
      "Epoch: [527/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.040\tAvg Loss: 1.040\t\n",
      "Epoch: [528/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.147\tLoss: 1.036\tAvg Loss: 1.036\t\n",
      "Epoch: [529/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.146\tLoss: 1.044\tAvg Loss: 1.044\t\n",
      "Epoch: [530/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.145\tLoss: 1.083\tAvg Loss: 1.083\t\n",
      "Val Epoch: 530\tTime: 0.156\tData Time: 0.147\tLoss: 1.051\tAvg Loss: 1.051\n",
      "Epoch: [531/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.144\tLoss: 1.081\tAvg Loss: 1.081\t\n",
      "Epoch: [532/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.144\tLoss: 1.101\tAvg Loss: 1.101\t\n",
      "Epoch: [533/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.084\tAvg Loss: 1.084\t\n",
      "Epoch: [534/619]\tBatch: [0/47]\tTime: 0.178\tData Time: 0.142\tLoss: 1.081\tAvg Loss: 1.081\t\n",
      "Epoch: [535/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.143\tLoss: 1.074\tAvg Loss: 1.074\t\n",
      "Epoch: [536/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 1.078\tAvg Loss: 1.078\t\n",
      "Epoch: [537/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.116\tAvg Loss: 1.116\t\n",
      "Epoch: [538/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.145\tLoss: 1.093\tAvg Loss: 1.093\t\n",
      "Epoch: [539/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.142\tLoss: 1.094\tAvg Loss: 1.094\t\n",
      "Epoch: [540/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.145\tLoss: 1.108\tAvg Loss: 1.108\t\n",
      "Val Epoch: 540\tTime: 0.157\tData Time: 0.148\tLoss: 1.039\tAvg Loss: 1.039\n",
      "Epoch: [541/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.062\tAvg Loss: 1.062\t\n",
      "Epoch: [542/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.146\tLoss: 1.106\tAvg Loss: 1.106\t\n",
      "Epoch: [543/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.144\tLoss: 8.517\tAvg Loss: 8.517\t\n",
      "Epoch: [544/619]\tBatch: [0/47]\tTime: 0.177\tData Time: 0.142\tLoss: 36.354\tAvg Loss: 36.354\t\n",
      "Epoch: [545/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 1.109\tAvg Loss: 1.109\t\n",
      "Epoch: [546/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.149\tLoss: 1.079\tAvg Loss: 1.079\t\n",
      "Epoch: [547/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.084\tAvg Loss: 1.084\t\n",
      "Epoch: [548/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.142\tLoss: 1.095\tAvg Loss: 1.095\t\n",
      "Epoch: [549/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.145\tLoss: 3.800\tAvg Loss: 3.800\t\n",
      "Epoch: [550/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.146\tLoss: 1.082\tAvg Loss: 1.082\t\n",
      "Val Epoch: 550\tTime: 0.154\tData Time: 0.145\tLoss: 1.059\tAvg Loss: 1.059\n",
      "Epoch: [551/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.088\tAvg Loss: 1.088\t\n",
      "Epoch: [552/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.150\tLoss: 5.011\tAvg Loss: 5.011\t\n",
      "Epoch: [553/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.077\tAvg Loss: 1.077\t\n",
      "Epoch: [554/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.146\tLoss: 1.073\tAvg Loss: 1.073\t\n",
      "Epoch: [555/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.147\tLoss: 1.100\tAvg Loss: 1.100\t\n",
      "Epoch: [556/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.144\tLoss: 1.099\tAvg Loss: 1.099\t\n",
      "Epoch: [557/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.146\tLoss: 1.100\tAvg Loss: 1.100\t\n",
      "Epoch: [558/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.146\tLoss: 1.102\tAvg Loss: 1.102\t\n",
      "Epoch: [559/619]\tBatch: [0/47]\tTime: 0.178\tData Time: 0.140\tLoss: 1.120\tAvg Loss: 1.120\t\n",
      "Epoch: [560/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.121\tAvg Loss: 1.121\t\n",
      "Val Epoch: 560\tTime: 0.160\tData Time: 0.151\tLoss: 1.084\tAvg Loss: 1.084\n",
      "Epoch: [561/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.148\tLoss: 1.136\tAvg Loss: 1.136\t\n",
      "Epoch: [562/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.146\tLoss: 1.112\tAvg Loss: 1.112\t\n",
      "Epoch: [563/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 1.104\tAvg Loss: 1.104\t\n",
      "Epoch: [564/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 1.124\tAvg Loss: 1.124\t\n",
      "Epoch: [565/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.146\tLoss: 1.125\tAvg Loss: 1.125\t\n",
      "Epoch: [566/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.149\tLoss: 1.137\tAvg Loss: 1.137\t\n",
      "Epoch: [567/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.134\tAvg Loss: 1.134\t\n",
      "Epoch: [568/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.144\tLoss: 1.115\tAvg Loss: 1.115\t\n",
      "Epoch: [569/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.145\tLoss: 1.123\tAvg Loss: 1.123\t\n",
      "Epoch: [570/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.134\tAvg Loss: 1.134\t\n",
      "Val Epoch: 570\tTime: 0.158\tData Time: 0.149\tLoss: 1.100\tAvg Loss: 1.100\n",
      "Epoch: [571/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.136\tAvg Loss: 1.136\t\n",
      "Epoch: [572/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.147\tLoss: 1.104\tAvg Loss: 1.104\t\n",
      "Epoch: [573/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.145\tLoss: 1.117\tAvg Loss: 1.117\t\n",
      "Epoch: [574/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.143\tLoss: 1.120\tAvg Loss: 1.120\t\n",
      "Epoch: [575/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.105\tAvg Loss: 1.105\t\n",
      "Epoch: [576/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.147\tLoss: 1.095\tAvg Loss: 1.095\t\n",
      "Epoch: [577/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.104\tAvg Loss: 1.104\t\n",
      "Epoch: [578/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.112\tAvg Loss: 1.112\t\n",
      "Epoch: [579/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.149\tLoss: 1.121\tAvg Loss: 1.121\t\n",
      "Epoch: [580/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.144\tLoss: 1.087\tAvg Loss: 1.087\t\n",
      "Val Epoch: 580\tTime: 0.160\tData Time: 0.151\tLoss: 1.078\tAvg Loss: 1.078\n",
      "Epoch: [581/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.144\tLoss: 1.096\tAvg Loss: 1.096\t\n",
      "Epoch: [582/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.148\tLoss: 1.100\tAvg Loss: 1.100\t\n",
      "Epoch: [583/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.144\tLoss: 1.132\tAvg Loss: 1.132\t\n",
      "Epoch: [584/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.144\tLoss: 1.126\tAvg Loss: 1.126\t\n",
      "Epoch: [585/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.099\tAvg Loss: 1.099\t\n",
      "Epoch: [586/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.146\tLoss: 1.176\tAvg Loss: 1.176\t\n",
      "Epoch: [587/619]\tBatch: [0/47]\tTime: 0.180\tData Time: 0.141\tLoss: 1.199\tAvg Loss: 1.199\t\n",
      "Epoch: [588/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.145\tLoss: 1.171\tAvg Loss: 1.171\t\n",
      "Epoch: [589/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 1.165\tAvg Loss: 1.165\t\n",
      "Epoch: [590/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.171\tAvg Loss: 1.171\t\n",
      "Val Epoch: 590\tTime: 0.158\tData Time: 0.148\tLoss: 1.134\tAvg Loss: 1.134\n",
      "Epoch: [591/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.144\tLoss: 1.165\tAvg Loss: 1.165\t\n",
      "Epoch: [592/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.147\tLoss: 1.185\tAvg Loss: 1.185\t\n",
      "Epoch: [593/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.146\tLoss: 1.201\tAvg Loss: 1.201\t\n",
      "Epoch: [594/619]\tBatch: [0/47]\tTime: 0.195\tData Time: 0.147\tLoss: 1.198\tAvg Loss: 1.198\t\n",
      "Epoch: [595/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.146\tLoss: 1.213\tAvg Loss: 1.213\t\n",
      "Epoch: [596/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.151\tLoss: 1.202\tAvg Loss: 1.202\t\n",
      "Epoch: [597/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.191\tAvg Loss: 1.191\t\n",
      "Epoch: [598/619]\tBatch: [0/47]\tTime: 0.181\tData Time: 0.143\tLoss: 1.188\tAvg Loss: 1.188\t\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: [599/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.152\tLoss: 1.177\tAvg Loss: 1.177\t\n",
      "Epoch: [600/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.148\tLoss: 1.213\tAvg Loss: 1.213\t\n",
      "Val Epoch: 600\tTime: 0.170\tData Time: 0.151\tLoss: 1.255\tAvg Loss: 1.255\n",
      "Epoch: [601/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.145\tLoss: 1.275\tAvg Loss: 1.275\t\n",
      "Epoch: [602/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.146\tLoss: 1.285\tAvg Loss: 1.285\t\n",
      "Epoch: [603/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.284\tAvg Loss: 1.284\t\n",
      "Epoch: [604/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.147\tLoss: 1.271\tAvg Loss: 1.271\t\n",
      "Epoch: [605/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 1.244\tAvg Loss: 1.244\t\n",
      "Epoch: [606/619]\tBatch: [0/47]\tTime: 0.196\tData Time: 0.147\tLoss: 1.267\tAvg Loss: 1.267\t\n",
      "Epoch: [607/619]\tBatch: [0/47]\tTime: 0.183\tData Time: 0.145\tLoss: 1.270\tAvg Loss: 1.270\t\n",
      "Epoch: [608/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.152\tLoss: 1.276\tAvg Loss: 1.276\t\n",
      "Epoch: [609/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.153\tLoss: 1.249\tAvg Loss: 1.249\t\n",
      "Epoch: [610/619]\tBatch: [0/47]\tTime: 0.200\tData Time: 0.150\tLoss: 1.263\tAvg Loss: 1.263\t\n",
      "Val Epoch: 610\tTime: 0.161\tData Time: 0.151\tLoss: 1.227\tAvg Loss: 1.227\n",
      "Epoch: [611/619]\tBatch: [0/47]\tTime: 0.190\tData Time: 0.151\tLoss: 1.245\tAvg Loss: 1.245\t\n",
      "Epoch: [612/619]\tBatch: [0/47]\tTime: 0.188\tData Time: 0.144\tLoss: 1.250\tAvg Loss: 1.250\t\n",
      "Epoch: [613/619]\tBatch: [0/47]\tTime: 0.186\tData Time: 0.146\tLoss: 1.249\tAvg Loss: 1.249\t\n",
      "Epoch: [614/619]\tBatch: [0/47]\tTime: 0.191\tData Time: 0.148\tLoss: 1.240\tAvg Loss: 1.240\t\n",
      "Epoch: [615/619]\tBatch: [0/47]\tTime: 0.192\tData Time: 0.148\tLoss: 14.347\tAvg Loss: 14.347\t\n",
      "Epoch: [616/619]\tBatch: [0/47]\tTime: 0.184\tData Time: 0.145\tLoss: 1.263\tAvg Loss: 1.263\t\n",
      "Epoch: [617/619]\tBatch: [0/47]\tTime: 0.185\tData Time: 0.146\tLoss: 1.240\tAvg Loss: 1.240\t\n",
      "Epoch: [618/619]\tBatch: [0/47]\tTime: 0.182\tData Time: 0.144\tLoss: 1.258\tAvg Loss: 1.258\t\n",
      "Epoch: [619/619]\tBatch: [0/47]\tTime: 0.187\tData Time: 0.147\tLoss: 1.233\tAvg Loss: 1.233\t\n",
      "n_epochs = 620\n"
     ]
    }
   ],
   "source": [
    "n_epochs = start_epoch + 620\n",
    "train_loss=[]\n",
    "test_loss=[]\n",
    "print('Training ...')\n",
    "val_freq = 10\n",
    "for e in range(start_epoch, n_epochs):\n",
    "    losses = train(train_loader, model, criterion, optimizer, e, n_epochs, log_freq=100)\n",
    "    train_loss.append(losses)\n",
    "    if e % val_freq == 0:\n",
    "#         end = time.time()\n",
    "        loss = validate(validation_loader, model, criterion, e, log_freq=100)\n",
    "        test_loss.append(loss)\n",
    "\n",
    "\n",
    "start_epoch = n_epochs\n",
    "\n",
    "print('n_epochs = {}'.format(n_epochs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.save(model, 'LSTM_pos_workin2_v2.torch')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "=== Test Training Dataset ======\n",
      "gt_poses = (1536, 7)\n",
      "pred_poses = (1536, 7)\n",
      "Translation(T) error in meters and Rotation(R) error in degrees:\n",
      "T: median = 0.066, mean = 0.072\n",
      "R: median = 4.047, mean = 4.604\n"
     ]
    }
   ],
   "source": [
    "def model_results_pred_gt(model, dataloader):\n",
    "    \n",
    "    model.eval()\n",
    "\n",
    "    gt_poses = np.empty((0, 7))\n",
    "    pred_poses = np.empty((0, 7))\n",
    "\n",
    "    for idx, data in enumerate(dataloader):\n",
    "        batch_images = data[0].cuda()\n",
    "        batch_imu = data[1].cuda()\n",
    "        batch_poses  = data[2].cuda()\n",
    "        batch_images = batch_images.to(device)\n",
    "        batch_imu = batch_imu.to(device)\n",
    "        batch_imu = torch.tensor(batch_imu, dtype=torch.float, device=device)\n",
    "        batch_poses = batch_poses.to(device)\n",
    "        batch_poses = torch.tensor(batch_poses, dtype=torch.float, device=device)\n",
    "        out = model(data[0:2])        \n",
    "        loss = criterion(out, batch_poses)\n",
    "        bp = batch_poses.detach().cpu().numpy()\n",
    "        outp = out.detach().cpu().numpy()\n",
    "        gt_poses = np.vstack((gt_poses, bp))\n",
    "        pred_poses = np.vstack((pred_poses, outp))\n",
    "\n",
    "    gt_poses[:, :3] = gt_poses[:, :3] \n",
    "    pred_poses[:, :3] = pred_poses[:, :3] \n",
    "    \n",
    "    return pred_poses, gt_poses\n",
    "\n",
    "print('\\n=== Test Training Dataset ======')\n",
    "pred_poses, gt_poses = model_results_pred_gt(model, train_loader)\n",
    "\n",
    "print('gt_poses = {}'.format(gt_poses.shape))\n",
    "print('pred_poses = {}'.format(pred_poses.shape))\n",
    "t_loss = np.asarray([np.linalg.norm(p - t) for p, t in zip(pred_poses[:, :3], gt_poses[:, :3])])\n",
    "q_loss = np.asarray([quaternion_angular_error(p, t) for p, t in zip(pred_poses[:, 3:], gt_poses[:, 3:])])\n",
    "\n",
    "print('Translation(T) error in meters and Rotation(R) error in degrees:')\n",
    "print('T: median = {:.3f}, mean = {:.3f}'.format(np.median(t_loss), np.mean(t_loss)))\n",
    "print('R: median = {:.3f}, mean = {:.3f}'.format(np.median(q_loss), np.mean(q_loss)))\n",
    "\n",
    "# Save for later visualization\n",
    "pred_poses_train = pred_poses\n",
    "gt_poses_train = gt_poses\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "=== Test Validation Dataset ======\n",
      "gt_poses = (384, 7)\n",
      "pred_poses = (384, 7)\n",
      "Translation(T) error in meters and Rotation(R) error in degrees:\n",
      "T: median = 0.041, mean = 0.052\n",
      "R: median = 2.215, mean = 2.545\n"
     ]
    }
   ],
   "source": [
    "print('\\n=== Test Validation Dataset ======')\n",
    "pred_poses, gt_poses = model_results_pred_gt(model, validation_loader)\n",
    "\n",
    "print('gt_poses = {}'.format(gt_poses.shape))\n",
    "print('pred_poses = {}'.format(pred_poses.shape))\n",
    "t_loss = np.asarray([np.linalg.norm(p - t) for p, t in zip(pred_poses[:, :3], gt_poses[:, :3])])\n",
    "q_loss = np.asarray([quaternion_angular_error(p, t) for p, t in zip(pred_poses[:, 3:], gt_poses[:, 3:])])\n",
    "\n",
    "print('Translation(T) error in meters and Rotation(R) error in degrees:')\n",
    "print('T: median = {:.3f}, mean = {:.3f}'.format(np.median(t_loss), np.mean(t_loss)))\n",
    "print('R: median = {:.3f}, mean = {:.3f}'.format(np.median(q_loss), np.mean(q_loss)))\n",
    "\n",
    "# Save for later visualization\n",
    "pred_poses_val = pred_poses\n",
    "gt_poses_val = gt_poses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "\n",
    "def calc_poses_params(poses, pose_format='full-mat'):\n",
    "    \"\"\"Calculates min, max, mean and std of translations of the poses\"\"\"\n",
    "\n",
    "    p = poses[0]\n",
    "    allp = extract_translation(p, pose_format)\n",
    "\n",
    "    for p in poses[1:]:\n",
    "        allp = np.vstack((allp, extract_translation(p, pose_format)))\n",
    "\n",
    "    p_min = np.min(allp, axis=0)\n",
    "    p_max = np.max(allp, axis=0)\n",
    "    p_mean = np.mean(allp, axis=0)\n",
    "    p_std = np.std(allp, axis=0)\n",
    "\n",
    "    return p_min, p_max, p_mean, p_std\n",
    "\n",
    "def set_3d_axes_limits(ax, poses, pose_format='quat'):\n",
    "    p_min, p_max, p_mean, p_std = calc_poses_params(poses, pose_format=pose_format)\n",
    "    ax.set_xlim(p_min[0], p_max[0])\n",
    "    ax.set_ylim(p_min[1], p_max[1])\n",
    "    ax.set_zlim(int(p_min[2] - 1), p_max[2])\n",
    "    return p_min, p_max, p_mean, p_std\n",
    "\n",
    "def extract_translation(p, pose_format='full-mat'):\n",
    "    if pose_format == 'full-mat':\n",
    "        return p[0:3, 3]\n",
    "    elif pose_format == 'quat':\n",
    "        return p[:3]\n",
    "    else:\n",
    "        warnings.warn(\"pose_format should be either 'full-mat' or 'quat'\")\n",
    "        return p\n",
    "\n",
    "\n",
    "def draw_pred_gt_poses(pred_poses, gt_poses):\n",
    "    fig = plt.figure(figsize=(8, 8))\n",
    "    ax = plt.axes(projection='3d')\n",
    "\n",
    "    ax.set_xlabel('$X$')\n",
    "    ax.set_ylabel('$Y$')\n",
    "    ax.set_zlabel('$Z$')\n",
    "    ax.view_init(50, 30)\n",
    "\n",
    "    all_poses = np.concatenate((pred_poses, gt_poses))\n",
    "    p_min, _, _, _ = set_3d_axes_limits(ax, all_poses, pose_format='quat')\n",
    "    \n",
    "    draw_poses(ax, pred_poses[:, :3], proj=False, proj_z=int(p_min[2] - 1), c='r', s=60)\n",
    "    draw_poses(ax, gt_poses[:, :3], proj=False, proj_z=int(p_min[2] - 1), c='b', s=60)\n",
    "    for i in range(pred_poses.shape[0]):\n",
    "        pp = pred_poses[i, :3]\n",
    "        gp = gt_poses[i, :3]\n",
    "        pps = np.vstack((pp, gp))\n",
    "        ax.plot(pps[:, 0], pps[:, 1], pps[:, 2], c=(0.7, 0.7, 0.7, 0.4))\n",
    "        \n",
    "    plt.draw()\n",
    "    \n",
    "def draw_poses(ax, poses, c='b', s=20, proj=False, proj_z=0, pose_format='quat'):\n",
    "    \"\"\"Draws the list of poses.\n",
    "    Args:\n",
    "        ax (Axes3D): 3D axes\n",
    "        poses (list): Poses list\n",
    "        c: matplotlib color\n",
    "        s: matplotlib size\n",
    "        proj (bool): True if draw projection of a path on z-axis\n",
    "        proj_z (float): Coord for z-projection\n",
    "    \"\"\"\n",
    "    coords = np.zeros((len(poses), 3))\n",
    "    for i, p in enumerate(poses):\n",
    "        # coords[i] = p[:3, 3]\n",
    "        # coords[i] = p\n",
    "        coords[i] = extract_translation(p, pose_format=pose_format)\n",
    "\n",
    "    # Draw projection\n",
    "    if proj:\n",
    "        if len(poses) > 1:\n",
    "            ax.plot(coords[:, 0], coords[:, 1], proj_z, c='g')\n",
    "        elif len(poses) == 1:\n",
    "            ax.scatter(coords[:, 0], coords[:, 1], proj_z, c=c)\n",
    "\n",
    "    # Draw path\n",
    "    ax.scatter(coords[:, 0], coords[:, 1], coords[:, 2], c=c, s=s)\n",
    "\n",
    "\n",
    "def draw_poses_list(ax, poses_list):\n",
    "    \"\"\"Draw list of lists of poses.\"\"\"\n",
    "    for poses in poses_list:\n",
    "        draw_poses(ax, poses)\n",
    "\n",
    "\n",
    "    \n",
    "# print(pred_poses_train)\n",
    "# print(gt_poses_train)\n",
    "\n",
    "# Draw predicted vs ground truth poses\n",
    "draw_pred_gt_poses(pred_poses_train, gt_poses_train)\n",
    "plt.title('Train Dataset')\n",
    "plt.show()\n",
    "\n",
    "\n",
    "draw_pred_gt_poses(pred_poses_val, gt_poses_val)\n",
    "plt.title('Validation Dataset')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}